Should the location of this file change please use the keyword "divdev7"
in a search engine to find it again. For convenience using search-engines,
use keyword divdevtips to target these files.
A number of the pictures are now apparently not downloadable, because the hosts have disallowed
remote linking although not saying so. To view them , you have to remove the picture file name
from the picture URL and put this .htm file name in its place
and scroll down to the relevant pic.
Renovating pot with push-pull switch.
Due to hardened lubricant on the DPDT slide switch contacts
If anyone happens to know a generic name or name of the manufacturer?
push pull sw on pot
Logo? is a Studebaker logo "lazy S" in a circle, M in a circle and then FE
also use of J without a top bar as eJ.
Used in Carvin and Trace amplifiers.
If I was doing it again I'd feed a wire through the point marked "<"
over the boss "B" that connects with the axial shaft of the pot , and twist
over the body to keep the boss from moving and dislodging the spring.
And no need to bend down the rear closure of the housing, 1mm graff paper.
push pull sw on pot
Didn't lose the spring but just in case dimensions are 0.55mm wire, 9mm
arms , ends 9 mm apart, 2 turns minus angle between arms and 3.3mm o/d of
ring.
At mid position , maximum spring force , so not stable position marked
between the > and < giving the flip-flop action. One section moves one way and the other
moves the other way and vice versa.
push pull sw on pot
The flattish section with 2 nibs goes opposite the boss part of the other
section.
Hold together enough to add a small dot of hot-melt glue on each corner on
the surface that slides in the housing, there is a bit of space there.
Position the 2 parts as in the knob pulled out posistion , before gluing
and check that it will go in the housing and over the pot shaft end.
Now for the spring.
Place the glued parts in an engineers cramp or similar to give some more
temporary closing force, and easier to hold.
Push spring, ring first, in through the side with the narrow slot leaving
the arms outside enough to grab one , to force into the slot in the flattish
PTFE-like part.
Push both arms in , restrained by finger flesh in the larger hole of the
other side.
With a needle point or similar, through the narrow slot, force the other
tail/arm into its slot.
Remove the cramp carefully , in case the hotmelt is not suifficient, and
so loose the spring.
Replace in the housing and wire down as described earlier , just in case
the glue spots fail before finitshing off the slide switch refixing.
Break glue spots by pushing pot shaft.
Which leaves the question - how do they
assemble originally?. Something to do
with the small diagonal slot in the casing and
the large gap between the 2 white pieces that
I did not really find a use for, and its
non functional in use.
Back to making rubber drive bands.
First some data for , mainly , flat bands.
Trio KP 1022 record deck 640mm x 5 x .5mm
Marantz TT 120 400mm stretched to 430mm , 1.3mm
Toshiba SRB20 660 to 690 mm, 5x.75mm
For some cut to 4mm wide 0.85mm bike innertube
50 percent elongation for 300gm load.
On one such join it took a proof load of 900 gm
Obtained a Kichen Craft pasta machine, for the cutter
section although the rollers with variable gap
could be useful, clever mechanics, offset-centre roller inside a roller ?.
Similar Imperia m/c
http://www.verynice2.com/images/prod/PASTA-590.jpg
2 sets of intermeshed roller cutters for cutting
bicycle and motor bike inner-tube down to neat strips for rubber drive band
making (then standard "bean slicing" to cut them down narrower). 2mm ones at
top of pic ( although 4mm for cutting rubber for some odd reason) and 6.5mm
below. Probably 4mm slicing because the material is too thin,
for 2mm pack out with sacrificial cardboard , before feeding into the slicer.
You can only neatly bean slice once you have neat parallel sided strips of
rubber
It is all metal
construction and a neat hidden innternal mechanism, I've not thought how it
works, for varying the gap in 10 lock-down steps and still allow
contra-rotation of the steel rolls.
Good strong construction, overmade
for pasta. For this use the interlink is far
too weak and mount the cutter section in a vice to
use . Pad out with blocks either side of the unused
pair of cutter rollers, as the frame is too
irregular to mount in a vice. For some odd reason cutting
bicycle inner tube rubber in the 2mm wide cutter tends
to produce 4mm wide belts with no or partial scored
rather than cut centre cut. Allow for overlapping the rubber.
Vulcanise cement together and lightly compress with a clothes peg,
that has 2 pieces of manicure sandpaper pre-glued in the gap.
Then reinforce the join with some 0.5mm rubber cord stripped
from some bungee or habberdasher's elastic, knot on the outside.
Made some slicers from thin Al tube of various sizes. Cut a length about 10
mm long and slightly flatten so internal dimension is slightly
less than the width of rubber needing slicing. Will only successfully
halve rubber, not 1/3 to 2/3 slicing. With thin disc in Dremmel
cut a slanted slot centrally to take a razor blade. Pull the rubber through
the tubing, by needle and thread. Preliminarily slice the
start of the strip by hand so the razor can be mounted in a
slot in some wood, at the rake angle. Then with gloves and goggles
on pull the strip through , keeping a bit of back tension to
keep the tube in place, but otherwise does not need careful mounting/holding.
If single glue joint is not enough. Make up a band and then a
larger one of about fully extended length to go around
the first one 180 derees staggered. Place a couple of cylinders
on the arms of some old lockable vernier calipers to
stretch the first one, then glue the second to the first
and stretch a bit more while the rubber cement goes
off. If a non elastic and protruding lump is ok then one band scarf joined. On the inner
scarf slope grind 2x 0.5mm slight slots just inside each end of the slop
by bending the rubber in the middle, slots just about 1/3 width.
Enough to loose some thread , sewn through to the outer side
of the band and knotted. So 2 threaded loops, perhaps slots
on the outside as well. Finfd a cylinder , a bit bigger than the smallest
pulley size, wraop with some PTFE. Slightly stretch thejoint
part of the band around the cylinder and anchor off.
Dab some super glue on the exposed edges and let capillary
action take it in.
A source of square section rubber (usually white though)
is pulling out the cores of bungee/shock cord,
sometimes it is in the form of strips but maybe not rubber.
Another possibility looked into is plaiting
http://www.strawcraftsmen.co.uk/project10.php
www.strawcraftsmen.co.uk/pdf/7_straw_flat_plait.pdf
A reply from there as to whether Dunstable plait
could be returned on itself to make a loop.
"Yes, just start with loose ends (rather than by folding or tying) and finish
by overlapping the 'ending' straws over the 'beginning' straws. It will be
bulky at that point and that is very difficult and skilled to avoid. You may
need a needle to work the straw ends through."
But small section round rubber cord is not easy to
work with compared to straw. Tried 7-plait of rubber
but not very flat result. Requires a soft clamp
at either end of the run. If small but rectangular section
is available/made then may be worth returning to
for flat belt making. May try making a 1x1mm to 1x0.5mm
bean slicer sometime. Some other untested ideas.
Self amalgamating rubber tape for thin rubber material.
Maybe possible to cut holes along the centre of a wide band, with too
much tension, to reduce the tension and still function. For
thicker bands, double up the above construction
by vulcanising gluing one inside the other.
Thin but wide rubber bands. A source could be
vinyl deck bands cut and glued/ground as above.
A "bean slicer " in the thickness sense does not work.
For 4mm wide bike innertube rubber (0.8mm) set a turned
pin DIL socket as a guide under a dremmel and grind stone.
Put some tension on the rubber and pull through under the dremmel .
Repeat with less and less tension till right thickness.
Another method that may work is set up a ring of PTFE or silicone covered
ring and then core material from bungee ot habberdasher's elastic
wound round and round to build up a width and then bond with
vulcanising cement, works with round section rubber but square would
be better, dust with talcum powder afterwards.
Silicone bake-ware cup cake moulds. Ok crinkled and conical but
I'll give it a go. Cut the base out, stretch over a cylinder of some
sort and cut a ring 5mm wide. The ones I used gave 1.2mm thick x 5mm
wide band from the widest part. Stretch is 70 percent for 1Kg, single
piece, about 200mm circumference for about 240mm stretched.
The smaller end would give about 130mm, small moulds
for smaller bands, 1.25 to 1.75 inch diameter and tiny ones 25 to 35mm .
Silicone bakeware small loaf one circum 630mm. 100 to 240 nn
Ring mould would give quite flat bands in 100 to 120 mm and 220 to
240mm and pin punching on one edge for anything in between.
Tried with a motor and bulbous pinion and a receiving pulley,
it did not ride up and down the pinion noticeably, stayed central, but
to fingers a lot of vibration. I tried with a proper band and
much the same vibration - I'd not realised how much motor
jerkiness was transferred to the receiver. Swapped with an ex-tape
recorder flywheel pulley and no worse vibration noticed with the
finger test between proper band and cockled cup-cake band.
So why not try flat rubber sheet cut to a ring and relieve
some of the inner tight side by piercing before "up-setting" and
using as a drive band.
My leather punch does not go down small enough diameter so
I made a punch from an old office size stapler, removing
the top plastic cover piece.
Found a 2mm diameter steel roll pin and ground a conical
cutting face. Roughened the surface of the roll pin and the
exposed active end of the stapler. Tied the two top section
together (eventually soldered together ). With a 150W soldering
iron soldered the roll-pin into the vertical gap at the head of
the stapler and reinforced with some fine copper mesh.
Worked a treat, giving
holes something like 1 to 1.2mm diameter in 0.8mm rubber sheet.
As a try out, cut some 2mm wide rubber and made holes along
a part of it about 4mm apart. Stretching , the unholed run,
for 140gm had had an elongation of 88 percent and the holed
section 106 percent. Will have to make a double scalpel
cutting "compass" but for the moment cut a 5mm wide ring of
rubber, 67mm small diameter.
Tried under tension and the difference , with a force gauge
each side was about 15/25gm to flip up the edges.
I pierced 1mm holes 1mm in fromn the edge, 5mm apart all round
of the innermost edge. Repeating the lift test then about 15/15gm.
Tension required before the sag was taken up was about
180gm doubled-up so 80gm single, so could not be used with
band tension that or less. Again tested with a flywheel ,
driven pulley , and no riding up and down the bulbous
pulley and no finger observable vibration, will have
to wait , to be done in ernest, for the next situation of requiring a flat band
that is only obtainable with a minimum order of 30 GBP.
A sort of working cutting compass, needs the right sacrificial
material either side of the rubber sheet to prevent rucking.
2 scalpel blades either side of a brass nut , as a spacer,
clamped into the jaws of an engineer's cramp. Then 2 large brass
screw-down cable terminals/ binding posts , hi-fi type , with knurled clamps.
One fixed to the main postional threaded bar of the cramp
and then the other terminal clamped to that one to go through
a pierced hole in the rubber. Then ptfe washers , large washers
and sacrificial padding material. Perhaps large netal disc with central
hole with double sided foam type carpet tape laid over , in path
of compass at least, will protect scalpel tips and stop rucking.
Alternatively another metal disc slightly smaller than compass path
and perhaps a band of flexiform grommet around the periphery
to ensure holding the sheet firmly near the cuts , when clamped to
the main disc,
UBL , China has a good range of this silicone bakeware, including
0.9mm thick flat sheet 16 x 11.25 inches, 8 inch and 10 inch
diameter pie moulds, muffin pans (75mm to 52mm unstretched diameters over
slant height of 30mm and 1.5mm thick ) mini muffin 40 to 56mm
diameter 0.75mm thick, butterfly cake mould would give 530mm
circ and 1.25mm thick, 52mm diameter cake mould and
pie moulds that could give 8.5 inch or 11 inch diameter bands
and poaching cups. Sugarman Jamie's kitchen baking sheet
is 370 x 270 x 0.5mm and other in range is thinner also.
Another one a number of small pie ones in one surround
giving 60mm bands potentially. Stretch over vcr cylinder, baked bean
cans etc and turn around on a flat surface with a razor clamped
at various heights to cut rings, widest diameter ones first.
First go at using Silicone bands for an Aiwa AD F450 cassette tape,
see repair3.htm file, long term durability not known yet.
The 0.9mm flat sheet elongation for 4mm x 0.9mm is 200 gm for 65 percent.
Unfortunately tends to be lurid colours, some permanent ink felt pens
will dye after heating with hot air and some need multiple
goes and still not that permanent. Neoprene rubber sheet, 1mm or thicker readily
available from the likes of Portmere and 0.5mm orderable but
12 GBP per square metre and even heavy duty black type
kitchen latex gloves maybe useable perhaps a bit
better than office rubber bands, white on one surface
so "inside" in use or felt-tip colour, 6mm wide , 0.5mm thick has single strip elongation
of 20 percent for 200 gm. Cut with razor and straight edge
clamped down but fold in another position to the moulded in fold or the ends will joggle.
Another rubber cutting method.
Some "worm cutting" triple scissors from fishing shop eg make Bexfield.
Consist of 3 pairs of blades ganged together. Bolt and nylock nuts so
easily adapted to narrower spacers 2 to 5mm range and remove one pair
altogether.
For measuring drive band path where access is difficult. Try some polyester
gummed tape cut down to 5mm wide or so. Dust most of the gum with talcum
powder and stick to pulleys and turn around by hand grabbing the tape when it
re-emerges. If sticking to an existing but stretched band
compensate for the band thickness
Latest cutting compass starting from a standard student type
pencil compass. Cut the point down to just enough to enter the metal clamp ]
of a choc-bloc connector core. 2 scalpel blades separated by a piece
of brass (i Used UK mains "square" pin plug pin) , just space with a bit of grinding
to go in the pencil hole and a switch back nut to even out pressure on the
inner surface. Disconnect the top screw and separate point and pencil
parts with 2 pieces of holed Al smmall bar and bolts to make trammel-like.
Set the free end of the choc connector in the centre of a ball race interference
fitted into a large thick metal disc. Glue the choc bit into the
ball race at an angle of about 45 degrees to the plate and then another 15 degrees
of set on the scalpel blades is about the right cutting angle for the rubber.
Cut central hole in the rubber sheet and fix to the plate with double
sided tape to do the cutting. Advance inwards for a series of bands. Perhaps
scoring rather than complete cuts and finish off cutting through
by hand. Use tangential direction of force at the scalpel blades for
cutting with accurate track around.
My test for whether bypass switches, found at the external
loop Send or Receive sockets between preamp
and power amp, need attention , even
before dismantling the amp.
Take one cocktail stick/toothpick . Hold at one end between thumb and index
finger and poke the other end in the socket - if the contact breaks before
the cocktail stick breaks ,about 150 gms over length of 65mm,
needs attention.
Where twizzle stick fails to locate intermittant fault.
Try heating with a hot air gun after preliminary
heating with use of a dummy load.
Try vibrating boards etc with a hand-held engraving
tool with a nylon bolt fitted in place of the metal tip.
Sharpen the bolt to a point with a pencil sharpener.
Beefing up 1/4 inch jack sockets
The notorious problem with band/stage amps, the guitar/mike leads making
poor contact inside the sockets.
I don't see the point in replacing with much the same, that will also weaken
at the bend, over a few years of use.
The tip contact has to repeatedly lift over the tip and then drop down into
the groove. The ring contact/s don't have to do this so rarely a problem
there. Coupled with more "leveraged" movement at the tip with lead tugging.
Requirements some 2mm silicone rubber cord from a fishing/angling supply
shop
Hama/Perler/Pearler beads about 2.5mm internal bore, (small coloured plastic
rings) for making "mosaics" on a sort of pegboard from kids/craft shops.
The outside diameter of the bead/s size chosen seems to move easily in the
gap in the moulding of the socket for all the ones I've tried it on, mono and
stereo.
Bit of looped copper wire as a "needle threader"
Artery forceps, again from an angling shop, or similar tool.
Thin nose pliers
Assuming chassis mounted , then cut a piece of cord 100mm long.
For pcb mounted ones then about 120mm and suitable 4x 2.5mm holes drilled
through the board. Being rubber you can safely drill through a track or two
if constricted space (caution if for speaker sockets).
No need to desolder any wires.
Pics below are against 1mm square graph paper.
1/4" socket bodge
Some people may think that the socket shown has a carbon insert as contact,
not the case, a dimple that the lighting suggests is black.
First one takes about 10 minutes to complete and about 5 minutes each, when
familiar, for the chassis mount socket type.
First shows the "needle threading" action.
Second is the main assembly, thread 2 beads (all red in the pics)
onto the doubled-up cord leaving
a loop.
Place a third bead in this loop and pull the tails to tighten in place.
Place in orientation over the contact as in the pic.
,placed at the "hinge" end of the contact.
Take one tail around the pin edge furthest from the mounting panel, under
the socket and back up the other side of the other pin.
Grab this tail with pliers, 10mm back from the free end, and slide through
the bead in the loop, the cord will be under a bit of tension.
Anchor the free end with forceps.
Lace the other tail to form a cross under the socket
Reuse the threader , pushed through the bead in the loop.
Trap the free end of the second tail in the threader, a few mm back, and
pull back through the bead.
2 such pieces of cord will anchor themselves quite well in a bead but to be
sure, balance up the 2 free tail ends and anchor both together with the
forceps, while finishing off.
You can then pull the bead through the loop to its final locked in position.
Position all the sections of cord and top beads so near enough equal tension
throughout.
Remove the forceps and melt a spot of hot melt glue between the 2 tail ends
The insertion and pulling-out force for any 1/4 inch plugs is then noticably
greater and much more reliable contact in use.
Makesure the hot-melt or beads are not near any hot components when
reassembled.
And guitarists can throw away those cans of contact spray
My guitarist clientele are much approving of the technoque although it seems
somewhat crude, but then the action of these sockets is pretty crude.
Beefing up a different type
1/4" axial type
no info of any sort on the socket, no maker or generic term I'm aware of.
This one from an Ashdown effects return with dodgy bypass switch.
There are 2 places in the plastic body that drilling through coincides with
holes in the metalwork almost as though designed to wrap some silicone
rubber cordage (blue in pics) around the switches to increase the closure
force, located in place by curves in the contact parts. Then cleaning the
contacts. One hole in the metalwork is a bit too small for the
shown cordage and needs pushing through, directed down through
the inlet and pull through enough to then push back and around the
contacts.
Tip contact to the left and ring contact at bottom of first image.
Beefing up open Switchcraft type.
Twist some tinned wire around the 2 outermost solder tags , twist together
and solder in , forming small radial projections.
L>oop the cordage behind the switch contact and around the
2 outer terminals , so avoiding the plug touching the cod.
Tie into a loop.
Attending to a different type
Logo Y with a bar and a right rotated C in a large C almost a circle .
Patent nos 3536870 976235.
Has a clear plastic panel on the face away from the pcb.
Drill ionto the melted plastic rivets, then open out
, clearing the rivet with a matrix board cutter.
Lever up the inlet end while desoldering the
pin at the far end that goes to the ring contact.
Clean and pack with bits of silicone rubber where it will fit.
Replacing band amp 1/4 inch jack sockets
pcb mount type ones with plastic barrel so barrel bends and pins break when
they try pole vaulting over the guitar lead.
I replace them with upside down tagged ones (plenty of give in the wiring
then), but require awkward insulated crossing-over and under of the
additional wires from tags to pcb. For high quality
(proper contacts) Cliff type ones anyway. Its an easy job releasing and
removing the metal parts and replacing on the opposite side of the barrel
and then straight runs of deliberately slackish wire into the pcb
Poor digital TV digibox reception on a few channels.
In an otherwise good reception area and good aerial.
Try placing a non metal box/packing under the box to
take it further away from any recorder or tV it is resting on
Returning a band amp back into its case
Without tearing up the heavy duty aluminium foil screening stuck to the
sides. Just what you don't need rattling around in the amp ,
bits of aluminium foil.
The solution I use is a sash-cramp used in the reverse sense
across the back of the casing to open the gap a bit and a sheet of celluloid
draped over each side , removed afterwards.
Preceeded by cutting off jagged corners to the chassis, for personal safety
reasons, if nothing else.
To determine value of burned out resistors.
Well approximate values anyway as excessive heat tends to
corrupt the value before going o/c.
For types that have split appart,desolder,connect
meter to wire ends and press the broken ends together
with thin aluminium foil in between.For spiral wrap types
scrape the laquer in a line along the resistor to expose
the conductive track,with prods measure the maximum
resistance from either end to the point of break and add
a guesstimate for the width of the missing section.
Worked example:
1 watt metal oxide resistor discoloured rather than charred.
Apparently reading brown brown black or 10 ohms.
The "black" band at the centre of the body will
be the most affected and perhaps the outer brown one
was brown originally, as least temperature affected.
Scraping axially to make contact there were 5 sections of about
200 ohm and one for the break so say about 1200 ohms
The value , this time from schematic, should have been
brown green red or 1.5K
Unusual metal machining process
Called skiving
Aluminium heatsink used for mounting electronic components to , to conduct
away heat to the air . Sort of chisel cutting advance and repeat.
http://www.diversed.fsnet.co.uk/chisel_vane_heatsink.jpg
a steel rule laid across it,
plus a close up showing the slightly angled cuts
http://www.diversed.fsnet.co.uk/chisel_vane_heatsink2.jpg
The remaining uncut spine is about 4mm wide and the original bulk of metal
must have been about 9mm wide to be cut down, each side, to a 4mm spine.
The fins are only 0.7mm thick and no sign of machining marks just a slightly
granualar looking surface on the concave side of each vane and a smooth
surface on the convex side.
the fins are about 27mm along the curve.
A milling process would not leave those 5 degree or so off axis cuts along
the spine.
This is an end on view
http://www.diversed.fsnet.co.uk/chisel_vane_heatsink3.jpg
showing the granularity which probably suggests a shearing action.
Looks as though the first operation was 4 knife cuts about 2mm deep
(assuming 9mm thick slab to start with ) which when chiselled? up give those
(2 sense) tapering ends to the knife cuts
The lines on the spine probably show a guillotine to chop to length.
The back face of these vanes is quite smooth in appearance mainly.
A very efficient process in the sense of no material loss in machining.
A small exploration of pulleys
A picture of a random collection of pulleys
http://www.diversed.fsnet.co.uk/pulleys.jpg
plastic brass and aluminium for use with rubber drive bands in tape
recorders, VCRs etc. They all have a small groove at the root of the V
section of the pulley. Anyone know the function or reason for this groove?.
I thought it may avoid an aquaplaning type possibility of air
getting trapped under the rubber band snd so wow/flutter possibiliy.
But more likely it is to avoid the corner of a belt bottoming out
in the groove and loosing friction contact with the V faces.
Shaping a cylindrical carbon brush from larger square or
rectangular section. Grind down to square and then
slightly oversize octagonal. Then with appropriate size
nut spinner or socket set socket held in fingers spin
it around the carbon to shape to round.
A fudge for replacing a broken pot stem , say on a large
mixer panel, to avoid stripping apart.
Find a bit of plastic rod, drill axially to
take a long self tap screw. Drill into the remainder of
the stem. Screw barrel and screw into the hole, it
may break past the end stop but hopefully
not graunch up.
Silica gel crystals
If sugar white in colour then spent/hydrated. If cream
to yellowish colour then active/de-hydrated.
Activate by keeping in an oven for an hour at some
temperature over 100 degree C.
Tag terminal desoldering technique
I find desolder braid or vacuum suckers only work well enough on flat pcb
pads not bulbous solder and wire/s around tags of older stuff.
Firstly assuming that the component lead length is not enough to cut and
re-use, further along the wire. So definitely a matter of desoldering and
re-using what is there, maybe more than one wire , all quite properly,
looped through the tag before the original soldering.
The first thing I do is fix some small (ratchet clamping) medical artery
forceps / angler's hook remover tool/s on the exposed bit of the metal of
the wire/s up close to the tag to act as heatsinks to avoid melting the
sleeving , due to prolonged solder-iron heating of the tag. Then with what I
call a needle-probe, a large sewing needle, set into wooden dowel handle, I
explore while melting the solder blob and unhook the "free" end of the
wire/s and then pull through using the forceps.
Cleaning desoldered SM pads before resoldering
To leave a clean solder surface.
After removing the original place a bit
of plumber's PTFE tape stretched over the pads
and remelt the solder , with the iron, through the PTFE
Taking apart "white goods" eg microwave oven, toaster etc
As there is often jagged metal , use a pair of stout
gloves. Especially if having to pull on some section
or use force on a screwdriver etc.
TO220 , TOP3 etc mounting bushes/washers
Not the usual nylon or whatever ones that I never trust for high
temperatures (soften and deform over time, loosening thermal connection,runs
hotter, melts more ......
Does anyone make them from whatever the plastic/mineral? is used for the
body of power transistors themselves, what is that material , what is it
called? . Maybe because of brittleness they would have to use smaller
screws/bolts allowing larger washer thickness in the hole.
I've got in the habit of cutting off the non-metal mounting parts of TOP66
type transistors, opening the holes of TO220 or TOP3 and using them as
washers.
Hacksaw through the body just down from the hole,
cut into the copper, to release around the hole, to remove the "bush".
Then open out the reciving hole using "drill saw" (
part drill part rotary rasp/burr) rotozip, torx wrench
of Allen key form (also using the nurled part) file etc
finished with a countersink bit to clean the edge.
Use a stout glove to hold the transistor.
That is if its not a practical solution to hold down
with a spring clip. Too crowded is the usual problem.
But if drill a hole for spring clip mount make sure
you mask off the area around the drilling to avoid
getting swarf into the electronics.
Just because they call it high temperature, does not
necessarily mean it does not compress/creep in normal but highly elevated
temperatures, leading to premature failure of device.
I'm after some insulated and mechanically strong enough
alternative to plastic without having to drill into heatsinks for spring
retention.
If a source of PTFE rod suitable for the insert sleeve part
when drilled,
unless having to restrain abnormal lateral forces .
I wonder if fibre reinforced phenolic board would be
suitable for the washer, I'll try heating some. Of course that washer part
does not have to be circular and cutting it square with a hole would be
easier.
The washer part could be a disc or 2 of that stiff plastic material found
inside power supplies where there is a chance that some component will touch
the casing, next to a steel washer under the bolt/screw head.
For that matter I don't see a problem of using a mica disc or 2 then a steel
washer for that part.
Cutting down height of Al cased plastic knobs
Or presumably plastic knobs also. Owner of Roland piano had a flight case
made up for it but he did not allow for the knobs. Hacking into the foam
still leaves the knobs pressed against the outside face of the case.
Replacement knobs not possible as would have to be only 8mm high and do not
exist that shallow.
How to mount in a lathe to cut down ? Cutting tool sufficiently sharp but
grip the knob in the jaws too tight and the Al of the knob deforms. Too
loose and it slips of course. Jubilee / hose clip around it and not enough
space between the jaws for the clip boss.
Nylon cable tie worked well enough to form a neat level line to then hand
hack saw / and bench-grind down to, but not enough force to fully lathe cut
the height down.
Replacing Jubilee clip screw with a grub screw, no goer, all threads are too
fine and needs a bearing surfce flange.
But cutting the bolt-head off the existing one, leaving the flange and
cutting some off the other end and cutting a slot across for turning with a
screwdriver. Cut off the retaining tang parts of the clip to shorten (no
longer holds the bolt in place) should work down to about 12mm diameter knob
, holding securely in a standard 3-jaw.
If cutting tool is sharp enough , then mounting the knob on a
piece of scrap pot shaft of right form should work, also for
taper/conical form vertical faces.
Loose 1/4 inch spade crimp connectors
Find a pair of small nylon cable ties with the tails about
0.75mm thick. Glue them together side by side. Place
in the socket before squashing with parallel action pliers.
Pull out and re-use.
Desoldering hook-up wire
from binding posts, turret tags, valve base terminals etc especially now
with use of lead-free solder on otherwise 1950s wiring practise.
Because of the extra temperature and heat involved it is no longer possible
to grab the bared end with thin nose pliers ,to sufficiently heatsink and
help in removing the wire. The sleeving will always melt and split etc. Any
tips for doing so ,neatly, where there is not enough slack to cut the end
off and resolder as a shorter wire.
The following does not stop the melting of the ends , but stops the
splitting / obvious raggedness, any other advice?
Some PTFE , through chassis , press fit vias/terminals, for the PTFE only or
some ptfe rod. Drilled axially to a range of diameters and split axially
with a razor and then clamped to the sleeving with a scaled down clothes peg
(apparently for suspending christmas cards) from toy/craft shop. Then soldering iron and dart point to extract the wire.
Anyone freeze the sleeving ?
Before de-soldering colour coided wiring
For marking the board with single letters , to denote colour
of wires before desoldering. Because of grey/green, brown/black/blue,
purple/pink - I use V for vert, T for tan, N for noir,L for lilac (purple)
Bodge repair for failed conductive epoxy joints
or where LCD zebra strips have lost contact because
pcb has warped slightly or whatever.
Push the section back with folded up
strip of kitchen pan scourer, held in place against the
casing / bracing and cable ties to pcb or something.
The coarse open random
mesh of plastic fibres is insulating and
quite consistently compression springy.
RoHS components with tin tinning of pins
For using tin tinned new RoHS components destined for both old boards
and Pb-Sn solder or new boards and lead-free silver solder, burr the tin off with a
"diamond" cintride burr on a Dremmel, before tinning with the intended
solder. A solid layer of tin is asking for trouble - turn to
grey dust tin-pest.
Silver does not supress the growth of tinpest , unlike lead,
but it is much reduced compared to the cheaper lead-free solder.
Tin pest can only start to occur when the
temperature is at or below 13.2 degree C, ie unheated room
temperature, so garage or shed storage can start the process.
Once started then it is self catalising. For temperatures much lower
than 13 deg C then the process is very much faster.
Temperature indication by effect on plastic part labels?
Valve amp with the three main lumps made by the same manufacturer and the
same sort of 0.05 mm soft white plastic part labels on the outside , so PVC
or polythene perhaps.
Output Tx label as original , choke just starting to show signs of
ruckling/cockling, and the mains transformer noticably ruckled , still same
white and footprint is the same ie the edges are in the original position.
So mains Tx has got the hottest presumably, any guesses as to what sort of
temperatures, and what if the label was seriously ruckled with distorted or
unreadable script or discoloured and shrunken footprint ?
From trying to find thermochromic materials, this is some data I got along
the way, from trying Mylar film (or metalised plastic anyway) ;
bubbles/wrinkles 75 deg C, shrinks 80 degree C and shrunk to 1/4 95 deg C so
pvc and polythene probably similar
Storing epoxy glue.
If both tubes are kept together in a box
make sure they lay in opposite directions
so any leakage does not interact.
Making an adaptor , for a reversible cordless drill,
for undoing and doing up the hundreds of bush nuts
on large mixers.
Find a suitable double ended box spanner (that use a tommy bar)
that fits the nut and has plenty of length to take the pot shaft.
Then a 1/4 inch drive, ratchet socket , suitable size to just not
fit into the other end of the box spanner. Gringd 6 flats
, slightly tapered so it will now go in the box spanner.
With a long throat vic squash them together, laying
along the vic threaded bar, to guarantee alignment.
2 common sizes for a 15mm nut find a 14/15mm box
spanner and a 12 mm surplus socket.
For 7/16 inch nut find a 7/16 / 1/2 inch box spanner
and force a 9mm socket into it.
Then a standard 1/4 inch square to hex adaptor to fit the
drill.
Small pot knobs for 4.6mm spline shaft pots as used in audio mixers
Of the thousands of salvaged knobs I've collected over the years not a
single one for 4.6mm . Unfortunately all mixers that come my way for repair,
get returned, and no scrappers.
Fudge is to pack out small 1/4 inch recess knobs with
2 layers of electrician's 0.6mm thick earth wire sleeving. Cut axially and
both coincident for D section ones and overlapped for round ones. Bot very
professional but holds together ok on the splines without slipping
Plastics failure mechanism
I've never seen reference to this mechanism, so transcribed and placed
here.
From (UK) Hacker GP42 record player of 1966, technical bulletin, not dated
but probably close to '66
"Investigation over a period of time of reports of knob breakages have shown
that the adhesive used to fix the chrome trim has affected the material and
caused the knob to fracture after a period of time. A further investigation
by the manufacturers disclosed that the material used in one batch of knobs
was an incorrect type. DIAKON was the material specified and POLYSTYRENE was
used in error. Unfortunately there is no way in which we can easily
differentiate between these materials other than by a flame test. "
Where I used to work , there was a flow-chart wall-chart for identifying
plastics by flame test, smell/smoke character when burnt and other basic
characteristics. Something similar
http://www.boedeker.com/burntest.htm Identifying types of plastic
Forming splined form to round aluminium pot shaft
The pattern repeats for only one diameter and a certain
degree of intrusion. Aiming here for 16 around a nominal 6mm spindle.
Proof the concept works. Robbed the steel gripper wheel from a butterfly can
opener, sharpened the points a bit and opened out the central hole. Mounted
in the pipe cutter and produced a neat ring of indents. These were about 2mm
apart and 7 equally spaced around, when cutter advanced far enough in. So a
cog with 1mm spacing should work even easier as long as not too wide. Actual
splined shafts have a 1mm spacing and 16 around.
If nothing else a thin approx 1mm tooth spacing cog from a clockwork clock
mechanism would make an excellent way of marking pot shafts before making
axial cuts with a grind wheel.
I've opened out the slot in a Rolson pipe cutter from 3.5mm to just over 5mm and it will now hold
neatly and freely rotating, a 16mm diameter x 5 mm ball race.
I know I can grind into the outer race with a Dremmel and thin disc
(freehand) and cutting 1mm spaced teeth is possible as proved by cutting
,freehand so rough, some 1mm spaced teeth around a steel washer. Running a
24 TPI hacksaw tightly around a pot shaft produced a ring of about 19 small
indents, so 1mm is about right for 16. Next a bit of trig to find a whole
number of rotations into 16 mm diameter to give about 1mm spacing.
Too many belts in the lathe drive train. Settled on a doubled large cable
tie around the chuck, clamped to the lathe bed, as a brake and
anti-backlash. And marking around the chuck 50 divisions covered with a
needle, as indicator, on an arm to a magnetic base. Pulling at the motor V
belt to step round. Mounted the Dremmel on 3 short pieces of Dexion to set
on the tool post with a 45 degree set. 3 rounds of cuts to get to final
profile.
Next time I will spend more time fiddling around DTI setting in the 4-jaw
before starting as there was eccentricity.
There is a lunate form to the cutting edges, maybe due to flexing of the
0.6mm thick grinding disc or I did not pay attention to changing direction
of cuts on "odd" and "even" teeth on each pass. Perhaps I should have done
another round of cuts but at the previous depth setting to cut into the
lunate form more. Perhaps it was due to axial play on the Dremmel. Anyway it
works well enough for this purpose. No more than 5mm splining or the
tightening force would be more than finger force on the knurled head bolt
and the aluminium of the holding frame would snap with anything more than
finger force. It is possible to re-engage the spliner further along, synched
with the first cut to form extended splines.
In use leave the pot spindle full length so the 2 round rollers of the
existing pipe cutter can react against the extended shaft, and
hold the free end of the shaft with molegrips, or cut
a slot to introduce a screwdriver blade, in case the
back torque is too high and the pot end-stop or something else internal breaks up. Afterwards cut to
length and the diametrical cut. Remember the length is back to the
mounting face, as bush length varies. For the slot
start with thin grindwheel in Dremmel and complet with a hacksaw.
If the slot needs narrowing place some strip in the slot before squashing
or you will have no control it and will fully squash.
16 evenly spaced cuts around the pot shaft
in the bottom view. If not deep enough then at least a good accurate guide
for starting some deeper cuts done freehand with a Dremmel. Didn't think to
measure the diameter of the cutting disc before starting and the others vary
in diameter, but by comparing to them, somwhere between 0 and 0.2mm erosion
of diameter in use
I like a challenge and may learn something along the way. A week ago , if
someone asked me - can you cut a 0.5mm wide slot in the steel of a ball
race, using mundane equipment, I would have said no.
That brown glue stuff
Used for fixing otherwise floppy bits to boards.
Is it hygroscopic ?
IC with 2 glued pins are almost corroded through
but not obvious before removing the IC, and have presumably conducted
moisture, capillary fashion, into the IC.
On first sighting of this latest case, I tested the glue and it was more
than 30 Megs , just retested a bit and over 500 megs.
Will hot-melt
glue go the same way after a quarter of a centuary.
Is it bulking filler, corn starch or something organic rather
than mineral like talc in these glues that cause the problems ?
Melting temperature of hot melt glue.
Of course in critical use a matter of placing a sample on a heated
block with a thermometer mounted inside.
But as a rule-of-thumb for the more industrial 11/12mm size
Low melt ones , start melting at 90 degree c, translucent
or dyestuff coloured ones that easily double over in the length of
195mm requiring a force of about 1Kg (on set of scales).
High temperature ones tend to be naturally creamy colour ,
melting at 150 degree C, and much stiffer to bend, about
2.5Kg over 195mm length.
Testing LED/LDR coupler/isolator
With amp unpowered, for an LT3011, but more usual Spectrol ones probably similar
2 DVMs, 3V over probes in "diode test" over the LED pins and 150 ohm drops
to 142 ohm measured on another DVM, over the other pair of pins. Change from
diode to 200 ohm scale and the resistance changes to 146 ohm. Reverse the
active probes and stays at 150 ohm.
Exploring failed toroidal audio power transformer,
blowing mains fuses
48-0-48 approx 3A & 15-0-15, approx 0.5A
Size of final central space for shuttle winding/
unwinding 44mm. Used a 3/8 in end mill in a hand drill
to mill out the mineral filled epoxy as a chain drill
and then on hammer and chisel easily broke up.
The outer 15 V one is bifilar wound 0.45mm 2x 9,5m, ok for low voltage
48-0-48 is 1.06mm approx 260g approx 190 turns approx 32m per winding times 2
primary approx 2x 446 turns, 2x 60m of 0.7mm , 425gm
wound bifilar so relying on just 2 thicknesses of varnish
and wires welded about midlength of bifilar winding.
High temp plastic tape 0.04 mm thick
Size of ferrite core 1.65Kg 51.6mm x 98.8mm , 18.3mm thick
with tape covering.
Size with primary windings and 4 overlap layers of
tape over primary, size 56.6 x 103mm
tape layer between 15V and 48V wiring
Thermal fuse under the tape.
Exact replacement transformer DVM readings
240V primary 4.9R, both 48V 0.8R, both 15V 1.7R
Another one toroid for Ampeg B1, 2.6Kg to go in 1U case
150mm diameter , 35mm high
Complicated primary probably 100 + 17 , 100 +17 V
and 2 secondaries , one full wave rectified to give
perhaps 11.5-0-11.5V and main power rails
Inner winding , secondary of 2 layers, outer winding circumference 125mm
of 1.2mm wire , of about 140 turns per layer
and by winding some test wiring over it as a primary
would saturate at 2x 35V.
Over that 61.1 m of 100V primary and 13m of 17V primary then 65.5m of
second 100V primary and 14.4m of second 17V primary wound on winding
circumference of about 134mm.
2x 8m of 0.28mm wire for heater supply.
Final results for 240V were 2x 12V and 2x 28v , no load
6.8R, 2.1-0-2.1R, 0.3-0-0.3R
Yamaha EMX68S
3.8Kg, 57-0-57 10.5A 17.5-0-17.5 0.7A
2 x 29.1m of 1.38mm wire , 57V per section on outside
2 x 9.1m of 0.54mm , for 17.5V per section
2x 45m of 0.86mm wire for 240V primary
Untwisting bifilar after unwinding. Assuming wound onto a small
shuttle. Run out along a "rope walk" tying the open ends
separated so you can keep an arm in the division and counter twist the
shuttle as you go.
For unwinding weigh the wire on the shuttle before
rewinding on a drum. Make sure you don't
form tight loops in the wire in this process or later. Straightening using a piece of hollow
nylon , mounted in some rubber gas tube , to enlarge and take the
pressure off the fingers. Clamp one end of the wire and with gloves
on pull the wire tightly through an angle through the nylon.
With one gloved hand providing some back tension, forces varied to
suit the gauge of wire.
For winding/unwinding I made 2 shuttles from paxolin and hex pcb standoffs,
differrent lengths for different situations, covered
with cloth tape to soften edges, and countersunk screws to hold in place.
Paxolin is kind to the wire and to your fingers and strong enough.
1/ 2 paxolin slabs 25x3x200 mm
2/ 2 paxolin strips 8x1.5x 450mm , with glue reinforcement at the
joins, for small interior passage. Many many hours of tedium.
Used a B&D workmate with rubber sheet on the jaws for soft
clamping. A slat of wood between the jaws so you cannot
drop the weighty Tx onto the floor. Working at the end rather
than normal.
Pass the shuttle through in the direction that alows
undoing a turn, so goes with the flow. Hold the loose loop
up while transfering the shuttle to the other hand. For fine wire join
to some larger wire to bring up through the bulk and eventual
soldering to the external tails. Remember woven sleeving
needs stiffening up if it is to add structural resistance to stretching
as well as for insulation purposes. Make sure any joins , in the
bulk, are well twisted and soldered with no spikes and
then well insulated with ptfe/silicone sleeve and positioned
where it will not cause a layup problem , usually the outside
circumferencial face.
For even layup of wires stick a small
piece of high temperature double sided
tape on the periphery and mark off the spacing that corresponds (ie x2 or x3)
to touching of wires at the internal periphery (number
of turns per layer set by the internal opening periphery).
Use some siliconized label backing sheet as a marked off template
to mark the active adhesive surface with marker dots.
Also holds fine wire in place and interlayer insulation
in place and the marks are useful if on doing half a
layer you find the spacing is wrong , you can adjust the spacing
to get the right end point. A set of sensitive kitchen scales is useful
for weighing the remaining wire to adjust spacing, I always
underestimate the number of turns remaining. A tailor's taspe
measure is useful as well. It is preferable
, with their own insulated sleeving , to bunch leadoots for
each set of windings into one overall sleeved , to pass through the
later layers, or separated by an inch. Not closely separated
as it makes passing the plastic insulation strip through the gaps awkward.
Some mesh reinforced double-sided
tape can take elevated temperatures. To handle it use a sheet/s
of the siliconized backing sheet from sticky/address labels
to stick to the open face of the tape, before passing through
a pasta machine for fine even widths. Use high temperature sleeving to cover
the leadouts and allow for mechanical binding , for fine wires, by
using woven sleeving that can be bonded in better.
I find the following best for shuttle use. Pass shuttle through
after forming one turn, holding down the wire while passing the
shuttle through twice, to allow for a second turn to be formed.
Feel, with finger, the wire through on the unseen underside
of the taurus so tight and no overlap and mount a small piece
of plastic mirror tile in a suitable place to reflect the view.
Check for any problems , at the completion of a section
by powering on a variac , to the saturation magnetizing level,
and also high voltage insulation testing. If doing this a lot I
would make the mirror pivotted and sprung loaded to a fixed mouint
so it moves out of the way on shuttle passing. Where the windings have to
go around underlying layer leadouts, bring the windings in closer
on the outer circumference before and after passing the leadout.
Toroidal coil winding machine - partly.
I've now tried both (hand winding) methods , hours and hours later decided
some improvenment in technique is required if I ever try it again.
Shuttle and loading ring methods .
Loading ring is where you have a splittable ring. Shaped like a bike rim in
commercial winders, I used some polythene pipe , cut along one side, heated
over a piece of dowel to form into a U section. Reform , taped, through the
torus and wind the required wire while through the torus.
Works but without a third hand or at least some sort of support for the
loading ring it is awkward. Another problem with it, you need a way to hold
the hank of wire on the ring without any of it coming off when the required
wire is unloading. I just used a cylinder of rubber that just about held in
place in the U, but has to be released every 5 or 6 turns. Some sort of foot
controlled motorised system to wind back the excess wire onto the loading
ring, for each manual pass of the loop through the torus, would help. The
loading ring, permanently in the central torus hole, gets in the way of
manually posistioning and tightening the turns.
Shuttle works quite well, but slowly, as I found you could only sensibly
form at most 2 turns , then 2 combined shuttle passes. I've now tried some
different ways to do say 10 shuttle passes and then return to manually
tighten all 10 in one go. It requires a way of gripping the wire after each
pass and then the next loading turn, staggered along some gripper thing
that can be removed when returning to manually form the turns proper. With
enough slack to grip and tighten but not so much as to knot-up. I tried
Hellerman sleeving , pieces of rubber, slid along a tension spring. Held the
ire well but difficult to remove the rack. Some rubber, sliced with 0.6mm
grinding wheel , alright for larger diameter wire like 1mm or more and is
easy to remove before manual stage.
Turning the whole
thing through 90 degrees would make it much simpler. ex-VCR mode switch
motor gearing driving a spool and 2 free-running ex-vcr spools on the
inside of the polythene loading ring , fixed to the bench , rotating round
in the horizontal plane. And toroid mounted vertically. So motor drive could
be used to load up the wire initially and of course the main use, slack
take-up, on each turn. If the circumference for a wiring turn is 150mm and
the loading ring is 900 mm then to form one turn requires slipping 900 mm
off the loading ring and taking up 750mm of slack each time, and shifting
the anti-spillage bung every 6 turns.
Well the prototype bodged up loading ring system works upto 1.4mm diameter
wire.
http://home.graffiti.net/diverse:graffiti.net/toroid_winder.jpg
Using a mode switch motor+ gearbox and spools from VCRs , bits
of motorcycle inner tube, hosepipe, dictaphone footswitch and bits of Dexion
etc. Only a test run so ignore the uneven layup.
As long you start the retraction winding at the same time as slipping the
large loop of wire, then no problem with turns of wire unravelling from the
loading ring. I suppose its something you can get into the rythym of , once
practised, as it is a continuous process. As it happens the retraction
rotation speed was about right for a novice, 8 seconds for one rev of the 1
foot diameter loading ring. Loading up the ring intitially with wire, so
running the motor continuously , did not even make the motor warm. With
practise I suppose you could up the motor power and speed for faster actual
coil wind. I would not fancy making the hook transfer system and tensioned
forming system to fully automate.
If you load up the ring , with copper wire in the wrong sense ,
it does not matter , just flip the ring over.
Using milliohmeter to trace a ground-short on a digital , ground
planed and multilayer board. Using an AC LCR meter
Avo B183, replace the black lead with a lead
without a probe and solder the end to the ground plane.
This meter measures to 0.01 ohm , test components that
may go short circuit before trying to find any shorts
within the board , perhaps by slightly bending the board.
By finding the minimum reading to ground.
Another silly household tip
Supermarket plastic carrier "bags for life" are all very good but they don't
crumple up conveniently into a coat pocket. And standard thin ones
wear out after 3 or 4 uses. Place one standard one inside another,
draw the handles together and fasten with a 3 inch long piece of
nylon spiral cable wrap. If used for mainly tins then 3 bags perhaps.
When eventually worn to a hole in one , transfer the spirals
to a new pair of standard bags. Refinement - if you object to
being an unpaid billboard , turn the outer one inside out.
Damp-proofing the end of a single conductor cable
To avoid condensation creeping, via capillary action, between copper
conductor and the sleeving from an open end into otherwise sealed section.
How to make or what sort of term or construction to look for in cable specs?
Had a go chopping off a 2 inch length of hotmelt glue stick. Ground a crude
chisel on one end of a length of 1mm rod and drilled/melted through the
stick, pulled out while still warm. heated a copper wire cable core and
pushed through.
Covered with some heatshrink and a ceramic bead at the open end to keep
roughly centralised. Heating the copper and sleeving with hot air gun and
allowing the hotmelt to ooze out until contracted down, then cutting off
bead section.
Cut it open to check and seemed a good bond but will repeat and stand in
some water with potassium permanganate on the upper end of the hotmelt
section and leave for a few days.
Isolation transformers for testing power supplies etc
If you intend using a builders site transformer for this purpose,
beware the secondary may be centre tapped to ground.
Also the term isolation can mean noise isolation in other circumstances and so
earth/ground continuity to the secondary side.
Removing eboxy from potted boards
First try hot air gun on a low setting, large separation.
Fix or bury a 1N4148 to the epoxy and monitor "diode"
fashion on a DVM until voltage is equivalent to 70
degrees C. If not softened then try 100 degree C.
Beyond 100 is up to you, but a large thermal mass
involved with potting.
If that fails then
I tried a number of household chemicals in test cells with flakes of 2
week old cured epoxy, over a couple of hours, all at room temperature.
The only ones showing promise are
acetone as nail varnish remover and 2 types of paint stripper paste Polycell
with Dichloromethane and methanol and an "151" containing Methelene Chloride
and methanol, probably the same active chemicals. Acetone and paint
stripper mixture look most promising producing a gooey mess and reduced soft
test piece in less than an hour. All with no mechanical action at all.
For pcb use I would recommend repeated sequence, motorised wire brush then
this mixture on the track side.
On the component side Dremmel and very small ballmill or burr to grind
around obvious faces to iCs, large caps, and transistors so you could prize
off the section of epoxy over any markings, then mixture of Dremmel or brush
and epoxy softener. Sustained elevated temperature may help the chemical
process.
Replacing brittle grey ribbon with rainbow ribbon cable.
Cut half in or so along each trough with end snips and then by hand tearing.
Strip back to longer than required conductors.
Twist the strands and tin just the ends of each wire and snip
off the ends but still in the soldered length.
Brass rivets used on switches , pots etc.
What to replace with, after grinding off one end to gain access.
Assuming a loop of nichrome heater wire and washer is out then find
some 1 or 1.5mm, to just clearance suit hole size, high current transformer/choke wire.
Make a rivetting tool for a standard engraving tool. Find a
small bolt that will fit in the "chuck". With a small ball
burr in a dremmel, grind a hollow into the end of the bolt.
Hold the wire in some double action pliers , just enough exposed
to dome-up /compress a bit, too much and wire will bend. Repeat
a number of times. Gauge the right amount to cut off after inserting
into the hole, place a dolly at the other end, and form some sort of head at
the other end.
Cheap and nasty croc-clip ended test leads,crock going stiff.
Not the pivot jamming up but due to the plastic shroud loosing plasticity.
Unless anyone knows of a simple healthy chemical re-plasticising
process, cut into the plastic shroud from the wire end up to about the pivot.
Footswitch for indirect use.
ie under a foot pad rather than direct
Adapt a standard push/push otherwise miniature toggle switch
Clamp around the thread part with a potentiometer shaft
lock. Pad out to fit existing hole.
Make a cap from a 3mm long type clinch "nut" soldered
to a brass insert, after heating for ssolder
introduce some hot-melt to fit to pn of the switch.
Fix the switch mechanism with hot-melt inside the box.
Advantage of this is if someone really stands on the
switch it should slide in a bit through the shaft lock
Old dried out solder paste
Will not stick on surfaces. Find some straight soldering flux
and mix with the paste.
Peak limited "white" noise generator
Before ditching an audio effects unit Boss DD6 ( main 100 pin BOSS chip
failed) (see repair files)
I wanted to see if the DAC was still working . I cut data-in line, pin 4 of
DAC AK4352VT from the main Boss chip and connected , via 1K, to pin 9 , data
out of the ADC AK5353VT.
Feeding sine in the ADC, both channels, the datastream varies on frequency.
No data into the DAC, then nothing out. Connect datastream and there is 2V peak limited white
noise on the analogue outputs, left and right channels. Varying sine input
or even disconnecting makes no observable difference to the noise output.
Scope shows no obvious periodicity 100Hz to 100KHz, and sounds like white
noise and a peak to peak amplitude maximum of about
half power rail. Used 4.3V as supply , datasheets for the ADC and DAC from
the www.
pins 1 to 16 DC, M is some mid voltage
AK4352 M,H,M,M,M,H,L,L,L,H,H,M,M,M,M,L scoped frequencies , so about 400K on pin 3, 50K on p5, 11M on p11
AK5353 M,M,M,M,L,L,H,L,M,M,M,M,H,L,L,L
obvious interlinks DAC to ADC, maybe others, 2-13,3-12,5-10,11-7
Making a tiny endoscope from a webcam, not tried yet
Anyone successfully removed a glass fronted ball-grid-array CCD from a
webcam and extended out. The one I was looking at is a CCD 4mm square with
6x6 BGA array , 0.5mm spacing of contacts. I can get some loops of 0.12mm
nichrome wire under the chip to help lift off, when heated on the other side
of the multilayer board. Only a couple of SM Rs on the other side and easy
to focus heat on the approx 4mm or so square patch of the other side.
Assuming it could be removed without damage to chip or traces, what would be
the limitations of extension wiring? inches or feet? Would a blast of
freezer spray on the CCD damage it , before applying heat on reverse. ?
function of the relatively large square piece of coated
glass behind the tiny lens UV filter ?, I'm
assuming that to remove it, to make easier fitting into some sort of smaller
barrel, would be of no great significance.
as it has a reddish coating it is probably an IR filter so removing should
not only be ok but could have added (non endoscope) functionality , for
hotspot location. eg blink comparator imaging of same board and position ,
one powered other unpowered.
Latching footswitch replacement.
If the unit is only ever used off mains supply
then it could be replaced by a simple momentary
switch and a relay plus electronic latch.
Especially if DPDT let alone 4 pole footswitches
are not too easy to get hold of.
Captive nuts in combos
To avoid the possibility of them dropping in the works,
spire nuts or caged nuts, wrap some gaffer/duck/duct
tape around each one and pierce through both sides
for the bolt/screw.
Replacing pot with integral frame for pcb mounting.
Not chassis mounting. Salvage the frame from the
broken one by removing the tanged back cover and grinding
off the splays on the end of the shaft to remove the
wiper disc etc. To avoid the following , the next time,
I will drill 4 or 3 holes in the pcb to take
up the displaced footprint, 3 holes for the active
pot or 4 holes for the displaced, dependent on
surrounding components/pot shaft length etc.
Place the mount over the new pot using the bush
nut to keep in place. But this introduces a relative
displacement of the pins. So , remove the frame and grind back part of the
bush plate so the back cover tangs just about
make it inwards over the bush-plate plus the
anchor frame, and solder or glue the back tangs in
place. Then joggle pot pins and frame pins
to align with the original holes.
Multichannel mixer bush nuts
Use a cordless screwdriver adapted to take
a socket set socket braized to a shaft to undo.
On replacing just lay the nuts haphazard over
the bushes . Use a screwdriver handle , adaptor
and socket to do up or nut spinner if one large enough.
Vast majority of the time
the nut will locate itself and tighten. But occassionly
will cross thread , so don't use a powered drive
to do up unless you start each nut by hand. Perhaps
a spring or something inserted inside the socket would
guarantee initial threading.
Reusing or swapping dichroic reflectors etc
Use a tapered centride burr in a dremel to
grind out the fire cement.
To remove any UV circular cover disc grind
into the periphery with a centrided ball burr
to get purchase under and crack or lever off complete ,
protecting the reflector edge, maybe.
Probably held on with RTV silicone so
not that firm a bond.
Reset the plain bulb with fire cement in
the reflector.
For linear lamps, usually mounted in sprung
pairs of pin connectors. Grind off part
of the white pocelain ends until the rest
can be broken away to allow mechanical wire fixing
to the crimp terminal.
Removing rubbery potting compound
1/8 inch ball mill in a Dremmel for that sort of
stuff, localised heat and tugging action hepled in removing from surfaces,
. But still time consuming and you have to know there is no little vital
stuff buiried within it. For protective potting around large lumps -
make a pilot hole with a thin steel rod, chisel ground at the end,
used as a drill bit.
Then pass through it one of those carbide covered wire saw , garotte type
things kept under tension.
Or after a pilot hole then a current carrying length of straightened heater
element wire perhaps would work to cut through.
Formula for rubber drive band path length
ie the stretched length where it is too restricted to pass a thread around
to measure. Or to measure without taking apart a worker, to obtain a spare,
before having to take apart to replace. Using callipers, pair of rods and
ruler, radius gauge templates or whatever to take measurements of diameters and so radii and gap,
, whatever intrusions will allow.
2 pulleys , small radius "a" and big radius "b" and gap between the 2
pulleys "c".
So distance between pulley centres is a +b +c
Straight belt section between the two pulley rims (doubley tangential) is x
N is the angle, in degrees, between the line-of-centres and the b radius to
the tangent point of the large pulley. Also the angle between this extended
line and the radius a of the small pulley.
Parallelogram with parallel sides a, b and the other two sides x and (a+b+c)
and two right angles.
So reducing the parallelogram by lenghth "a" down to a right angle triangle
of lengths x, b, (a+b+c) and N between b and (a+b+c)
x = sqrt [(a+b+c)^2 - (b-a)^2]
and N = cos(-1) [ (b-a)/(a+b+c)]
Then band length = 2x + 2*Pi*a (2N/360) + 2*Pi*b[1 - (2N/360)]
cos(-1) is arcosine
Confirmed practically in one case anyway
Repairing large desktop mixer amps
To safely work on and move around.
Brace with Dexion marked D or Handy Angle
(slotted angle) marked H , H, supporting the angled top
panel. Anchor the ends with cable ties to
make a fairly rigid part cube structure.
May have to make up extendors or add to leads a bit.
After taking this pic, decided to cover the
the angle with gaffer tape, piercing the holes ,
to give a bit of insulation , should anything slip
as the aluminium of the case can bend a bit when
opened up like this. Remember, only put a bare
minimum of bush niuts and screws, no knobs,
back in place before doing "final" checks.
Send / Receive problems on band amps
While operating , push a toothpick inside each.
If very little force just from a matchstick size
is enough to break contact then they it needs attention.
Cassette deck, capstan spindle, speed calibration
Assuming my test tape is stretched (I doubt it , as it would have to be even
stretch the whole length).
I normally test speed monaurally so the beats are unmistakable. That is 1KHz
test tape output and floating accurate 1KHz sine signal both fed into one
headphone cup. Easily gives 0.1 percent resolution (assuming test tape is
ok).
If I record accurate 1KHz tone onto a blank tape for exactly ( to human
reaction time accuracy) 100 seconds. Then replay for beats of 1Hz or so and
output lasts for 100 seconds +/- reaction time then is the speed correct? I
suspect that repeating this process with the speed regulator changed , say 5
percent, would also give "correct" speed, that time also
Things counldn't be easier with a "known good machine " but how to check a
known good one?
eg quartz controlled (so low tens ppm) f-meter coupled strobe fed to
rotating capstan if diameter/s of the spindles are manufactured to specific
dimensions of precisely known tolerances.
I found a large quantity of salvaged capstans+spindles.
26 off on my mike (rather than callipers before) measure 1.99 +/-0.01 of
reading accuracy and 2 nearer 2.00
5 off measure 1.79mm
1 at 2.19mm
9 measure 2.49mm
So I will give a few to my engineer mate and ask him what they measure and
the temperature at the time.
The machine in question uses 2.49.
Assuming it is actually 2.5mm then doing the maths and strobing with quartz
f-meter calibrated strobe then my test tapes are 0.4 percent out from
calculation via 15/8 ips etc.
spindle rotation speed of 6.05 rev per second with an error of about 0.4
percent , needs longer gate time or repeating more times to bring that
accuracy up (rounding errors? so probably longer gate time required than my
meter has).
If 2.49 diameter then 0.6 percent out.
Hopefully will revise these figures when I get a bit of cross-calibrated engineering
accuracy but above is probably fairly representative and scale
for the different spindle diameters.
I used a white LED on a stable audio genarator coupled to f counter.
I found monitoring strobe at 6 Hz difficult but just painting
"tip ex" over about 1/4 circumf of capstan and monitoring for strobing
at 4 times frequency was much easier and better for the f meter gate timne also.
Another possibility as a test tape - record some constant tone, any f, on a
few minutes of tape, pull out a long length and pass a magnet over two parts
a measured distance apart. retract, and then time the interval between dips
in play mode.
Re-commissioning an old laser printer
Not been used for more than 10, years and been stored in a UK garage.
Although SMPS , powered up slowly via variac and the printer works fine. But
damp must have got to the toner. It clings to the opc giving a dusty
background rather than white. Removable opc and toner cartridge.
Removed the bung from the toner hopper area and placing in a sealed bag
with activated silica gel for a few weeks , then clean opc
and adjascent roller .
Scored lines on speaker voice coil former
Due to dirt in use or at manufacture?
Carlsboro powertone 10 inch ,open circuit, so burrowed in to explore. On the phenolic
cylinder that is the former for the voice coil . 4 pairs of scored
ines( stand out white against the amber colour of the phenolic). Each pair
about 6 to 7mm apart and arranged pairwise diametrically.
I suspect 4 pieces of sharp edged metal shim used in the original cone
setting and scored on removal. All the lines go axially , fully from one end
to the other , so if made in use would require the diaphragm (spider?) being
driven 10 mm into the magnet housing. Apart from the nearly symetrical
distribution of the lines.
the lines are on the smooth internal surface that glides over the central
section of the magnet, not the coil surface.
I doubt the lines are anything to do with the failure, that is well outside
the coil area. Probably due to stress on epoxy curing on a supposed stress
relieving bend of the coil runout wire inside a fillet ring of epoxy where
the cone meets the phenolic cylinder and before the join with the pigtail,
so repairable.
2 layers of 45 turns, 12mm wide, 46.1mm ext diameter of former and coil.
Overall thickness 0.54mm and phenolic former 0.1mm
There is no overheating , scoring or any damage to the coil or elesewhere
just these lines.
http://home.graffiti.net/diverse:graffiti.net/voice_coil.jpg
The woven and corrugated yellow part I've
labelled purple "S" is called the spider. The cone is to the other side of
it.
"L" is the lead out of the coil and "G" is some sort of optical guide laid
in the layup, on the manderel, of the phenolic for placing the lead-outs to
then meet up with the pigtails at the cone. Moire fringing is where the coil
is. A second sight guide and second of a pair score line is part in shadow
off to the upper right.
There is a third and fourth pair of lines diametrically opposite these 2
pairs.
If it was scoring from expandable/contractable mandrel or a chuck of some
sort I'd expect it to be engineered and pairs of lines exactly 6mm or
something precise apart, these are about 6 to 8 mm apart and some lines have
a nearby accompanying line. They are white as in scoring not grooves set in
at moulding . Or more likely no marks as what is the point if the process
leaves such serious imperfections where you don't want them.
Flxible wire
Twist up, in a drill, 6 strands or so of lengths of 30-35AWG/35-40SWG
or so gauge , about 15 times the final length, and cut into 3. Plait up
those 3 lengths and then cut in 3 and
plait together the 3 lengths of that intermediate. Keeping the "flat" side
in the same sense which will be the sense through which the
the main flexing action takes place. Similar process can be
used to make flexing ribbon connector if the wire is
laquered. Alternatively there is a plaiting technique
of 7 strands, but no knowledge , other than it is used
in straw boater manufacture.
Light duty impulse tester for clicking the sides of
SM resistors in a consistent manner if indications
are a component failure of one then check the
other s of same batch. measure R before and after testing each
I have a small adjustable automatic centre punch (spring loaded ratchet
release for easy centre popping of sheet metal). The minimum imulse setting
of that is 9Kg but found a couple of much weaker springs to try inside .
one gives range 1.5 to 2 Kg before the impulse, the other 2.5 to 3Kg range
of adjustment.
1.5Kg sounds worse than it is. Trying on the back of my hand is not painful
or self-mutillation.
Fusible resistors, case study of one that
had failed due to kit being dropped
Found this info for one supplier
http://www.peterparts.com/%5CCatalogPages%5C73%5C3123.pdf
and 5 band colour coding
Markings are to be shown on the resistor body by color coding.
1st, 2nd and 3rd color codes : nominal resistance value.
4th color code: resistance tolerance-J (±5%) - gold color.
5 th color code : fusing characteristics
green color -8 times
blue color -12 times
white color -16 times
violet color -32 times
if 1W then white band means subjecting to 16W will fuse in under 1 minute
Surely any normal 1W resistor subjected to that sort of overload will have
burnt up quicker than in 1 minute or have I misread something ?
Removed the grey coating and remnant parts of the metalisation/ MO sum to
about 1.5 ohm. The joining section is about 1 x 1 mm with a neat hair-line
crack across it and no sign of any heating there.
So mechanical breakage, probably a torsional failure due to
one end of the resistor swinging one way and the other end
swinging the other way.
http://www.diversed.fsnet.co.uk/crack2.jpg
Before removing coating, was definitely 2.2 ohm markings with white band
(if) for x16 presumably referring to that 1 x 1 mm section, so x32 violet
banded one would perhaps have a something like 0.7 x 0.7mm intercept region and a
green band x8 would have more like 1.4 x 1.4 mm fusing area.
http://www.diversed.fsnet.co.uk/crack.jpg
Is a microscope view of the crack. Viewing magnified it can now be seen
the fault developement.
The just left of centre part of the crack must be part of a crack that
formed initially but enough current flowed in the other parts making
touching contact until there was too much localised fusing.
1 to 1 marks the 1mm or so of conductor between the 2 straight line
etching/millings ? into the conductor to form part spiral paths.
The C marks the end of one of these formed breaks and there is another one
on the other track , off the top left of the image, giving about 1 x 1mm
central fusible conductor.
Nowhere on the remaining conductor is there any discolouration due to
overheating.
Anothe rcase study
Was in a valve /tube amp HT line as a dropper carying about 250V , no
obvious heating of the body or surround of this 10K resistor and no obvious
burning of the resistor track, microscope view
http://home.graffiti.net/diverse:graffiti.net/MO_resistor.jpg
B is the original blue body colour, C is part of the ceramic coating that
got left after my scraping. My scraping was axial so not the cause of the
tapered loss of metal oxide in the spiral between * and *, actual break just
off pic and a slight trace of this loss on the spiral above *-*. What caused
this tapering loss ? electrochemical? manufacturing fault ? All other 1/3 W
resistors in the amp look the same manufacture so wonder if being in the HT
line is significant or could any/all of the other 1/3W resistors fail in
similar fashion
I should have said there is a waveyness to the terminator between surviving
MO track and the erroded section of the taper on the original full
resolution pic for this
http://home.graffiti.net/diverse:graffiti.net/MO_resistor.jpg
shown here
http://home.graffiti.net/diverse:graffiti.net/MO_resistor2.jpg
So conclusion maybe runout manufacturing error
Small animals (cats in this case but others maybe) chewing on mains cables
Assuming vet/psychologist cannot change their behaviour.
Wrap nylon spiral wrap around the cable to at least make the live
conductors further away from wet teeth.
Gummed up potentiometers, as distinct from worn.
Indications are all pots become "noisy" about
the same time, eg each channel gain pot of a mixer-amp plus
pa pots. And perhaps turning the spindles become stiff.
Bathe the interior of the pot in methylated spirits,
then with precautions against fire , warm up with hot-air
and rotate the wiper a number of times.
Preferably make up an aspirator from a small thin walled
plastic bottle, bung, various ,size stepped, thin tube leading down to a
fine piece of sleeving that will go in any hole
in the rear of the pot casing. And a hole in the bottle
, covered with a finger, to act as a flap valve.
If no convenient hole then make a bath from a 35 mm
film container , cut down a bit to take the spindle
and angle board with pot , dunking in the meths,
and rotating wiper.
There is a vague , last ditch chance, to use this
for mechanically worn tracks where it is not
possible to replace or renovate. Add graphite
to the meths added via aspirator and not rotate the wiper , allowing
meths to evaporate and maybe leave a graphite deposit over all
interior , so low voltage use only. Not personally tried.
No glued heatshrink, but you do have plain.
Wind and "solder" the ends of some hot melt glue string
around the object before heatshrinking down the sleeving.
Making a Y intersection/split in screened leads etc. Find a line cord
connector cable entry gland. If not large enough
for the doubled-up cables , warm up with
hot air and expand with 3 prong sleeving pliers.
If a point breaks through, no matter, you will be covering
the gland with glued heatshrink after joining
the cables , robustly, and covering with the gland.
Synthesizing non-linear potentiometer characteristics
Not perfect but maybe a work around in awkward situations
My can RFI/EMI screening of is RS stock number "551-570"
other than it contains MEK and xylene, nickel and air drying acrylic resin
I had a go at adapting a linear 250K ohm
slider pot to 10K log.
With 2 narrow pieces of "masking tape" I laid one, leaving just the outside
edge of the 250K track exposed and another one on the paxolin, leaving a 2mm
strip un-masked .
Squirted out some "RFI/EMI" spray into a pot and then with a toothpick laid
down a thin line of conductive coating along about 40 percent of
the track (for comparison purposes) . Silver paint would have done instead,
but just a prooving exercise.
When dried that measured 23 ohm , over its length, and about the same for
that portion of the pot track via the 2 relevant terminals.
With a razor made 3 cuts, so divided the aquadag track into quarters. Then
some nail varnish just over the aquadag and onto the remaining paxolin just
to confirm no adverse reaction and a suitable fixative and rassembled.
I suspect the characteristic would be a bit wavy rather than linear, if
properly plotted out, but results
5 percent 10 ohm
10 % 400 ohm
20% 1.8K
30% 8K
40% 42K
50% 66K
and then much as the original 120K to 250K of course
I didn't bother with the following, but
For the upper section, for a log characteristic, then lay down a full length
aquadag track and make more , closer spaced cuts, and also a 1K and 20K
resistors or whatever , to fit, external to the pot.
Other non-linear characteristics , other than log could be so fashioned. And
as the thickness of the aquadag track is not critical then rotational pots
could be so converted almost as easily. The only criteria are higher ohmic
value to start with than final requirement, low wattage and enough space to
lay down these extra tracks.
Closer checking with vernier calipers and DVM there was an abrupt jump at
23%,
from 1.5K to 3.4K, so not a perfect solution
Renovating yellowed with age ABS plastic cases.
Originally white or cream and now yellowish or brown.
Not tried it myself but search on keyword retr0bright
and the use of Hydrogen Peroxide.
Vero stock codes of 1980s
2 or 3 digits and then a space or a dash
eg for otherwise anonymous packs of connector pins starting numbers
227-,252-,17-,266-,14-,908-,41-,15-,928-,903-,904-,906-,83-,901-,920-,266-,
273-,275-,229-,27-,270-,272-,900-,251-,63-,269-
Maybe when they became BICC Vero the space/dash became a dot in the codes
Bakelite knob seized on corroded pot shaft, nut seized grub screw.
Wanting to save the knob and consequently all the knobs, for matching purposes.
I found a piece of metal round bar about the same diameter as the knob,
drilled through axially on a
lathe, 1/8 inch hole. Mounted co-axially over the knob. held there with
Jubilee / hose clip and some padding.
Then 1/8 inch drilled into the knob and into the shaft. Saved the bakelite
dust to mix with epoxy to make good afterwards. Luckily the 1/8 inch
drilling was enough to loosen the knob and filled the shaft with epoxy as
well, in the end. Otherwise it would have been a matter of drilling larger
holes until freed
Motorola supplementary TO3 punch stamp codes
May or may not be of use where
printed marking is faded/damaged/burnt/ground off for cross-correlation
A batch of 1988 MJ150024 with punch stamped code "D14" around a mounting hole
batch of 1988 MJ15025 stamp code "D13"
1979 MJ15003 double stamped "CD"
I don't know if UK amp maker "Exposure" uses rebadged Motorola but marked
EXP8P (pnp) and stamped "C5" in similar position
EXP8N (npn) stamped "C4" in an Exposure VIII of 1988
A batch of 1972 (7293) MJ2955 stamped "27T"
An MJ8505 of 1984 (8414) stamped "1H6" or "IH6"
"BP" on 2N3055
"DID" on MJ15024
"DIB" on MJ15015
"34P" on MJ15003
"CFF" and "CFFY" on MJ423
"CAH " on MJ1800
Cutting back a recess in a plastic know to cover pot bush
These knobs not solid having an open cylinder dividing
inner and outer sections. Found a piece of metal cylinder to seat into
this ring.
Mineral insulation pyro cable termination seal marked
"CC 4H 2.5" was ideal, 14.7mm od and 13mm id to push
into the cylinder and 10mm id at the other end taking as a guide
a 10mm drill bit to recess back neatly.
Cutting down round Al pot shaft to D section.
Hacksaw down to length, check position of flat especially
if the pot or board is to be mounted upside down.
Grind down a tapered flat with a bench grinder.
Making a dual concentric power connector plug.
If not of the set of "standard" ones,
pin .75mm/outer 2.35, 1.35/3.5mm.1.5/5.5, 1.7/4, 2.1/5,2.5/3.5
This one was 1.5mm pin/4.5mm outer.
Stripped down core of a gauge 20 QM pin socket
was ideal for the central connector. Outer made
by rolling some brass shim with enough overlap
to solder. Cut all 4 corners back to protect exposed
edges. Roll around a 4mm drill bit , should open itself
out enough to fill the existing socket. Check all
contacts are ok and the diverter switch functions in the socket.
Solder the exposed shim end then the rest and check still enters the socket.
Ball mill the solder / exposed shim edge to fairly smooth.
Make sure no slits left for epoxy glue
to migtrate onto the socket pin by covering enough
of the socket pin with a small length of rubber sleeving.
Solder the wirees to the pins and place concentrically in the socket.
4 small pieces of over long card to part cover
with epoxy and introduce into the space.
Mount th ewire ends in moulding as below.
Salvaging small surface mound devices like LEDs and SO23 size
devices - gradually undermine the pads by cutting with a scalpel,
the aim is not to stress the component into fracture.
General tip for replacing bearing in a small dc motors
This was the small motor in a CD unit (platter motor). To remove
the CD platter itself first measure gap under platter with feeler guages,find
scrap of metal with 10mm hole
to lay over the platter coaxially to protect rest of unit and heat
with a heat-gun on low temp setting and pull off the platter.Take apart
the rear plate of the motor(prizing off) then by pushing spindle push
out the rotor and brush plate.Again with the 10mm plate over the sintered metal bearing
heat up with hot air gun and with a punch knock inwards the bearing insert.
From a scrap motor i could find a bearing of right bore but smaller
outer diameter but packed with small shim pieces and glued into
the original casing
Strobing of compact flourescent lamps - UK bayonet cap.
Continuous strobing from switch on.
One reason maybe poor surface contact in the
socket. Like the positive pips on some batteries
being not the right length to make good contact through casing.
The dimensions of CFLs seem to be slightly different to
conventional ones.
Then much higher inrush current for CFLs.
Add a small lump of solder to each blob on
the bulb, don't fully melt the existing blob
so as not to disrupt internal join.
Do the same for those annoying batteries that
just fail to make contact.
Damp sealed extruded aluminium box section enclosures with plate ends.
Come to replacing , if foam inside as a sealant then cannot remake the
seal. Cut a pair of over and undersized squares of rubber from
inner tube. Push half a Qtip through each corner of the end plate
to hold the rubber in place, as it curls of course.
Locate the stems in the extrusion tapped holes and replace
one by one with the screws.
Moulding a low power/ low voltage line connector to a cord.
Find a tapered nozzle from a mastic tube.
Cut diametrically , 2 cuts to give 2 conical
moulds, coat in silicone oil. Fix grommet/grommet in a grommet
over the functioning end of the connector so
it will rest inside the wide end of the cone.
Drape, overfill hotmelt
over the end of the connector/wire and place in
half the cone, overfill over the top , place
other part of cone over and push together by hand.
If black required then warm up the surface only
with hot air gun and then dangle
in copier toner or powder dyestuff or pre-prepare
coloured hotmelt glue.
A basic test for rogue "tin pest" RoHS soldered resistors
And with all lead-free boards I now routinely put a 2Kg pull at each end,
via thin-nose pliers, of all 1W or more watt resistors, any resistors mounted
off board or any that look a bit discoloured.
Practise pulling a spring balance with the pliers to
get a feel for 2Kg
Before cleaning a front panel
Especially stainless steel metal with screen printed
lettering. But best to check in all such situations
, whatever solvent or polish you use.
Try some of the chemical on a point of a toothpick
on a inconsequential piece of the script first.
Connecting banana plugs to "cooker cable" size speaker wire.
Cut about 1.5 inches of insulation off.
Tin over the exposed copper.
Cut the 1.5 inches of sleeving into 2 and in one piece
cut a spiral cut along about 1/2 inch leaving 1/4
inch length uncut. Push this over the copper with
the spiral overriding the full insulation and perhaps
pull the wire core out a bit more and clamp off
on the sleeving. Enough to be able to
plug to the wire. Screw cover on and push
the small bit of insulation back and re-wrap, cut
a bit off the end if necessary.
Then some clear or non-black heatshrink heated over
the spiral section , part banana cover and a bit
of the full sleeving. Perhaps some hotmelt string
around the central spiral area before heat-gunning
the heatshrink.
Harwin type connectors sealed against vibration with too
much nail varnish and down between the 2 parts and
not external bridging.
Push a compass point down between the affected faces and then
with something like the flat end of a 6 inch steel ruler
, between the ridge of the removable piece and the top of
the fixed , twist one way and the other, to lever out.
Valve bases with small holes in the ceramic only just larger than the valve pins
as used in Laney valve amps, good for allowing the use
of only valves with well spaced/aligned pins but no help if the
internal socket sections need forcing in to make better contact.
Remove the skirt , if necessary, to allow a bit of rotation
and angling for needle point probe entry.
With a course angled conical "diamond" grinder in a Dremmel
grind out each hole a bit and then you can push in a needle
point to bend each socket tang inwards a bit.
No longer gauged entry for valve pins
Re-using enamelled wire, magnet wire, from transformer etc.
So angular bends in the copper. A job for the garden
as a good long straight run.
Drill a hole, a bit bigger than the wire diameter
in a piece of nylon or soft plstic rod and
chamfer the internal and external edges.
Anchor one wire end with a clamp to something solid.
With a pair of stout gloves pull along the
wire , removing some of the irregularity.
Then insert the rod and with some gloved grasp,
giving some back tension, holding the rod at an angle
reform the bends. Use the minimum angle to remove
the bends, different for different wire thicknesses.
Failed heavy duty 1/4 inch jack plugs.
If problem with the connector part , retain
the heavy duty barrel and leadout and more than
likely the screw thread of an ordinary 1/4 inch
jack will screw in as replacement.
Engineer's Blue
Vary rarely need any but an adequate substitute for such
rare use is a permanent ink felt tip pen over the relevant area.
Broken sub-1.5mm drill bits , don't throw out.
Grind the end flat and then grind 2 angled faces.
Presumably roughly , but can be used on plastic.
pcb material and copper board traces, if not metal proper.
Heated screwdriver for Mackie powered speaker enclosures
These Mackie screws, 17 of them, (2 slightly shorter either side of the bass
driver) , in my opinion , are bordering on the threshold of shearing without
such heating, they deliberately used undersized holes probably.
Heated screwdriver worked better than I thought
http://home.graffiti.net/diverse:graffiti.net/heated_screwdriver.jpg
Pic shows the original soldering iron heater,
for unknown model. The one used with flange ground
off , added silicone sleeved wire and PTFE covering
the heater, slid over the end of the driver shaft.
And one of the long screws, mm /10mm/ 20mm graph paper.
Used about 10 watts, could have gone higher as no
plastic melted onto the PTFE.
While heating one 10 inch recessed screwhead, extracted
the previous heated one. Further general tip - slide some grommets
onto the second extracting driver to locate easily
on the head, for deeply recessed screws.
Plastic ring binders/ comb binders
I picked up a 300 page manual on the wrong side and by front card
only. The self weight of the
paper, in freefall, meant the spine unzipped leading to a cascade of loose
sheets.
a/ anyone know how to open up the spine, introduce the sheets and let it
roll up again , without access to a binder machine. ? Doing each tyne, one
by one, the self curve makes it want to break out of the punch holes and
straightening destroys the binding action.
b/ this spine was in proper working order but any old broken ones I come
across I will do the following as well as all thick/heavy manuals. Darning
needle and some cable-loom lacing cord/tape threaded through the pages,
looped a few times and knotted, when aligned with a couple of small
screwdrivers, repeated a few places down the spine. Then replace the spine
or a new one. Then it does not matter if the spine breaks or tries to unzip
itself. Someone elses's advice
Build a sandwich with your 300 pages
in between. Wrap with big rubber bands, straps, bungi cord, belt, or
whatever it takes to keep everything in place. Align the holes. The
idea is to keep the pages and holes all aligned, thus freeing both
hands to uncurl the Velobind backing.
Removing pcb clamps in awkward/restricted place.
Those white plastic pcb standoff and sprung clamps
that often hide behind components.
The cover off a banana/wander plug, cable end is about
right size to place over and release the clamps, where
there is not space to use a 2.5mm nut-driver.
Making a "set of rolls" for flattening down
round copper wire for edge woiund voice coils
http://home.graffiti.net/diverse:graffiti.net/rolls1.jpg
Nippy engineering vice for the strong moving function,
assorted ball reaces, mainly linear bearing of 15,8 x 22.1 mm
with the rollers removed and 2 simple races set inside.
As not used in a linear sense but for its length so a handle H
can be fiitted to it. Large bearing, this one 47x 14.4mm
to give the space to use the existing vice jaws.
Aluminium stepper arm from an 8 inch Winchester conveniently
had a clamp arrangement (next to the L) so
just needed drilling out a bit to accept the
linear bearing. A couple of flat face pulleys
and a distant one with some light elastic tension as the wire
extends in pressing , so go up a gauge in size from calculated size.
2 small bearing loose in the large bearing and clamped
by the large washer, all very bodgey. Also a sheet of
steel behind the static jaw to pack out to make the
bearing surfaces more parallel. Vice easily converted
back to original if you keep the original bolts with it.
Knot the 2 ends of the wire and run back
and forth till knot hits the guide pulleys
and gradually reduce the size by winding in the
vice jaws and checking with micrometer.
eg .42mm wire flattened to .18 to .20 mm.
Useful for creating fine solder wire for small spacing ICs,
if you rarely have to , so no requirement of a regular stock of
fine wire. With a very fine conical soldering bit.
So a length of 0.8mm multicore solder
squashed down to 0.25mm and then roughly cut widthways into
about thirds with a razor, short lengths at a time.
Radios of 1950s or earlier , rear wander plugs
for aaerial/earth/gram inputs etc
Probably noominal 1/8 inch but usually between 3.1
and 3.4 mm diameter that have no modern equivalent.
Find some 3mm brass cotter pins/ split pins and
open out between the pins, and solder wires to the brass.
WKnobs for slider pots
Plastic beads from craft shop. Heat up and
squash to form the hole to oval.
Woven mesh sleeving, not flameproof of course
for simulating cloth covered wire.
Mainly for old radios
Nylon cordage from a hardware shop and pull out
the strand/strands of core string to leave just the sleeving.
For a source of larger diameter of such tube, "toy hammock"
for "children over 3 years" for hanging their teddy bears in,
has larger such cordage. Also hobby shops have cord
and ribbon that the cores can be pulled out from
and camping shops for tent guy rope.
Paint with , preferably kopal , varnish after threading
in place over leads or looms. For simulated cloth covering
start with modern pvc covered wire then some of
this nylon mesh sleeving. Beforehand dye the mesh
with coloured broad tip permanent filt-tip pens
, different colour either side when layed flat, for additional colour coding.
Then pull the pvc sleeved wire through.
Haberdashers have blue, green, yellow cordage and black
or white tubular elastic all of which can have the cores
pulled out. Elastic cores provide 1mm or so diameter rubber.
Maybe then dull
down/lock in place with brown varnish. Not the subtley
of strand-coded cloth covered wire but part there.
For 2 colour , at start melt an end , then pull one
core string to ruck up the sleeving and then coat with
felt tip pen. Then release and coat overall with a
lighter and contrasting colour.
http://home.graffiti.net/diverse:graffiti.net/old_sleeving.jpg
Top 2 , plain white cordage with the core strings removed. The second one
rucked up with one string remaining in the core before felt tip penning.
Dry wipe , non-permanent felt pens work for this purpose and
tend to be less dense colours.
Remainder is some cordage with a green stripe running through it, if penned
with a light colour , like the central orange, then looks a passable
imitation of old sleeving with contrasting stripe. Then slid over modern pvc
wiring. Another source of short lengths of expanding mesh covered elastic
is hairdressing "hair loops" for tying back hair but relatively
short lengths.
For anyone who has to re-use a 1/4 inch jack socket, just replacing
an active pin. From the solder side push a thin blade screwdriver
between pin and plastic to bend/force the nearest swage out , and
with luck the force required will dislodge the other one,
that is set going with the flow.
Phenolic ribbon cable repair
Shellac coloured coating around the conductors, which
a 400 degree soldering iron will not melt.
OK down to 0.1mm conductors with 0.3mm spacing.
Make doubley staggered pairs of scrapings with a
scalpel, down to the copper, using a scalpel blade.
One either side of the break to bridge and
the other pair to check for continuity/ lack of s/c
before powering up. Perhaps physically
reinforce over the break with hotmelt glue sheet on
top and bottom, before doing any soldering.
For that, coat the scrapped patches with solder paste
and apply iron. Use 0.2mm prepared enamel copper wire,
if only pairs of 0.3mm spacing conductors, eg FlatLink
"stripline". After checking conductivity etc put
dabs of epoxy over the solder/wire points.
Bending phenolic cable
One thing I did notice, luckily, is that the bend is always on the side
where the fine conductors are, not on ground side, so forcing the ground
side to stretch , not the conductors, compressing them if anything.
Used a 300 degree C soldering iron barrel, no damage/ melting to the
phenolic and took it in 3 stages, heat bend 1/3 way and repeated twice
until a tight radius bend giving about 1/16 inch gap. Just did freehand in
the end, no 1/16 inch wide edge. At 300 deg there was still a fair amount of
resistance to bending, hence 3 attacks.
Permanent bend and no breakage in the 0.1mm conductors
Useful info from visual inspection of fuses
As far as mains glass-barrel fuses , its annoying when the owner
throws away the blown fuse.
Inspection can tell a lot, from
2 remaining sagged pieces of wire with a small break in the middle, could
be due to fair wear and tear in effect, sagging over time and current carrying
capacity going down over time, combined with some vibration/jarring perhaps.
Sometimes no observable failure of the fuse-wire but just the
process of removing fronm the holder is enough to make the
weakened wire fail in another place and drop away into the barrel.
via still extant wires at the ends and one molten ball , no smoke
blacking (x2 or so overload)
via a few sputter balls in the tube , some blacking (x5 or so overload ?)
to totally black and perhaps even cracked glass (> x10 overload?)
Terminating fine wire of order 45 SWG/ 40 AWG.
Stripping the varnish/enamel off the wire.
A small square finest grade wet and dry emery cloth.
Experiment first with different grades of paper and card.
Glue a pair of small padding squares of paper to some card
and one side of the sanded side of emery to a packing piece.
Slide the wire end in the space and lightly push emery
and backing card together while pulling the wire through.
If more than a few ounces of back torque then adjust.
When correct spacing/action etc turn the wire
and repeat to abraid around the wire. Better than burning
the insulation off , which weakens the copper.
Where you need thicker gauge wire tails to the fine wire.
Strip back a length of the fine wire to allow 3 or 5
flattened loops to be made of stripped wire.
Strip the coarse wire and lay in the end of the loops
and twist all together, the length of the loops,
then solder where the 2 grades of wire overlap
but leave the remainder just twisted. Avoids
a stressed single joining point.
AWG wire gauge info
A=2(50-g)/3 and A=Pi * D*D/4
where: Pi= 3.14...
g = AWG (American Wire Guage ) number
A = Cross-sectional area in circular thou (a circular thou is the area of a circle 0.001 in. in diameter).
D = diameter in thou (1 mm = 40 thou for these purposes)
Neutrik RP8 connectors
Certainly for pins with female function, removal process.
It may be possible to remove them by pulling on the crimp/solder stem
as there are not ratchet type spring clips, probably expanding ring grippers.
Piece of 1.3mm tubing with 1mm bore fitted to some sort of
handle and pushed in the contact holes, pushes them out.
Adapting white LEDs as festoon bulbs
Where 5mm is the maximum dimension / available space of the original bulb so
you cannot turn aroud the LED. Would be useful for replacing awkward voltage
festoon bulbs.
The maximum beam of 5mm LEDs is axially but how to convert to transverse?
Grind an angle to the dome and melt on some mylar?
Wrap whole LED in mylar except for an axial slot?
Abraid the LED all round to make translucent rather than clear?
Then adapting AC or DC supply to suit
I have 27 salvaged 35mA ,5mm white o/p LEDs and no 3mm, and every now and
then need a festoon bulb that is usually not 12V.
The bluishness can be compensated for to a certain extent by coating with
varnish with a tiny spot of red dye mixed in.
So far the best conversion seems to be 1.25 inch fuse glass barrel with ends
removed, scored and cut down to size , heated and pushed over an LED and an
oval of muylar glued into the open end, forming a 45 degree or so mirror.
Placing back the fuse ends for soldering to, with wires taken external to
the barrel..
Then a wash of vaguely red tinted varnish over the glass.
With the larger ones there is no great problem finding some sort of filament
bulb to replace it with.
But small diameter , high wattage ones, like the those built into meters are
a different matter.
3mm LEDs, in there, would give a spot light which would not be distributed
by the white plastic housing, giving a very poor meter illumination
I've never tried melting an LED before , not that easy.
In the process I found that filling the glass barrel with translucent hot
melt glue gives quite a good linear diffused illumination.
Air entrained clear glue or plastic would probably be a bit better but
hot-melt is quite convenient.
Comparison of 2 in series
http://home.graffiti.net/diverse:graffiti.net/festoon_LED.jpg
http://www.diversed.fsnet.co.uk/festoon_LED.jpg
as original LED and then a half length, 1 1/4 inch fuse barrel filled with
hot-melt glue.
A bit more scientifically measuring the original transverse, at a given
distance over matt black carbon/foam , was about 30 lux, modified LED about
180 lux and for comparison a 12V 6W festoon lamp at the same distance about
1000 lux.
Comparison of illumination , by readable text, at a distance of a foot I
would say the modified LED was equivalent to the 12V bulb running at 0.1
amps rather than 0.2A.
So at least 2 such LEDs at 30mA to have any chance of equalling the
illumination of the festoon.
So a white LED at each end of a filled glass barrel , at least , needed to
equal.
The mylar mirror just gave 2 bright spots rather than an illuminator, but
"frosting" of the glass may still work
Lifting heavy amps etc
I'm aware of lifting with the legs, not the back and trolley for horizontal
movement, rather than carrying. Any other advice assuming no human assistant
available ?
Travelling ceiling mounted block and tackle, swivable mounting frame on the
bench ? All little problem if the same amp enclosure each time but what
about universality of clamp arrangement or occassionally working on a heavy
and long 7 octave Yamaha piano and other odd shaped objects.
I was dealing with a large mixer amp, requiring working with
amp propped up slightly one way and the mixer propped another way (if
regular appearance I would make up extender ribbons). Then a 5 foot long
Yamaha piano full of key weights etc, the two main sections are very
difficult to separate and reassemble safely both for keyboard and repairer.
I find most problem, not the lifting to
bench height but turning stuff around and orienting at awkward angles while
on the bench. Instead of item being close in to your body , you end up
lifting most but not all its weight perhaps 15 inches away from your body.
So far I've seen nothing to beat pulley block/ ceiling roller track / heavy
duty bungee/ softened-faces G-cramp as hook and a paired strop of some sort for
other situations, supported each side and under. With mixer
amps one of the 4 strops is useful for clamping the upturned
mixer section to.
How many peole work at a bench with all round 360 degree access ? eg like
photocopier repair where you have to have 360 degree access.
With an island work position , especially if on wheels , you then have the
trailling leads problem and also an inherently unstable bench. Perhaps
engineer a floor-fixed lock-down point, for the table/trolley , so it cannot
tip over so halves that problem (you can still pull stuff off the nebch),
plus associated island power sockets, but it still leaves trailling power
and probe leads to scopes etc.
Converting metal cased TO3 transistors for mounting as plastic cased TOP66
IRF.... for IRFP.... etc
Except for the most thermally demanding situations.
Find some brass or steel nuts,
2 hacksaw cuts to make a slot. Or some flat metal bar with
a couple of slots cut in.
Cut down some PTFE rod to discs , one around
each TO3 pin then one of the cut nuts around that.
Solder another one or more to those 2 using
a broken TO3 as template to take the heat ,
fill the holes with screwed up tin foil.
Transfer to the wanted one, solder some silicone
sleeved wire through the nut slots and one to
soldered to or solder tag bolted to the TO3 case.
wires usually go the short major axis of TO3 body for G,D,S pinning.
Use silipad
insulation rather than mica to deform slightly and
slightly locate the nut assembly or use mica/ grease
for probably better heat transfer.
Insulate a spring or bar retainer from the cap of the TO3.
If a pair with central bolt (with yoke) then perhaps cover the
bolt in silicone sleeving.
Modifying a knob for contrasting axial and radial pointer line type.
Assuming can find and colour a resonable size and colour
match for the body.
Cylindrical rasp in dremmel to cut 2 slots,
dot in some paint and laquer over when dry.
Doesn't protrude like the proper ones so
swap for a little used control .
Trying to find hidden fixing screws on equipment.
Perhaps hidden under stuck down trim.
Find a rod magnet from a burglar alarm door
or window reed switch and hang by a thread and
"divine" over the casing.
Is that lead-free solder ? some simple, but not definitive
tests for lead-free solder and other observations.
Lead-free production probably started in Japan in 2001 eg
Japan passed the Electric Appliance Recycling Law, April 2001
eg
http://www.japanfs.org/en/pages/011541.html
"The soldering on every aspect of Sony's DCR-TRV 30 digital camcorder,
released in March 2001, is 99.7% lead-free, including all supplied
accessories. " Europe joined the clan in 2006 (RoHS).
So probably sets the earliest date ie
any board dated 2000 or before, then no chance of lead free stuff.
Genuine , not just assembled, manufactured in USA
is probably lead free solder , up to 2009 anyway.
If it says "product of ....USA" , is post 2006 and has
"made in (Asian Basin) " then probably contains PbF.
If a green RoHS sticker on the outside or PbF or RoHS
marked on the pcb overlay , then lead-free solder.
A sticker in the manner of road traffic prohibition sign stuck to the pcb.
Circle with a diagonal bar and letters P and b either side
means lead free, "slashed out Pb".
Some hints for deciding, probabalistically, if no labels found.
If the solder joints have a conical form with sharp
points where component leads emerge then probably lead-free,
compared to a more domed form for leaded.
If more temperature is required to melt the solder
then probably lead-free.
If the quality of the solder joints make it look as though
it was done by a trainee on their first day , then probably pbf.
Is the surface appearance of the solder a dusty grey or white long
before you would otherwise expect it to look old,
making it look like its been stored in a damp shed or garage.
When you desolder some with a 700 degree F iron, is it "claggy"
ie pastey and lumpy in texture ? and before heating looking like
solidified volcanic larva (wavey differential cooling in layers perhaps) .
If when molten you push in a stainless steel sewing needle and extract it
before the solder hardens and leaves a trace on the needle,
leaded will only require a finger nail to remove it
but lead-free will more likely need pliers to remove it
from the needle.
WEEE/RoHS/PbF rogues gallery
http://home.graffiti.net/diverse:graffiti.net/pbf_crack.jpg
My micrograph setup only allows one magnification.
This one on an amp of 2006 with the tell-tale green RoHS sticker on the
back, only in for a dodgey speakon connector. Usual looking pbf solder that
looks as though its been in a damp shed for 10 years.
But magnified visual inspection of the pcb showed this 1/4 of a circle crack
in developement around the pin of an 80V, 6800uF. About 1mm between X at pin
centre and the crack, left hand end shows the complete solder. Any guesses
how long before total failure if unattended to ?
I suppose the purists here would say leave it alone, as it is what the
designer/manufacturer specified.
I've not seen a loose IC but I was surprised to find the simplest of
components loose once - a pbf "soldered" wire link, ordinary 1/4W resistor
gauge of wire. I'd like to know what that whitish-grey powdery deposit on
the solder is, that makes these boards look as though they've been in a damp
shed for years. Is it tin-pest, coming to the surface, the powdery allotrope
of tin? that even medieval church organ builders were aware of.
I see its the work of the devil, just googling, the first one
http://dwb.unl.edu/Teacher/NSF/C10/C10Links/chemistry.about.com/library/week
ly/aa040300a.htm
"The gradual disintegration of shiny tin metal into a crumbly, gray powder
was observed hundreds of years ago on tin organ pipes housed in unheated
northern European cathedrals. With allotropy unknown, the phenomenon was
attributed to the work of the Devil and was variously denoted by the terms
tin disease, tin pest, tin blight, or tin plague. The reaction is
autocatalytic; in other words, the presence of the gray alpha-tin
accelerates the process. Tin disease manifests itself in pure tin. Antimony,
bismuth, or lead retards this transformation. Tin alloys are thus more
resistant to tin disease."
I was reading up recently concerning damage due to the excessive heat for
PbF soldering of pcbs. The most common fallout in immediate post-production
test is multilayer ceramic caps.
Is this delayed damage on this German made Dynacord mixer amp, definitely
using pbf construction.?
A 100nF, 100V cap incinerated itself , in an area that did not affect
routine use of the amp.
The failure of the amp was due to another of this batch of capacitors with
less than 20V standing DC, going ohmic, measuring of order 20K to 30K with a
DVM when I measured it but varying considerably.
in the parts list as
KO-KER 0.10MF 100V 20% cap ceramic 100nF
it is mustard yellow epoxy cased, axial leaded, 2.2mm diameter, 3.5mm long
at longest axial dimension, with markings (requiring x30 microscope to read,
and not that clear, so maybe wrong transcription)
R1E
104
MGR
619
104 for 100nF and 619 for week 19 of 2006, R1E for 100V ?
I've not found the manufacturer yet under KO-KER or MGR yet, but I assume
from the size that they are multilayer ceramic.
I can of course replace these 2 but what about the remaining 40 or so?
the sort of caps involved, not this company necessarily
http://img.en.china.cn/1/0,0,434,3172,234,179,b831706c.jpg
I forgot about the mixer part, 103 of these same caps in the whole amp. So 2
failed in 3 years, where's a statistician when you need one ?
I've not fully found out why MLCCs fail due to the extra heat of PbF. Its to
do with thermal mass of the ceramic , but failure would seem to be cracking
rather than going ohmic. Or does the cracking allow adjascent metalisations
to bridge or approach very close. These were wire ended not SM, and failed
by going progressively ohmic, the one that burnt out did not blow 1.6 amp
fuses with 40 volt over it, the other one was down to 20 to 30K ohmic
causing problem.
This seems to explain the mechanism
http://www.era.co.uk/news/rfa_feature_07b.asp
but that is for SM and I've not managed to confirm the ones I'm dealing with
are definitely MLCC, how/why round format ?
So metalalisation creep down the cracks is the mechanism, so would take
years to develop. I checked this retained cap a few weeks later
and it showed >30M so mechanism probably requires application of
DC every now and then.
The sign of the devil is not 666, it is the green label on this image
http://www.labelident.com/images/product_images/thumbnail_images/341_0_pb17_
pb18.jpg
I could not find the single green only version of the "RoHS Conform"
sticker.
When resoldering failed RoHS joints , if using silver solder which is
by definition also leadfree (but more expensive) introduce the solder at
the pin not the pad, as its there the problems usually occur. Hopefully any
tinpest powder will be dispersed but to be more certain its a matter of
desoldering and mechanically scraping back pin and even maybe pad , before
resoldering.
Aerosol can nozzles
Before throwing away any aerosol cans , remove and stow
the fine delivery tube, cap/s and the extender tube that goes
down into the can to depress the valve.
At some point you will be faced with a broken nozzle
on a can of something.
For delivery of aerosol fluid into small openings,
eg pots, slide switches etc , find some 1mm diameter
or so tubing eg "standard wall spaghetti tubing",
open out the standard aerosol tube with a dart point
and feed in an inch of a foot long piece of fine tubing.
Adapting a linear 250K ohm, suggestion as not required in the end
and only experimented with the idea.
Converting to 10K log. Because of the requirement constraints a
more suitable 20 or 50K one not available.
I doubt if the following could be used for the narrow tracks
inside a miniature one, but if space allows for a more conventional one as I
had anticipated doing.
With 2 narrow pieces of "masking tape" I laid one, leaving just the outside
edge of the 250K track exposed and another one on the paxolin, leaving a 2mm
strip un-masked .
Squirted out some "RFI/EMI" screening black (dries to slightly brown side of
black) stuff is , may be useful for very low ohm but probably too brittle
and does not adhere to impervious surfaces.
My can of it is RS stock number "551-570"
it contains nickel and air drying acrylic resin , MEK and xylene
Spray into a pot and then with a toothpick lay
down a thin line of aquadag, or whatever it is, along about 40 percent of
the track (for comparison purposes) . Silver paint would have done instead,
but just a prooving exercise.
When dried that measured 23 ohm , over its length, and about the same for
that portion of the pot track via the 2 relevant terminals.
With a razor made 3 cuts, so divided the aquadag track into quarters. Then
some nail varnish just over the aquadag and onto the remaining paxolin just
to confirm no adverse reaction and a suitable fixative and rassembled.
I suspect the characteristic would be a bit wavy rather than linear, if
properly plotted out, but results
5 percent 10 ohm
10 % 400 ohm
20% 1.8K
30% 8K
40% 42K
50% 66K
and then much as the original 120K to 250K of course
Closer checking with vernier calipers and DVM there was an abrupt jump at 23%,
from 1.5K to 3.4K.
I didn't bother with the following, but
For the upper section, for a log characteristic, then lay down a full length
aquadag track and make more , closer spaced cuts, and also a 1K and 20K
resistors or whatever , selected to fit after measuring
the fudged values, external to the pot. Also means a fall back
position should part of the aquadag track break away and leave
something close to an all or nothing non-pot.
Other non-linear characteristics , other than log could be so fashioned. And
as the thickness of the aquadag track is not critical then rotational pots
could be so converted almost as easily. The only criteria are higher ohmic
value to start with than final requirement, low wattage and enough space to
lay down these extra tracks.
Pinch wheel rubber
If about 14mm diameter then "8mm gas pipe" from camping shops
for LPG use may be suitable if mounted on some sort of cylinder.
Valve bases
For lightly abraiding the socket pins.
Some bronze multistranded picture frame suspension
wire. Heat up and melt solder onto a length of it
to form a rod. Gives a ridged structure, from the twisted wires, to run
into the pins.
Bulking up with solder , for the
wire as it comes then the diameter of the soft abraider would be about 1.2 to 1.5mm
. If you twist up tight before soldering then about 0.7 to 1mm so some
latitude on diameters. Untwist a bit would give slightly larger diameter.
Perhaps twisting 3 strands together before soldering would be suitable for
larger pins. Push/pull/twist in the sockets to clean without abraiding the
plating.
For in-situ replacement of fork type pins where
the gap is opened out or one tine is broken.
Bend the rearward extension of the pin back into
alignement with the internal part.
Cut close to the matrix with end snips.
Push with dart point/ pull the remnant through
and clean up the wire ends by cutting if
the remainder is long enough.
Rob some pins from another socket and introduce
from the inside out.
Replacing screened valve base when plain socket only is available.
Demount the original and grind off the swaged part that grips
the plain base. Open up those edge section a bit and it
will probably seat over the replacement base, deforming the surround a bit
and clamping together. May have to grind 3mm or so off the
bottom of the screening can to clear the new base flange.
Straightening tool for tube/valve pins
I will try a ring of 9 Souriau connector pins (just the right bore ) on the
pins of a brand new valve, setting the pins in heat settable fire-cement.
Then use preceeded by individual pin straightening with parallel jaw pliers.
A 20 minute job making a straightener and pins are now sober. It would have
taken that to track down a UK supplier and order one and then 3 days at
least.
Also my method could be used for non-standard pinnings
I use both types of fire cement for all sorts of uses.
But mainly the heat setting type as its cheaper for bulk use compared to
epoxy and it of course takes heat. It can crack so reinforce with car repair
glass fibre/fiber mesh if that is critical.
Recently bought, as I used up the previous 1Kg, tub of the heat setting type
called Pyruma fire cement, so still sold in the UK.
To cure, I use a hot air gun on a low setting for an initial surface
hardening , as in this straightener job. Removed the valves (I used one at
either end of the pins when it came to it, each covered with PTFE tape to
keep the valves clean) after intial surface set. Then heat up at full whack
to fully harden the resin , its then as hard as stone in a few minutes.
The other type , non setting, trade name may not exist any more Hermetite
flue jointing compound. Mine is probably 20 years old but still useable. I
tend to only use it when I've robbed a right V & W halogen bulb from one
dicroic mount and are having to, suck it and see, adjusting the position in
another size/shape of reflector, also robbed.
Desoldering compact slider pots that are surrounded
with surface mount and through-board vias all over the place,
with no colateral damage.
Cut off the casing tinplate tags as
far down as you can and then gring down to the board,
with a centride burr in a dremmel. Then desolder
the active pins. If repairing and reusing then
make electrical connection with the cut down stubs
and for structural integrity add hot melt glue around the casing wherever
there is suitable spaces.
Push on star grip circlip fastener
Perhaps doesn't have to be round but with a star-form 4 way cut
so when pushed onto a rod, the star bites-in to act as a circlip.
Ground a seriously rusty carpenter's chisel, ground to just 3mm or so
wide and cut through the steel laid on a copper block
and hammering.
NiMH batteries, recovery from 0 volts.
If run down to zero, apparently unlike
the Ni-Cad situation of metal dendrite whiskers
forming and not very long term successful fusing
them away. NiMH ones do recover quite well
from zero.
Hint for using silver loaded paint
When repairing multiple printed silver tracks on membrane
keyboards etc lay down a large patch of silver paint.Then
when dried, with a straight edge,razor and jewellers screwdriver
scrape away the paint between the tracks,much easier
and neater than trying to paint thin lines.
Salvaging small SMT devices,eg 3 or 4 leg transistors
Score round each leg with a scalpel then cut under the pad
of each leg to remove.Then clean up each leg with a soldering
iron when demounted.
Searching for a buried wire or even trace tracing in a
multi-layer board.
Where you can only connect a signal to one end.
Make a sniffer from the pole pieces and coil of a
high ohmic relay.
I used a shallow flat pack (small diameter solenoid) 48V
relay as you used to find on
telecom boards, this one an NEC MR48S 24, 4200 ohms,
fixed to an empty ballpoint pen barrel.
Connect to an op-amp with high gain and feed into
a crystal earpiece. Connect an audio sig gen set to
about 1KHz, square wave and high amplitude
and sniff with the pole piece/s in
translating and rotation to zero in on maximum tone,
position and direction of the conductor.
For more discrimination connect a DVM on AC range
in place of earpiece and set sig gen at about 8KHz
for highest response.
For single connection, ie not current through the
conductor then detection distance with earpiece
only about 10mm but about 50mm for 8KHz and DVM.
Miniature thermal switch
For temporary use only , for thermal runaway situations
when checking power amps or switching transistors.
Fix one of these to the body of a transtor temporarily
wired into the supply.
A small button click switch , where you've previoiusly
depressed the button while covering with hotmelt glue.
The button will release if the hotmelt melts
Source of ferrous shim of 0.03 mm or 1.2 thou/mil thick
Open up one of those anti-theft tags from shop
bought items.
Repairs relating to temperature effects or defects that appear
a few minutes after switching on the equipment.
Use a hair drier or hot air paint stripper on a very low setting
on areas of circuitry that could lead to the observed fault to
induce the fault earlier. Use freezer spray to try localising
by reducing the fault condition (cessation of oscillation
in temperature induced gain change parasitic oscillations).
Beware temperature effects do not necessarily warm the whole
component it can be a very localised effect at one junction say.
Poor man's conductive epoxy
I have some 20 year old silver loaded paint/ silver conductive paint
Electrolube SCP003 ,kept in a 35mm film pot, in fridge. It cannot be called
paint now, but not totally dried out, more like paste. I mixed a little with
the resin part of some also quite ancient but good epoxy , bulk 1kg purchase
about 10 years ago, also kept in fridge (red cap = resin ? any standard ?)
Left the mixture overnight and it was same pastiness, ie no/little cross
reaction.
Prepared a pair of 1.65mm wide , flat , brass pins, by cleaning with
methylated spirits only and drying off, deliberately no abraiding. One pin ,
wrapped 2 separated loops of 0.02mm insulated wire around as a precise
separator.
2 hours ago wixed with the hardener and placed between the spaced crossed
pins, held with a wooden clothes peg , holding force about 0.5 Kg.
I just tried measuring resistance between pins. Did not think , in advance,
about soldering fine wire to the pins for ohms testing. Connecting 2 croc
leads to the crossed pins, still with clothes peg, resistance varied between
3 and 35 ohm , holding in fingers. Highly sensitive to slight forces in
semicured state.
A day later
Minimum resistance about 200 ohm, shear or peal strength ? was 90gm at 13mm
distance to shear the joint.
Proportions were something like 2 Ag paint:1 hardner:1 resin , so perhaps
4:1:1 would be better and also newer product presumably. So for CMOS type
currents application not car windscreen heater repair use. In
earnest, abraid any amenable surfaces for more key.
For tracking down squeals and other noises in electomechanical equipment.
Obtain a medic's stethoscope (or make up an approximation from 2 pieces
of polythene tube). Remove the diaphragm and fix a small length of
polythene tube instead. Wave the open end in and around the piece
of machinery to find loudest point.
Accurately measuring diameter of very fine copper wir
For 2 reels of enamelled wire labelled as 2.4 thou/mil and 2.8 thou/mil and
my micrometer that has a 2Kg closure force (just checked via spring and
kitchen scales) before the torque clutch disengages.
Compressed the spring and measured the length when clutch disengages.
Find some nuts/washers etc of that thickness and with spring
inside and over kitchen scales , push down until level with the packing pieces.
I have to DIVIDE the reading by 1.15 for the 2.4 thou wire and 1.25 for the
2.8 thou wire. I suppose this is to do with the enamel thickness and the
wire is specified as the metallic diameter. I assume my wire gauge v
wight/resistance tables are for bare copper, not actually specified.
Proper engineers use feel , not the clutch, perhaps I will repeat sometime ,
presetting the micrometer in 0.005 steps until I can
feel resistance to the wire passing the faces.
PLCC chip removing tool
Another use for traditional sprung wooden clothes peg.
The spring, cut back one of the return parts to leave
a small stub . Hold in some large pliers and introduce
in/under, one corner at a time, of the chip, pulling
straight, not levering on the plastic surround which will
split. I don't
like trying to remove at both sides simultaneously.
Replacing audio cassette tape heads.
Although the engaging face dimensions of the head are the standard the position
of the mounting plate spot welded to the head is not standard when trying to
find a replacement. On the replacement head break away the mount and burr down
any lumps with a Dremmel and ball-mill. Do the same with the original and swap to the new head.
Bind the 2 parts together with thin copper wire and immerse the wire in epoxy glue.
Threading fine copper wire through sleeving
I needed to pull some doubled up , 0.22mm (9 mil, SWG 34, AWG 32) Cu wire
through small bore sleeving. Some 0.3mm nichrome heater element wire
stretched out until straight, worked rigidly enough, with "peristaltic"
movement as a mouse (does this term translate across the pond ?) . I was
probably lucky that "soldering" the copper to the ni-chrome made a
sufficient strength overlap join to pull 1 wire through but I doubt it would
work pulling 2 through, soldered or glued, because of added thickness of
overlap bulge, when I have to repeat the process for the double one.
Butchered mini-din socket pins are a bit too big diameter.
Angling supplies shop for trace crimps, 0.8mm internal bore seems the
smallest they go to. Grinding a lead-in chamfer on the leading edge , and
feeding through a tiny funnel of talcum powder did not work.
While at fishing shop , bought a spool of fine line. Just to try the
following but wasted my time.
Connect a source of vacuum at one end of the sleeving core and introduce
some fine nylon line at the other, set turning on a small motor .
In the end slightly flattened the end of the ni-chrome wire
on a small anvil. Curled it over into a loop that could take a
doubled up length of 6 lb nylon fishing line and
pulled that through. Moved the tug point and introduced
the doubled Cu wire and pulled back through.
Keep hands dusted in talcum powder, not soap or silone oil
, extended surface tension or whatever the problem hinders
with tight sleeving.
Caution about blowing out amps etc
Just as I was blowing out, with a 1KW Martindale, the crud of ages from an
amp and its 12V fans.
Thought - these fans act as DC generators if blown forcefully. Presumably
you could damage, if driving negative volts into ps or fan contol circuit.
Next time I will jam something in the blades before zapping one with 1KW of
draught.
As an experiment with no load other than DVM , neg to neg, 12V dc , 0.2A fan
, 1 foot from the nozzle of a 1KW fan then 17volt output from the fan I
tried.. Must have contained segmented magnets.
Turnuing the fan the other way in the airstream still produced "positive"
voltage.
I didn't try it any closer.
Making sheet/slab of hotmelt glue that I call Glass Reinforced Hotmelt Composite ( GRHC ).
If you use high temperature melt hot-melt glue then could
be used to somewhere between 100 and 150 degrees C, would soften but not melt.
Clamp one blade of a pair of parrot beak secateurs (plus
spacer if required) to a metal block and clamp block
to bench. Chop hotmelt glue stick into fairly even thickness
discs. Cover 2 sheets of flat and thickish metal with
covering of turns of thick formulation plumber's PTFE
or some other silicone or ptfe sheet.
Preheat them in a warm oven or hot air gun.
Bound the area with 2 rods of required sheet thickness.
Arrange discs between rods and fill in gaps roughly
with hotmelt from the gun. Heat the whole lot
with hotair gun on low setting and when fully molten
squash with the other plate plus weights.
Or use the extruder attachement to the hot melt
nozzle explained elesewher ITF. Measure volume of
stic/sticks to gauge the area for a given depth.
Needed to make some stronger sheet so reinforced
with some fibre-glass woven glass mesh
http://www.diversed.fsnet.co.uk/mesh_hotmelt.jpg
Notice surprising lack of bubbles in the cooled melt
except for line of bubbles where the melt has
spread beyond the original mesh covering
This use needed to go round an obstacle and then resealed
for damproofing.
So needed to cut it but then still not have a weak point
at the join. Hence the skewed mesh in the pic.
Formed a pleat, temporarily held with a staple
before swathing in hotmelt. Melted thinnish flat bands
of hotmelt (tip below) top and bottom over the mesh avoiding the pleat
areas by about 5mm. Rods here were 2.3mm diameter and mesh
.35mm thick. Removed staple, low temp soldering iron
melted along intended cut line, cut the mesh. After
fiting over the job , plenty of mesh overlap to remelt
together between a pair of small PTFE covered plates.
For plain sheet without glass fibre I've reliably made
1mm thick and thinner is possible. Ideal material for
coating the outside surfave of the cardboard of ordinary
speakers to make damp proof for (covered) external use.
Producing thin bands of hot melt string.
Rather than circular section explained elsewhere in tips.
Small length of copper microbore tube. Belled a bit
to sort of fit over the metal extrusioning end of the hot
melt glue gun. Drill a pair of holes in the copper
for wiring back to the gun with 1mm wire. Squash the
open end , with a thin spacer in place, in a vice to produce
a narrow slot, to extrude through.
Intermittant combo/ amp vibration fault inducer
Anyone have any advice beyond just rattling connectors , contacts,
components etc with a bit of plastic pen barrel., plus hot air gun heating /
running amp into dummy load?
Turning a blunt point to a 5mm nylon bolt, in a pencil shapener, and fixing
the bolt head in the collett of a powered engraving tool, with a 5mm back
nut, makes an excellent vibration fault indudcer. You do need to use phones
via an attenuator to monitor the audio out, over the engraver noise.
Cutting a hole in 1.5mm/ 0.06 inch thick steel
The sort of gauge used in casings.
Say something like a 3 inch diameter hole, fairly neat, so something neater
than chain-drilling then filing.
No fancy tools available eg no slow rev hand drill, fly-cutter on a mill etc
No proper mechanical engineer was around to see the following
Took about half an hour to cut a very clean hole.
Used a 72mm downlighting holesaw, working from one side only.
mm graph paper behind
42 teeth on the cutter so decided on keeping every 6th tooth and grinding
back the rest. Black felt tip line indicates one of the untouched teeth.
Hole is quite round and very clean, no wavyness, nicks etc and
eccentricity varying only
between 71.9 and 72.2mm in diameter.
Left image is the approach side and right image is the breakout side.
The grey smudgey line on the left image is my felt tip mark to show where to
put the plasticene bund to retain a pond of cutting oil,
not anything due to the cutter. I carried on
cutting until broken through 3/4 around and then hand snapped the remainder.
The breakthrough side had a thin and narrow swarf ribbon perpendicular to
the sheet which cut away/fettled easily with a stout knife, the only finishing I did
before taking the pics, other than washing off the cutting oil.
As a standard hand drill in a drill press, not geared down, I had to stab at
it, ie spin up / drop+cut / release/ spin up ...
Had to pack out under the packing slab of wood under the
steel as it was obvious the cut was uneven , missing part of the
ring.
For small (smaller than fan grills) for kit casing
Added vent hole covers - use a kitchen sink
strainer rivetted to the casing after cutting a
hole with Qmax cutter or similar.
Repairing broken shafts on hard to replace pots.
Remove the pot and disassemble. Assuming
reinforced plastic use a Dremmel and carbide
coated pointed cone tool to form an axial
hole to take a self-tap screw, with
head small enough to pass through the pot bush.
When confident the screw can bind well enough in
leaving a reasonable amount of thread exposed.
Surround with epoxy and with a bit of sleeving , axially split,
for later release to reform the missing spindle.
Owner butchered mains fuse holder caps.
Wrong sized scrwedriver blades hacking away
the bakelite even down to live metal.
I so often see it but rarely have an exact replacement due
to all the variables, 20/1.25,1/4 turn / screw, deep/shallow,
deep thread/ shallow thread etc.
For recessed type, cut off the end of a "proud" type that you
know has a recess inside the bakelite cap and has barrel not
larger than the recessed one.
Cut off the cap, clean up the cut surface
and glue with epoxy to the original.
Test tape for checking end of tape reversal / stop problems
Find an old cassette type computer games tape of short run length
and erase the programme
Identifying surface mount devices from the
alphanumeric codes printed onm the top
Ranking of keyphrases to hit SMT codes
in search engines, eg CO5 in conjunction with term "surface mount" so only a crude list
"device code" 10000
"marking code" 8040
"SMD type" 5400
"top mark" 2400
"device identification" 1360
"standard marking" 375
"device identifier" 645 ( not very SMD specific)
Also for Google searches, the NEAR AND function,
narrows the target
http://www.staggernation.com/cgi-bin/gaps.cgi
oddly just trying it you have to place "marking code" in the first box and
CO5 in the second, (in either order), 3 words, surface mount in the
additional box may also reduce target pages.
Diverse Devices,Southampton,England
Telephone number - the same number as it has been since 1988
but email is now the preferred method of contact so number deliberately not placed here.
I devote time each day to replying to emails.
(obscure/obsolete components,second hand test equipment,
schematics etc)
Postal:
66 Ivy Rd,
St Denys,
Southampton,
SO17 2JN
England
If this email address fails then replace onetel.com with fastmail.fm or
replace onetel.com with divdev.fsnet.co.uk part of the address and
remove the 9 .
Please make emails plain text only , no more than 5KByte or 500 words.
Anyone sending larger texts or attachments such as digital signatures, pictures etc will have
them automatically deleted on the server. I will be totally unaware of this - sorry, again
blame the spammers. If you suspect problems emailing me then please try using
my fsnet.co.uk account.
More hints & tips and repair briefs on
homepage http://home.graffiti.net/diverse:graffiti.net/