Repair broken traces Model M?

[redwald]

I took apart a Model M to try find why Ctrl and Shift key weren't working. I found that mylar trace sheet cracked where it bends before plugging into controller board.
.

.
I've never tried to repair these before and need some guidance.
Searches here bring up copper tape, conductive pens and even windshield heater repair kits.

The copper tape seems to be the way to go, but I have a question. Since the tape has an adhesive on one side, won't the adhesive insulate rather than pass current?

Would this be the correct one to get?

http://www.kaptontape.com/Copper_Conductive_Tapes.php

I'll need to cut it to about 1mm thickness and am wondering if it will stick well enough being so small.

Any suggestions on how to make a pro repair is much appreciated.

[tp4tissue]

I took apart a Model M to try find why Ctrl and Shift key weren't working. I found that mylar trace sheet cracked where it bends before plugging into controller board.
.
(Attachment Link)
.
I've never tried to repair these before and need some guidance.
Searches here bring up copper tape, conductive pens and even windshield heater repair kits.

The copper tape seems to be the way to go, but I have a question. Since the tape has an adhesive on one side, won't the adhesive insulate rather than pass current?

Would this be the correct one to get?

http://www.kaptontape.com/Copper_Conductive_Tapes.php

I'll need to cut it to about 1mm thickness and am wondering if it will stick well enough being so small.

Any suggestions on how to make a pro repair is much appreciated.

haha..

don't try to repair that cable. usually "NOT" reliably fixable.

Look for where the traces lead, and use a plain copper cable of "similar" length. bridge both ends.n solder on.

Peace of mind, you'll never worry about jiggling the board and having the connection get disrupted.

[kishy]

There is nowhere to solder a lead to...the mylar will practically evaporate if you put that amount of heat near it.

The only option is something involving a conductive adhesive. I fortunately haven't found myself having to repair one of these guys in a while, but I did have an old container of a conductive glue that worked for repairing small breaks. The problem is if it flexes, the connection breaks or becomes intermittent, and the resistance is somewhat high so you need a lot of surface area to make a reliable connection.

If you do go the route of some kind of adhesive, lightly sand (even just a single pass) the surface of the trace for better adhesion.

My first suggestion would be to confirm that it works 100% by bridging the break with a paperclip or something just to see that there isn't another issue elsewhere. Unlikely, but worth confirming.

Rear window defrost repair products should do it, but remember that those defrost grids are essentially big resistors so I don't know how reliable that fix is. I've definitely heard of it being used before though.

[tp4tissue]

There is nowhere to solder a lead to...the mylar will practically evaporate if you put that amount of heat near it.

The only option is something involving a conductive adhesive. I fortunately haven't found myself having to repair one of these guys in a while, but I did have an old container of a conductive glue that worked for repairing small breaks. The problem is if it flexes, the connection breaks or becomes intermittent, and the resistance is somewhat high so you need a lot of surface area to make a reliable connection.

If you do go the route of some kind of adhesive, lightly sand (even just a single pass) the surface of the trace for better adhesion.

My first suggestion would be to confirm that it works 100% by bridging the break with a paperclip or something just to see that there isn't another issue elsewhere. Unlikely, but worth confirming.

Rear window defrost repair products should do it, but remember that those defrost grids are essentially big resistors so I don't know how reliable that fix is. I've definitely heard of it being used before though.

oh... hmm those are the contact sheets.. I see, my mistake.. I thought they were just cables.

[redwald]

oh... hmm those are the contact sheets.. I see, my mistake.. I thought they were just cables.

No worries tp4 

If you do go the route of some kind of adhesive, lightly sand (even just a single pass) the surface of the trace for better adhesion.

My first suggestion would be to confirm that it works 100% by bridging the break with a paperclip or something just to see that there isn't another issue elsewhere. Unlikely, but worth confirming.

Thanks kishy...I think I'll first try the copper tape. If it doesn't work at least I've not ruined anything.

I just found this thread from hasu, who chopped up one of the membranes and used copper tape, so apparently it can be done. Of course his is flat and my repair is over that curved bend.

http://geekhack.org/index.php?topic=29483.0

BTW, I tested the controller by swapping it out with a known working controller so that's ok. Those broken traces are certainly the problem as they lead right to the Shift and Control key spots on the membrane. Confirmed by multi-meter.

[wcass]

Replacement membranes are only about $10 from Unicomp. They carry most of the old model membranes.

[redwald]

Replacement membranes are only about $10 from Unicomp. They carry most of the old model membranes.

I checked there first but this is for an SSK. I only see one listed and it appears to be for the 101 (104?) keyboard. Their picture is cut off on the right side, so it's hard to tell.

http://pckeyboard.com/page/Misc/MEM

Edit: I just left a message with them. I'll update when they call me back.

[dorkvader]

Unicomp does not carry SSK membranes (sadly). This appears to be one barrier to them continuing production of them.

You may be able to roll up a new membrane to fit in the case.

[kolonelkadat]

conductive paint and done. its expensive though.

[blueSmoke]

This http://www.radioshack.com/product/index.jsp?productId=3964901 worked for me. I have completely removed the numpad cluster and rewired this with the circuitwriter to make the poor mans SSK. I also found it is pretty flexible and does not fall off/come off as expected. HOWEVER you have to follow some precautions:

1. Get the newer pens, as the older pens may be somewhat dry and not so free flowing. Chitchat with the RSH guy and be nice to him/her and he will tell you if these are new stock.

2. Play with a scrap piece of plastic to draw. This does not have a pen like tip, rather a tube and a tapered bit at the end and when you press the pen down the internal taper of the bit opens up for the fluid to flow down. So it is continuous at a certain pressure and travel, and below that it is lumpy and above that it is spotty.

3. prepare the surface with a tiny bit of plastic modeling cement/acetone. Test a tiny bit of cement over some unused tracks. It SHOULD dissolve existing track/silver as well as leave a small residue. I found this actually leaves a somewhat compatible surface for the circuitwriter to hold on (the contains a little bit of acetone? that dissolves the surface but not enough to make a solid bond).

4. You have to draw 2 times, the first time draw in one pass and let it dry under a table lamp (10-12 inches from it), leave for at least 6-10 hours. Magically it will be conducting then. Then go over it using a lighted jewelers loupe and check for any restrictions in the track. You may go over this the second time and again leave under the light for another 12 hours.

5. Note that this creates much thicker lines than the existing few micron ones. Use a scale to rest/slide your drawing arm for a uniform line.

I could get about 50 ohms between a 2mm wide track over the foil width (8-inch?). So this is pretty good. I also tried to scratch this with nail as well as sandpaper and it is holding on. It worked for me great and I believe this will work for you very nicely.

If this work, leave a note and it will be very useful for others in repair situation.

Cost:
circuitwriter 23$ radio shack
jewelers loupe 7-10$ frys/online
a steel scale 5$? anywhere
==========================
Getting a broken SSK work: Priceless

[hasu]

This http://www.radioshack.com/product/index.jsp?productId=3964901 worked for me. I have completely removed the numpad cluster and rewired this with the circuitwriter to make the poor mans SSK. I also found it is pretty flexible and does not fall off/come off as expected. HOWEVER you have to follow some precautions:

Great guide! and good to know Circuit Writer worked.

I used Circuit Works conductive pen, and it didn't work for me. It was not flexible on curved surface and very brittle after it dried.

In the end I used copper tape to fix traces instead of Circuit Works and it worked. I'll try to use Circuit Writer and follow your guide if I have to do this job again.

Get Circuit Writer, avoid Circuit Works.

[terran5992]

It looks pretty alright with me

I would have never guessed that the tracers were broken

[blueSmoke]

This http://www.radioshack.com/product/index.jsp?productId=3964901 worked for me. I have completely removed the numpad cluster and rewired this with the circuitwriter to make the poor mans SSK. I also found it is pretty flexible and does not fall off/come off as expected. HOWEVER you have to follow some precautions:

Great guide! and good to know Circuit Writer worked.

I used Circuit Works conductive pen, and it didn't work for me. It was not flexible on curved surface and very brittle after it dried.

Show Image

In the end I used copper tape to fix traces instead of Circuit Works and it worked. I'll try to use Circuit Writer and follow your guide if I have to do this job again.

Get Circuit Writer, avoid Circuit Works.

I have a feeling that I have gone thru the brittle issue. See my notes #3, this is the problem. I suspect the seller sends with a reasonable amount of solvent/acetone base, but when we buy these things and apply to the mylar, amjority of the acetone is evaporated. I found it by mistake... after trying countless types of plastic solvent etc to glue plastic cabinets that they simply dont get the bonding strength not due to the bad glue... but due to a bad/unprepared surface. So you wipe the surface with acetone and it somehow creates a very thin melted surface or something. Now if you glue anything here, it is BETTER than than JB-Weld. I have broken model-M pieces that I had to break as I could not unbond a surface just after 5 minutes of joining with the standard plastic binder.

The other thing is once you use the jewelers loupe you will discover that the conductive material is not uniform enough, and so it possibly needs a second coat. Here is a picture of my SSK track check it out. You cant remove it without a blade. Period. Checkout the below photo. Solid conductivity and bendability. You must prepare the surface before apply.

[Wildcard]

This http://www.radioshack.com/product/index.jsp?productId=3964901 worked for me. I have completely removed the numpad cluster and rewired this with the circuitwriter to make the poor mans SSK. I also found it is pretty flexible and does not fall off/come off as expected. HOWEVER you have to follow some precautions:

Great guide! and good to know Circuit Writer worked.

I used Circuit Works conductive pen, and it didn't work for me. It was not flexible on curved surface and very brittle after it dried.

Show Image

In the end I used copper tape to fix traces instead of Circuit Works and it worked. I'll try to use Circuit Writer and follow your guide if I have to do this job again.

Get Circuit Writer, avoid Circuit Works.

I have a feeling that I have gone thru the brittle issue. See my notes #3, this is the problem. I suspect the seller sends with a reasonable amount of solvent/acetone base, but when we buy these things and apply to the mylar, amjority of the acetone is evaporated. I found it by mistake... after trying countless types of plastic solvent etc to glue plastic cabinets that they simply dont get the bonding strength not due to the bad glue... but due to a bad/unprepared surface. So you wipe the surface with acetone and it somehow creates a very thin melted surface or something. Now if you glue anything here, it is worse than JB-Weld. I have broken model-M pieces that I had to break as I could not unbond a surface just after 5 minutes of joining with the standard plastic binder.

The other thing is once you use the jewelers loupe you will discover that the conductive material is not uniform enough, and so it possibly needs a second coat. Here is a picture of my SSK track check it out. You cant remove it without a blade. Period. Checkout the below photo. Solid conductivity and bendability. You must prepare the surface before apply.

Thank you for posting that, and good work.

[redwald]

Unicomp does not carry SSK membranes (sadly). This appears to be one barrier to them continuing production of them.

You may be able to roll up a new membrane to fit in the case.

Yes, you're right. I just got an email back from them.


Thanks bluesmoke and hasu for all your suggestions, tips and pictures.
That should be very helpful for a better long term solution. 

Yesterday before you guys posted your experiences and while waiting for the email from unicomp I tried an out-of-the box solution which (surprisingly) works great.

Materials:

Clear 2" box tape. The thick kind made by 3M.
2 individual strands from twisted copper lamp cord cut to 1/8" (one for each trace)

Procedure:

The part of the flat cable that is attached to the mylar is two sided. It has a mylar coating over one side (just like the large membrane)  Fortunately the other side of it is a much thinner plastic coating. This is the side you must scrape away very gently with a fresh razor blade to expose the traces. I scraped away about 1/8" (1/16" on each side of the break)

I measured the distance between the traces which is 2.5 mm.

Then on the non-sticky side of the tape I drew two lines 2.5 mm apart with a thin tipped sharpie.
I made sure these lines were on the proper side of the tape so that when I turned it over the entire cable was covered with the tape.

I then placed the two tiny pieces of wire on the sticky side of the tape. I could tell where they need to go on the tape because of the lines I drew with the sharpie on the other side.

I stabilized the curved ribbon cable by taping it over 3/4" PVC pipe so when I placed the tape with the two tiny pieces of wire over it, it wouldn't move. I pressed the tape around the tiny wire pieces so they won't wiggle when flexed.

The tape also has the added benefit of adding some strength to the rest of the cable.

That's it, and even when I flex the cable a little bit my multi-meter needle doesn't even flinch when measuring resistance.


Now I know ya'll are gonna think thats a jerry-rigged solution of the first order. Conceded, but it works and cost me nothing. Perhaps in a few years the tape will start to disintegrate, so I'll have to do it again or perhaps try the CircuitWorks pen.

[0100010]

Wasn't there a thread where someone took a 101 key Unicomp membrane, and trimmed it for use in an SSK?  Will look - I am probably mis-remembering the thread I saw.

[0100010]

OK - the thread I was thinking of was wrong :

http://geekhack.org/index.php?topic=46106.0

But the thread that would probably be of interest is :

http://geekhack.org/index.php?topic=29483.0

[redwald]

But the thread that would probably be of interest is :

http://geekhack.org/index.php?topic=29483.0

Thanks 0100010, thats the thread I referred to in the 4th post.