Conductive ink or flex glue to repair keyboard

[soldar]

I have a keyboard which was damaged by water years ago. It has the commonly used flex PCB with traces and in two or three places they lost continuity. I repaired it by taping a thread of copper wire over the damaged tracks and it has worked well for maybe a couple years but it starts to fail after some time; I guess the copper oxidizes and stops making good contact. I have repaired it twice and this would be the third time.

In the past I asked about conductive ink and similar solutions but they were too expensive and / or complicated. I also tried home brewing but that did not work out.

So I am asking again for some cheap and easy conductive ink or flexible glue. Rather than just tape the copper wire to the plastic i would better glue it with some conductive glue.

And, just to preempt the suggestion: I have several reasons for repairing this keyboard and not just replacing it. I am not interested in hearing how cheap and easy to get a new keyboard is. This thread is about repairing the conductive tracks of the flex PCB. Please respect that.

If everything fails I guess I will just have to repeat the process of opening, cleaning and taping new wires.

[jpanhalt]

I have not done that type of repair, but I have been interested in conductive inks for years.  There are at least three types (maybe more) that are based on different principles.

The first and most recent type commercially available is similar to the very old Tollens reaction.  It deposits pure silver nanoparticles and shows conductivity about the same as pure copper.  It can be soldered according to its developers.  Although Tollens is a very old reaction and was/is used to add a reflective coating to mirrors, the new reagent can produce thicker layers.  It was patented in about 2012.  As you may not have access to the journaal in which it was published, I have attached its recipe.  It is available commercially and last time I checked, it was quite expensive.

The second type uses a conductive material (e.g., carbon or metal) paste in a supporting matrix.  That type, based on advertising,  seems to be the most common type commercially available.  It is relatively inexpensive. 

A third type is made of conductive organic polymers.  Here's an old review of those materials: https://secwww.jhuapl.edu/techdigest/Content/techdigest/pdf/APL-V15-N04/APL-15-04-Poehler.pdf  They may bemore available today.  I haven't checked.

The only keyboard I have disassembled appeared to use carbon paste paint for the conductors.  What did you use before going to metal strips?  If you need really good conduction, I would try the Tollens /silver nanoparticle product.

[soldar]

Thanks. My recollection from my last repair was that the tracks were conductive but did have measurable resistance. I tried some concoction with carbon or something but it did not work and I just reverted to taping the copper wire to the tracks.

Searching for conductive ink on eBay gives some results for about $10 and I would have to wait like six weeks and who knows... Last time I tried the carbon powder recipe and when that did not work I just reverted to the old system.

Now I am thinking I do not need to rebuild the entire track and it would be enough to assure good contact at the ends of the wires.

I don't know if I should try to make again some carbon-based paste. Anything with silver is going to be more complicated and expensive.

I might give it a try. It seems strange that this should be so difficult.

[jpanhalt]

The first conductive ink I mentioned ("Tollens") is distinctive in that it is effectively clear in the bottle.  The second type is opaque from the suspended particles.

[MegaVolt]

Soft pencil helps me out

[richnormand]

Tried all of the above over the years with mixed results. Particularly in terms of durability. Many time the conductive glue applied to the conductive trace or pad on the button would flake off or come loose or just wear out.
Best results using carbon glue was with Pelco colloidal graphite with isopropanol base. It is available from Ted Pella Inc. for scanning electron microscope supplies.

The lasting solution (for me) turned out to use kapton tape and click domes for the actual buttons. The idea is to have a low profile, low pressure snap dome with a large surface contact to short the traces (not a small single point contact that could miss the adjacent trace. This is unlike how domes are usually used)
Clean the contact area with IPA. Cut a kapton tape with a hole a bit smaller than the trace contact area and the dome outer ring with a scalpel. Tape should have area outside to fix it in place on the pcb. Put the switch dome over the hole making sure it is isolated.
Then another kapton tape on top to secure in place when aligned.
If done appropriately with thin tape and a low profile dome you can reassemble the original button on top.
You might have to reposition the dome slightly to ensure a good contact.

For example Snaptron has assortment snap dome kits at very reasonable prices.
I have done several keyboards, TV remotes, low profile switch on my car and computer mouse with this method.

[soldar]

It is a coincidence that this thread was revived just when I was thinking about this.  I have been searching for some carbon electrodes I have somewhere (last time I saw them they were near the welding equipment) because I was thinking I could scrape some carbon powder and it would have good conductivity. But I just cannot find them so I just keep hoping they will show up when I am looking for something else.

I am somewhat surprised to see something as simple as conductive paint is so complicated and so expensive. I am just trying to repair some keyboards but I should do something because I have one computer without a keyboard and I need to move keyboards around which is inconvenient.

Now that I remember, the central electrode of carbon-zinc batteries is carbon which should be usable for this purpose. I zinc.

I probably do not have any "heavy duty" batteries though. ...

Back to searching for the carbon electrodes.

ETA: I might just order from ebay
Silver Conductive Glue Wire Electrically Paste Adhesive Paint Pen PCB Repair Kit
https://www.ebay.com/itm/313931025507

[richnormand]

I tried  the silver conductive paint several times. Once dry it either came unglued or spawned flakes that made the contact erratic.
Each formulation and flex geometry is different so it might work for you.
Best of luck with it.

Best results I got was the Pelco colloidal graphite for SEM use.
An alternative might be the small copper wires but with bismuth based low melt solder if the flex cable can take it. Some alloys can melt at very low temperature.

[jpanhalt]

When "you all" refer to "silver conductive" paint or "I tried them all," please be more specific.  Was it pure nano-precipitated silver or some metal in a paint matrix?  In other words, was it the "Tollens" type of conductive application that is pure silver and is solderable?  It presumably can be built up in multiple layers to almost any thickness.

The original article in JACS  reports that the newer  version has conductivity almost equivalent to pure silver and is solderable.  I have done the Tollens reaction several times to silver mirrors.  There is no comparison between that and paint.

Unfortunately, the Tollens-based product is quite expensive, but it was published in JACS before being patented, and while patentable in the USA, I suspect you can prepare it at home and use it without worries of infringement.

[richnormand]

Good point. "I tried them all" ... there might be something new now but several of the stuff I tried was also made for special applications I wont get into.

The silver print coating I (still) use is the GC Electronics 22-202.
It is the 31 grams bottle. That means I cannot afford it now!

The other conductive glues I tried over the last 40 years both in the lab, space hardware, optronics and at home context I lost track of as it ranges from industrial to commercial and consumer use.
The colloidal graphite is about 10 kohms per mm or less. Quite suitable for a small crack in a low current sensing application (keyboard, remote, etc) and stable.

Just saying what works best for me after years of "testing/trying" ... YMMV...

[jpanhalt]

A distinctive characteristic of the reactive/Tollens-type is that they are practically clear. They may be slightly tan or turbid with age.  That is, they are like any solution of a soluble silver salt.  Silver nitrate was used originally.  They should be stored protected from light, so don't be deceived by a black bottle.  Of course, the price of silver is very high now. 

The silver depositing ink contains:
Silver acetate (AgCH3CO2)
Ammonium hydroxide (NH4OH)
Formic acid (HCO2H)

Originally, it was used to distinguish aldehydes from ketones.  Any aldehyde can be used.  For silvering mirrors, glucose was often used for the aldehyde.