Joining broken Flex Circuit material 1

[dananay]

My client repairs medical devices. We have 16 and 32 conductor flex circuits on a .004" pitch and .005" thick (substrate) that are fatigue-broken. (Replacing the flex circuit is out-of-the-question because the terminations are potted.) I believe someone may have expertise in the repair of broken flex circuits. I am interested in any applicable info or experience. I want to generate some ideas and try to develop a repair process. Can you share your experience please?

Thank you for your kind reply.

 

[itsmoked]

I cannot, in my wildest dreams, think of how to fix flex circuit that has broken due to flexing. Anything you attempt will create discontinuous points of non-flexibility that will shift great stress to adjacent flex areas causing them to rapidly fail. Better would be a redesign to reduce stress on the areas breaking.

Sorry..

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[IRstuff]

The standard copper used in flex circuits are not really suitable for applications where chronic mechanical flexing occurs. They need to have OFHC copper in the flex circuit.

You're almost better off de-potting the flex circuit and replacing it and re-potting.

TTFN

 

[MacGyverS2000]

If the flex circuit fails due to cracking, you are either flexing it beyond the rated cycles, flexing it beyond its rated minimum diameter, or both. More than likely it's a significant portion of the latter with a small portion of the former.

Flex circuits are in use every day in inkjet printers (to name a single device of many) that last for years of repeated heavy use, but their flex diameters are quite large (1-2" or greater). Can you provide some more design details, such as does the FC pass through a casing wall in the portion that actually flexes, what is the bend radius, etc.?


Dan - Owner

http://www.Hi-TecDesigns.com

 

[monkeysolder]

I've fixed a couple out of necessity / desperation before (hate those things... not durable at all). Here's what I did:

1) Delaminate (or strip away) the protective layer of plastic over the broken traces. A razor blade or exacto knife helps.
2) Cut a really narrow strand of aluminum foil about the width of the trace. A sharp pair of scissors makes the straightest cut.
3) Lay the strand on the broken part of the flex circuit, overlapping a few mm of trace on each side of the break.
4) You can put a tiny bit of glue on the ends of the strand before you lay it on the broken trace to help hold it in place. Not too much or you won't get a connection. If you have some single strands of copper wire, you could try using that in place of aluminum foil and solder in place.
5) Lay laminate back over trace and compression fit using a clothespin or paperclip. At this point you may want to add a splint of some sort to prevent further flexing at this joint and a few rounds of duct tape for effect.

 

[monkeysolder]

My fix wasn't for a moving part though. I don't think it would hold up very well over time and would transfer stress to an adjacent point, like Keith said.

If you were daring, you could try "slicing out" the offending trace at the potting junctions and replacing the whole trace with thin gauge, insulated wire. The idea is that you could then fix it to the flex cable at the potted junction, where there may be less flexing of the cable and mechanical stress.

Or, if the flex cable traces are a commonly used pitch, you could cut out the flex cable, attach an off-the-shelf flex cable header at both ends, bond those headers to the moving part at both ends, and substitute stranded-core wires for the flex cable.

 

[itsmoked]

Some nice suggestions there monkeysolder.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[ScottyUK]

There are some conductive epoxy materials which are widely used in the hybrid industry for semiconductor die attachment. They aren't desperately flexible, but if it is a crack you are trying to repair they may be a viable solution. Some are definitely less brittle than others - it's worth checking their properties rather than buying the first available one. Alternatively such an adhesive might let you attach a new section of conductor to bridge the damaged section. IIRC these flexibles are anything but solderable.


----------------------------------

I don't suffer from insanity. I enjoy it...

 

[jimkirk]

These are medical devices. Do lives depend on their function and reliability? 4 mil pitch! That's what, 2 mil conductor & 2 mil insulator? Human hair dimensions.

Personally, if I had to depend on that, I wouldn't trust anything but a full replacement of the flex circuit and/or any other circuit required due to potting.

If they are not life critical, then maybe some of the previous suggestions will work after a fashion, but flexibility and reliability will suffer.

 

[dananay]

Thank you all for sharing your experience and recommendations. To answer a few questions raised:
- The bend radii is <1/2"
- The medical device is a class II diagnostic instrument
- Yep - .002" spacing between the .002" conductors
- The duty cycle is fairly low in that for each diagnostic treatment, the flex circuit instrumentation is mechanically aligned only once or twice, but obviously the design life is exceeded.
Our proposal to date is this:
1. Trim the severed circuit;
2. Remove the top layer of insulation by scraping, shaving or polishing off the Kapton-like coating;
3. Place anisotropic tape (conductive in the Z-direction but only in stripes) on the exposed conductors, and
4. Marry another flex circuit "patch" on top.

Have you heard of anyone developing a process like this?

 

[MacGyverS2000]

Sheesh, I never noticed the 4 mil pitch spec in the original post... yikes! Many board houses consider that pitch to be at the high end of their capabilities (assuming they can handle it at all) when used on rigid PCB. To do the same on flex is scary to think about.

Be that as it may, you should have no problem getting a couple of flexes out of it without breakage, which leads me to wonder about the abuse they're receiving during the alignment process. Maybe THAT'S where the company should be looking at, its alignment handling procedures.

Before I felt comfortable doing the anisotropic tape method, I would verify that a misalignment would not cause an incorrect diagnosis. For example, let's say the track's were misaligned, or one track shorted with another, and the tracks in question were used for parallel data transmission. Suddenly bit 5 becomes the same as bit 6 because they're now linked electrically, with the resulting data declaring a terminally ill patient healthy as an ox (maybe their white blood cell count dropped by half). Major potential lawsuit there.


Dan - Owner

http://www.Hi-TecDesigns.com

 

[dananay]

Good point Dan.

We validate all of our equipment against known standards. If there was an error in the data due to cross-talk/opens/shorts,etc it would show up in testing. Each one is 100% inspected.

 

[itsmoked]

Wow I didn't notice the pitch either.. what's with us?

If you value your time at anything over $20 an hour I think trying to fix those would be a waste.


dananay; Is this a product you sell and are trying to fix returned ones? Or are you a user? How many are you going to desire fixing?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[IRstuff]

time for new reading glasses ;-)

My client repairs medical devices

TTFN

 

[itsmoked]

New brain...

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[itsmoked]

I wonder if my lips move when I read..

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[dananay]

Hello Keith:

We repair equipment and save our customers on the average of 80% compared to buying new. These products are not operational when we recieve them due to a multiplicity of reasons - broken flex circuits is just one failure mode that we currently do not have a process in place to repair. I estimate we could be repairing 5-10 per month.

 

[itsmoked]

I see. Thanks for filling in the blanks.

I have a thought. What about making a connector that the flex can essentially plug into. So you cut the flex with scissors. Plug it into a connector. The other end of the connector has elastomeric cable going to another connector. You cut the other end too and plug this connector on to it. Keeping this whole thing in the mass termination realm.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[MacGyverS2000]

Keith,

That's the only viable solution I see... it's difficult enough to get a clean strip of flex conductors once they've been laminated, but to do so on such a tight pitch is going to be hairy, at best. The only major problem I see at that point is finding a connector that accepts 4 mil pitch... could be a toughy, though I haven't spec'd anything like that before.


Dan - Owner

http://www.Hi-TecDesigns.com

 

[machmech]

Hi Keith

Would it be possible to use a ribbon cable from a ink jet printer? This link has many variations.

http://www.digikey.com/scripts/dksearch/dksus.dll?Criteria?Ref=66980&Site=US&Cat=33030934

Chuck

A government which robs Peter to pay Paul
can always depend on the support of Paul.
- George Bernard Shaw