Ribbon cable / IDC connector failure modes?

[Alex]

I need to build a cable tester and I am wondering what the failure modes of say 50-way ribbon cable terminated with IDC connectors are.
There are only so many ways a cable can fail; I am particularly intrigued by a failure mode where adjacent connections are shorted?
I see some cable testers checking for shorts, for example by individually powering each conductor. Is this something you have experienced?


Alex

[wraper]

It's nearly impossible for them to be shorted. Ribbon cable and connector need to be misaligned so much that shorting would be the least significant problem with that cable assembly.

[Tomorokoshi]

I've encountered these failures over the years:

1. Blade spreading
The blade into the cable slowly spread out over time, leading to a disconnection. Using a pliers to bend the blades back in fixed the connection.

2. Adhesive flow
The IDC connector had an adhesive on it. Over time it deteriorated, which flowed around the conductors. Once the blades spread, it made a nice insulator.

3. Crimp misalignment
Some of the IDC blades were misaligned, so when the crimping was done they got bent over instead of being held in alignment to cut through the insulation properly. This made for only one blade making contact.

4. Dangling external conductor strand
When the cable was cut, a strand of conductor was stretched out. If it makes it to a couple millimeters, at some point it could be bent over and short to an adjacent conductor.

5. Dangling internal conductor strand
When the cable was crimped, a strand of conductor was pushed through the insulation by the terminal blade. Hopefully these are caught in the test.

The problem is that these would have worked fine when first manufactured. The failures happened over time, but it's entirely possible that some failed within some reasonable warranty period, as opposed to decades later. A lot of IDC assemblies may not be electrically tested, only visually inspected.

Checking for basic shorts and opens is a good start.

[thermistor-guy]

I need to build a cable tester and I am wondering what the failure modes of say 50-way ribbon cable terminated with IDC connectors are.
There are only so many ways a cable can fail; I am particularly intrigued by a failure mode where adjacent connections are shorted?...

Yes, I experienced adjacent shorts not so long ago, when fitting 64-way twisted pair ribbon cable with DIN41612 connectors (64-way IDC), for a test jig.

The problem was the ribbon cable - the conductor spacing across the cable was a little uneven. So it would match the connector contacts at one end of the connector, but not at the other end.

The solution was to carefully inspect the cable alignment all along the connector, before clamping the cable down into the connector, and stretch the cable (in cross-section) slightly where needed. Sometimes I needed to do this a few times, so the cable would sit just right.

The cable was a major brand - Amphenol or 3M - from a recognized supplier - Digi-Key or Mouser. Don't have the exact details handy. DIN41612 connectors were Harting.

Once I paid attention to conductor alignment, I had no more adjacent shorts. I tested each cable fully, and now regard adjacent short test as a routine part of the assembly process.

[HwAoRrDk]

A fault with ribbon cables I have occasionally seen is where the ribbon has been slashed by a sharp chassis edge (or similar), leading to an open in one or more conductors. Sometimes the open is intermittent and only occurs when the cable is flexed.

Also sometimes where the outside edge conductors have had the insulation abraded and worn away, allowing the possibility of short to chassis or components. Obviously, no way to test that in isolation, though. But one would hope that such damage is easily visually apparent.

Other than that, as others have said, failure of the IDC connector - loose, etc.

[Alex]

It is, in a way, impressive how many things need to be right to get long-term reliability out of IDC and ribbons. I will make sure to account for shorts and intermittent connections in my tester. Thanks very much all!

Alex

[powerchisper]

Recently I had a machine with three ribbon cables, which looed ok.
Then I saw the bare tips of the cables, here it was cut, and they were green fom copper oxide, which had penetrated to the pins.
Instead of 0 ohms , you can get up to 12-15 ohms per strand.
Ths machine I'm woking on shuts down the CPU if any of the data/adress lines are cut and the code can not be executed,

I had to copy three new cables as the only solution.

[chickenHeadKnob]

Back when I was just starting out (1983 ish) the lab I was working for was using T&B Ansley ribbon connectors. The female (socket) version of these had the propensity to break. In particular the contact blades that are normally pressed against the inserted .1 inch header pin would loosen. This resulted in numerous intermittent failures in fielded equipment, What a pain in the ass that was to track down. We switched to samtec and the problems went away. That was nearly 40 years ago and perhaps Ansley fixed the design but I will never touch them again.

In real long term use  the ribbon cable insulation will go the way of all PVC/vinyl and crack as it loses its plasticizers.

[poorchava]

We had an issue with a 1.0mm ribbon pitch 50p connector. It was very fragile and production would very often break the strain relief when inserting the connector, sice it required quite a bit of force due to high number of pins.