purple/brown tinted discoloration on internal hookup wires

[Ken9053]

I have been reading this forum since before I knew I wanted to be an engineer, finding answers to whatever question had arisen from that weekends project. I finally have a question I can't find answer for, so I created an account!

The company I work for manufactures an rugged embedded computer. When the units come back for RMA or Upgrade, we have noticed that some of the internal wires, initially white, have become tinted purple/brown. Because there are no electrical engineers here, it is on me to solve this mystery.

The victim wires are all 24awg stranded copper hookup wires, 0.010" pvc jacket hookup wires, dearborn 172407-1000 family to be exact, 6-10" long. The discoloration is more pronounced at the ends of the wires, and less pronounced where the wire was covered by a label, tag or zip tie. In the case of particularly long labels (30mm) the discoloration is completely absent near the center of the label.

More details:
- Only these 24awg wires are discolored. Other white wires, 22 and 24 awg Dearborn pvc jacket hookup wire, are as white as when they were first assembled.
- All these 24awg wires carry serial data, 5mA max current, according to the serial board datasheet.

My first thought was heat, since the enclosure internal ambient temperature gets up to 80-85c on a hot day. However, the fact that other white wires are unaffected seems to argue against that theory.

My next thought was over current, but after looking at the datasheets for our serial communications boards, I don't see how that could happen. All the ports are optically isolated, all the units in question are functional, and the phenomenon is too wide spread to be a surge or current spike of some kind. Also, the fact that the discoloration varies along the wire, and fades under the labels seems to point toward an environmental cause, instead of an electrical one.

The final detail is that the enclosures are purple anodized aluminum. The idea that this is somehow connected has been kicked around here, but I can't find a process by which the purple dye could be transferred to the pvc jackets without direct contact, and why wouldn't the 22 and 24 awg wires be affected?

I am at a loss at this point. This is not my specialty, and I have been investigating based on common sense up to now... any insight or suggestions would be greatly appreciated!


Ken Schleich
Mechanical Engineer
B.S. Aerospace Engineering, IIT, '11

 

[DRWeig]

Two questions come to mind:

1) Can you post a picture or two?
2) Your discolored wires carry serial data. Where is the other end?

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[Ken9053]

DRWeig,

http://files.engineering.com/getfil...f9f4ed403017&file=IMG_20130703_101935_609.jpg

I took several photos trying to get the discoloration to stand out, but it doesn't show up well under our florescent lights... In this assembly there are only white and black wires, so all those grey looking wires started out white!

Regarding the serial data, the end devices are located between 10 and 35ft away. Some use RS-232, others are RS-485 busses. In the photo you can see the bulkhead mounted MS connector that brings the serial ports to the exterior of the enclosure.

- Ken

 

[btrueblood]

"Only these 24awg wires are discolored. Other white wires, 22 and 24 awg Dearborn pvc jacket hookup wire, are as white as when they were first assembled."

Are the wire types that get discolored a different type/brand/family of wire, or identical wire from the same spool as the discolored wires?

Yes, dye materials can migrate, especially when they get hot, and the leached or vaporized dye material may be getting absorbed by the pvc jacket. Similar things can happen when you put colored (or black) wires next to white wires, the dye/colorants from the insulation can migrate if the insulation temperature gets high enough. Other wires may have different chemicals in their jackets, or have things like drawing lubricants (silicone grease?) that inhibit absorbtion.

You might check the spec's. on the housing post-anodizing bake-out, or do a bake out in your own shop prior to assembly. Also try running a test where you put wire bundles into a warm oven and see if the discoloration is coming from wire insulation.

 

[Ken9053]

btrueblood, thanks for the input!

All the hookup wire is from the same product family, but I mistyped in my original post. That sentence should read

"Only these 24awg wires are discolored. Other white wires, 22 and 20 awg Dearborn pvc jacket hookup wire, are as white as when they were first assembled."

Your insight about dye migration is very helpful, I had not been able to find a straight answer about this previously. My purchasing manager mentioned our housing supplier also changed anodizers last year, if their bake out process changed that could explain why the problem is starting to show up now...

I have been questioning our assemblers/RMA technicians, and they brought to my attention a special case that my shed some light on this.

http://files.engineering.com/getfil...f677b50b2448&file=IMG_20130703_133101_465.jpg

As you can see in the photo above, some of the 24awg white wires are discolored, and some are not. The difference is that in the application where this unit was used, the serial port labeled COM 3 (Green circle) was in constant use, while the other (blue circle) was disabled. Could the tiny electric field generated by the serial signals influence the dye migration? More research is required.

We have a thermal chamber I can do a bake out test in. I am planning to install a cable assembly and run a low current through one of legs. What kind of time scale would it take to see a change? Would a weekend at 80c be long enough?

-Ken

 

[Pud]

Not an EE, but Google "PVC pinking" - it may be applicable to PVC wire coatings as well, but not causing so much a furore as one's window frames and front doors!

White colourants are not dyes - they are pigments which do not migrate. Dyes tend to be used for transparents and/or flourescents.

H

(sorry for the correct spellings....)



www.tynevalleyplastics.co.uk

It's ok to soar like an eagle, but weasels don't get sucked into jet engines.

 

[amorrison]

A quick and dirty answer might be to hold some new white wire over a candle flame at various heights and see what happens.
If you are worried about the candle flames "products of conbustion" maybe a 100 watt light bulb as the heat source or a hot soldering iron.

 

[MacGyverS2000]

If you've ruled out UV issues (completely enclosed case), are you sure the anodizer is sealing the cases after coloring?

Dan - Owner

http://www.Hi-TecDesigns.com

 

[LiteYear]

Fascinating.

If you can't rule out it being a thermal effect, then perhaps the cause is inductively coupled currents well beyond the current levels of the serial signals. Ground loops will significantly exacerbate the effect.

In other words, somewhere in the cable's 35ft journey it runs alongside a power cable, which inductively couples current into the signal cable. That current finds a path to ground via the operating COM port and returns to the other end of the cable, running round and round in a low impedance loop. If the frequency of the coupled current is sufficiently different to that of the serial data, then it's possible that no operational effect will be noticed. The cables themselves however, have to carry the much higher currents and will therefore become hotter.

 

[ScottyUK]

That is a lovely theory, but can it explain why the discolouration effect is less pronounced under the labels? There's something far simpler at work than mysterious induction effects - I suspect it is dye migration from the anodizing.

 

[MacGyverS2000]

One way to test the electrical theory... place a few inches (feet?) of the wire into the container. Don't hook it up to anything, just put it in the case. If it still changes color, it's not electrical in nature. Try placing some in a UV-resistant bag, too.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[LiteYear]

Good point Scotty - if anything I'd except the label/tie to exacerbate the effect due to extra insulation keeping the heat in the cable. I also have a feeling that it'll be something simple, but dye migration doesn't completely explain why only certain cables (of the same type) are affected either.

 

[ScottyUK]

Dye migrating to a black cable probably wouldn't be as apparent - maybe it is there, we just can't see it?

I wonder if it could be a manufacturing variance in the plastic insulation material, which is why some experience the discolouration and others don't? In the earlier post there are two or maybe three shades of red too, so there certainly seems to be some variation in manufacture.

 

[Ken9053]

Everyone,

Thanks for all the advice. I really appreciate everyone's input. I am going to be running some thermal chamber tests to prove/disprove the dye migration theory. I will post what I find!

-Ken