Is potting my boards with black pigment ruining them ?

[iso9001]

Hi,

I've got an odd problem where I have boards that are working PERFECTLY. But almost all seem to report back wrong/erratic adc readings after I pot them. (?!) Its a pic based cicuit. I thought maybe because 2 of the 3 signals I adc are 20k impediance that it was just too weak, but I now notice that another signal i have (500ohm) imp. is acting erratic too.

I'm not 'exactly' certain of what the problem is, but I had 3 boards here I've been testing all week, potted them up and 1 failed imediatlely and another is just starting now.

I'm using normal epoxy that I am adding carbon black pigment into for color (it says its burnt bone). Anyone have any ideas ?

Ihighly doubt its the CFE or Tg factor since all these are being potted, curing, and then failing at the same room temp.

 

[williB]

Have you tested the resistance of the pigment.??
like stick your meter probes into the solution..??

 

[chemelec]

Even if that Mixture Isn't Conductive, It will Definately Add UNWANTED Capacitance.

Why pot them in the first place?

If someone wants to Copy it, They will find a way.

 

[iso9001]

I'll try that resistance thing. I just don't understnad whats going on with it. Doesnt make any sense at all.


chemelec: If that were the case then why would ANYONE pot ? This is a very harsh (vibration, water, heat) application, and it needs to last years and years. I dont have a sealed case AND I dont want some backwater redneck making cheap knockoffs, I am 100% for potting but maybe you have a better idea ?

 

[williB]

how about two coats .. one clear the other with pigment..
carbon conducts electricity btw..so if your carbon black pigment has any carbon in it it will conduct..

 

[Styx]

yes coat them ist them pot them
The aerospace industry always coat their board to seal them from moisure.

Dues sound like the potting you are using is changing the boards characteristics. Get a coating formula then pot the board

 

[Agent 009]

what's the power rating for ur boards? maybe it's high enough to create capacitive coats, like chemelec is saying?

 

[chemelec]

There are Many Coloring Materials, Carbon Black is Not the Best one.

Additionally, I am Not sure if Your Epoxy is a Good Choice.

You could have a Coefficient of Expansion Problem.
**Different Materials Expand and Contract at Different Rates.
This could cause some circuit connections to Break apart.
Or a Heat Problem as it Cures.

They Do make Special Products for Potting Electronic Circuits.
I Know this Because I have also Potted many.
Not to Hide the Electronics, But to Protect it from Moisture, etc

The Epoxy I use is: "Sealtronic" #21AC-7V.
I Purchased this from "Industrial Formulators of Canada Ltd"

Hope this Helps..........Gary

 

[zevon8]

As rightly pointed out by Gary above, on top of the electrical "anomalies" that can be introduced to circuits when potted, there are the mechanical problems to resolve.

Unless specially formulated for the purpose, epoxies tend to shrink or expand on curing. Any part that generates heat on the circuit board may also create uneven stresses and damage other parts. This can completely destroy the potted circuit. Filler material can be mixed into the regular epoxy to control this but it is hit and miss.

Years ago I had produced some modules for transport trucks that needed to be potted. Several potting compounds were tried, and general pupose epoxy was one of the worst compounds tried. Most modules failed either immediately or under temperature variation.

One compound you can use for small batches is a machinable casting urethane. This is used for making prototype parts or molds. Materials such as the "master-cast" line from Kindt-Collins are very durable, have almost no expansion or contraction on cure, and are impervious to chemical attack.

I would definite use a pupose-made potting compound for circuit boards. Something like one of these:

**broken link removed**

Many companies produce urethanes just for this application, they are very durable, and flexible enough to allow thermal expansion without damage to the potted object.

If you choose a clear compound and want to color it, you can often use something like titanium dioxide to make it opaque.

 

[iso9001]

Thanks Guys,

I'll change from carbon black to somthing less conductive, however when I say I'm just using a pinch, I mean it. It takes an amazinly small amount to color it black. Is Ti02 conductive?

I'm pretty sure now the epoxy is screwing me up in some manner, but it must be in the cure process since that should be the only time this thing gets above room temp.

Its a nonaudio, low power, low heat circuit. Just a pic and a punch of dc inputs and outputs. No big deal there.

Can someone explain how the expoy can add capacitance, and what that might mean for an signal going into my adc?

I have some urethane I'm going to try. I'll have to look up the specs first but it never really fully hardens, so that should cure any stress issuses I might be having.

I'm really curious about the capactiance deal. Since I have OK adc readings in a noisy enviroment before potting, but horrible reads after potting. (?)

 

[Russlk]

The dielectric constant of epoxy is 3 to 3.5, which means that the stray capacitance will increase by that amount. That is with unfilled epoxy. The filling, if it is conductive, will increase the capacitance more. But, more likley, the conductivity of the filling is what is messing up the circuit.

I have used urethane floor varnish with good results. You could fill it with aluminium oxide or titanium oxide, both non-conductive.

 

[chemelec]

Without Knowing How much of that color or its exact properties, it is Impossble to know if that was the problem, Or if it is the Epoxy Shrinking as it hardens Which Now Sounds More Likely on the info given in your last post.

As to Adding Capacitance: Carbon is a Conductor, although a poor one.
Even if it is mixed into a non-conductive agent, And if your circuit has a high frequency oscillator, a few picrofarads can become a problem if its across the wrong points in the circuit.
Some Epoxys also can have a High Moisture Content. Especially True when potting in Wax.

Another Possibility Is Solder Flux:
I have found it Very Important to Remove All Flux and make sure the board is Absolutely dry Before Potting in Anything. Contrary to Most Solder Manufacturers Claims, Most Flux is Conductive. Additionally it can Chemically React with the Potting Material, As its Hardening to cause further problems.
This is Also Possible in the Long Term, After it Has Hardened.

I now Always use "Kester 331" Solder. It is a Water Based Flux which allows for removing ALL the Flux in Warm Running water.

Strange as it may seem, On my PI Metal Detector project, Without Removing the flux, it Doesn't work properly. It Causes Leakage Between Two Pins on one of the Transistors.

Your Urathane Foam should be OK, But I really Recommend Considering, Cleaning off the Flux.

I Hope this Helps......Gary

 

[zevon8]

One point, though it may not apply in your situation: potting is not a very good method of "hiding" your design or circuit. Many of the potting compounds can be disolved ( albeit it takes awhile ) with solvents.

Things such as Methylene Chloride will disolve many potting urethanes. It can take a few days, and is a bit of an archeological event, but can be done. I have used this method to discover why some of my potted devices had failed in the field.

 

[iso9001]

Methylene Chloride!? Geez, I've used that before to disolve glass fibre glass from some tools on another project. That is nasty stuff. It SHOULD eat right into the pcb board. I'm skeptical that pcb would be ok, also the labels on the part would be gone. Its a good deturent for sure. I'm guessing that if someone REALLY wanted it they could have it xrayed. But it needs to be done, these things need to be 100% element proof.

chemelec: I went and recleaned all the flux from my new boards (not potted yet). I'll try that. I'm using a ceramic resonator with caps built in, the timing doesnt seem to be screwy, its the inputs (esp on the adc) that are acting up. We'll see this next time with no epoxy, no pigment, and no flux.

I checked the pigment I was using. Its labeled "carbon black", but its really Calcium Carbonate (burnt bones), I checked it out w/ a dmm and could not get a reading even when I touched the two probes togather with some on each probe. It seems like its OK stuff. Just incase anyone wants to try it or somthing like it, I got mine from the university art store, $5, lots of colors.

I guess I could do urethane foam, but I also have some urethane thats used for gluing windows in panels that I'll try, its already opaque, doesnt exothem, and doesn't harden past a durometer of 40 or so. Its CHEAP too.

 

[chemelec]

Its labeled "carbon black", but its really Calcium Carbonate (burnt bones)

Well it may appear as Carbon Black, But to my knowledge, Real carbon black Isn't made from Burnt bones.

Just make sure that what ever you are using for Potting,Isn't Acidic.
That can cause Corrosion of the copper, Typically forming Cupric Chloride.
Which is Quite Electrically Conductive.

Keep me Informed how you make out........Gary

 

[iso9001]

Like I said...

Art Store :?

Maybe I should ask them what is electrical properties are

I'll pot one up using urethane goop tomorrow,

 

[zevon8]

If you use the window urethane, make sure the curing agent is not acidic or corrosive, but simply a moisture curing type.

 

[iso9001]

Yea it is. They call it somthing funny like Hyrdoocscopic or somthing like that.

zevon8: Have you used it for potting ?

 

[zevon8]

Not specifically, but some of window or glazing urethanes have chemicals in them to speed curing. Think of RTV silicone, some has acetic acid ( like vinegar ) in it to promote curing.

If the compound is room temperature / humidity curing, and does not absorb more moisture once cured, it should be fine.

The real test is trying it out. Empirical evidence is always best.

 

[iso9001]

I tested some dried window urethane and got crazy resistance readings. It appears to be very conductive.

I checked it out and apparently cars with aluminium bodies and electrical contacts in the glass have to use a special non-conductive urethane. Well how about that :?