Good coating for condensation protection?

Most air-dry silicone caulk has acetic acid, which is bad news for electronics.

Epoxy is good, or an acrylic casting compound. Use slow-set epoxy... the fast stuff gets hot as it cures. The resin they use to fiberglass boats is good and cheap. Mold-making (2-part) silicone is excellent.

For conformal coating, ordinary polyurethane varnish is pretty good.

John

I'd imagine whatever you seal things up with is going to effect the temperature response of the unit. Will that be a problem?

I would not rely on conformal coating on the circuit board. It is a good vibration damper and protects against, at best, light levels of dew formation. If you expect to have water condensation than I would recommend epoxy. But it will depend on if your sensor operates mostly in a narrow range of temperatures or across a broader range such as below freezing to 90 deg. F.

filling

might be

In my experience, it takes many,. many coats of spray to make a decent thick coating on a PCB, and each coat has to dry, making it take a long time. It's better to dip coat the part.

"John Larkin" wrote in message news: snipped-for-privacy@4ax.com...

Thanks John,

Epoxy seems to be the favourite. Although I imagine repairs would be a bitch.

How do you find the drying time for a polyurethane dip? I think it would take a while to harden? Do you favour any particular brand?

JeB,

The sensor is not on the board, it's attached to a short (30") cable. The sensor actually measures water temperature, hence the serious condensation. I used connector coating to waterproof the sensor, which seems to slow down the temp response a bit, but not too bad.

B. Stock wrote: "I've got a temperature sensor circuit which is subject to severe condensation. What's a good protective coating? I was thinking of filling the enclosure with silicone caulk, but I imagine some spray coating might be better?" ____________________________________ Re; I've used a product called "Plasticoat" which is sold in many hardware stores. I've dipped whole circuit boards in it. It's non-conductive, water proof, acid proof, and tough, albeit probably not flame-proof. It, and any other coating, will affect the response time constant of any temperature sensor and increase any self-heating effect/error. Additionally, you'll need to derate the power dissipation of any coated, heat dissipating components. Repairs are made difficult, if not impossible.

-Dan Akers

With my deep experience with silicone caulk (including leaky showers ;-) I'd say that the problem with silicone is that it doesn't "wet" many surfaces. If there's any motion or vibration, I'd be afraid that it would develop a small gap at the surface interface and draw up water via a capillary effect. I'd go with marine epoxy, myself.

One-off project or 100K/month commerial or ...?

be

How about using some conformal coating\lacquer made specifically for the type of task you describe?

Is that still effective if surface mount parts are used instead of dips?

Yes, but the non condensing atmosphere is outside, so then I've got snow/ice/critters to worry about. The sensor measures the temperature of an ornamental pond (coverered) and controls a heater to prevent fishsicles.

"Bill Stock" wrote

Is it possible to have just the sensor in the condensing atmosphere?

If not, then you will need a small 'Hoffman Enclosure' with a NEMA (I forget the #) rating.

If you use epoxy you may want to use marine epoxy, that's the stuff they make boat hulls out of, available at boat shops. Also called 'West Epoxy'.

Thanks Art,

I think someone else suggested wax too. Yes, the sensor (LM34) is at the end of a cable, covered in "connector coating". Seems to be similar to Plastidip. The sensor is in the water and the board/case is attached to the bottom of the pond cover's escape hatch. Water temp is around 40°F and the outside air (above cover) can be as low as -40°F. Although we haven't seen that yet.

I removed the circuit (board/case/sesnor) this morning. The solder side of the board, which rests against the case lid, was dripping in water. The component side seems almost completely dry. Once I dried the circuit out again, it started to work fine. On the bright side, all that water emulsified the flux residue. Making my coating job much easier.

I bought some Acrylic spray for my coating, they didn't have the dip. I may also put some Silicone in the case lid to be safe.

PVB is inside a case. Now conformal coated.

I'm using a titanium aquarium heater to do the heating. But it's thermostat starts at 68°F. My goal is to keep the bottom of the pond around 39°F (max density of water). This should prevent too much convection and be much cheaper to heat. These heaters aren't designed for prolonged use, they will burn out quickly. Right now it only runs about 10 minutes per day, depending on outside temp and snow coverage (insulation). The sensor is also hooked into the house's automation system, so I can monitor any problems.

"Bill Stock" wrote

There were completely submersible fish tank heaters - two test-tubes IIRC, one with a heater and one with a fixed thermostat, wired together and to a line cord. The test-tubes were sealed at the top with some green gunk. Heating element was packed in sand inside the test-tube.

Might this sort of construction be applicable?