how resistent are TO-220 cases to moisture

[Simon]

Say I have a TO-220 mosfet mounted to the side of an aluminium case that has moist air flowing through it. Will this harm the one TO-220 mosfet at all ? I'd rather keep it out of the control box so that the control box can be plastic and the heat from the drive mosfet not be a problem.

I could I suppose have a plastic cover over the mosfet with a gasket, do such covers exist off the shelf ?

[Rerouter]

likely being the hottest part in the box it should be fine, moisture only forms and settles on the coldest parts of a given air volume, though there is a little bit of concern if this thing is pissing away a few hundred watt, as then it could instantly vapourise any moisture inside it and potentially crack the casing,

in general anything with a deviation in temperature faster than 2 degrees per second is where you risk this,

[Simon]

it's for an air handler and I think it's all on the inlet, most moisture is formed around the air con evaporator.

[Psi]

You could probably use potting compound to seal it.
Would be easy if you could mount it in a depressed area, then you could just fill it up with compound and let it set.

[Simon]

well we have to copy the existing case. I was thinking to heat shrink the leads. It's on the air in so not a lot of moisture as that develops in the aircon bit that comes after, so just a splash shield if anything.

[ptricks]

Heat shrinking the leads could make it worse, moisture would be trapped against the leads when the circuit is cool and not be able to get out when the part heats because the heat shrink will expand becoming tighter when the part is in operation.

If you don't have different metals in contact with each other no corrosion is likely to happen.
http://en.wikipedia.org/wiki/Galvanic_corrosion

One thing you can use for some added protection without potting is clear coat sealant. You see it a lot on boards that are put into appliances like washers or refrigerators. It is a clear spray on coating that makes the board almost water proof. I don't know the name of it but maybe someone else does.

[Psi]

You can get some kinds of heatshrink which have a glue inside, when they shrink the glue activates and sets making the heatshrink join waterproof.
Maybe that could work.

Though mounting it to a case when its covered in heatshrink might be an issue

[Baliszoft]

conformal coating?

[Simon]

You can get some kinds of heatshrink which have a glue inside, when they shrink the glue activates and sets making the heatshrink join waterproof.
Maybe that could work.

Though mounting it to a case when its covered in heatshrink might be an issue

I just cant to heat shrink the leads so that water droplets can't short pins

[Jeff1946]

Easy quick coating -- clear fingernail polish.

[Simon]

this is for a commercial application

[SeanB]

BMW did this with the fan controls, and they fail with depressing regularity, either cooking to death because the fin is too small or corroding the semiconductors off.  Use a set of fins that will handle the dissipation in free air, and use a conformal silicone ( electronics grade so no acetoxy cure) to blob over the active device, its leads and the mounting tab so as to keep liquid water out. Vapour will still penetrate, but will take a long time to corrode it to pieces, and the heat of operation should keep it out. You will have to cover the entire device, not just the pins, but the whole body and tab. Whoever has to repair it will have to dig it out and curse you, otherwise it is not done right. I will recommend not using a silicone grease thermal compound, but to use a silpad or other premade heat transfer and insulating method. If you use an extruded heatsink then fill the channel totally after wiring and test, but before it is used.

[Simon]

I'm going to try and get it in the main case using it as a heat sink. Otherwise are there standard covers that can be bolted over transistors mounted on panels that I could put a gasket around the rim of.