Meltability of car power window switches?

[lpc32]

How likely are such switches to melt when de/resoldering?
And any idea what's that general type called?



Bad board design. To get soldering access to two of them you need to desolder a 13-pin connector:

[Stray Electron]

  If you use a temperature controlled iron and a solder with a relatively low melting point it should be ok.  Do NOT use ROHS solder that has a high melting point.  I use lead/tin solder with 2% silver made by Kester. IIRC the MP was 398 degrees and it instantly liquified with no slush range. Good luck finding that stuff any longer!  Don't use Chinese or Brand-X solder unless you're willing to throw the item away after you burn it up trying to get their junk solder to stick!

  The switches are probably custom made for that car manufacturer and it's very likely that they are used only in that one model. Car manufacturers are fond of doing things like that and they build/buy in such high volumes that there is no cost savings in using standard parts. To the contrary, using part made very specifically for their application usually saves money and assembly time so is cheaper in the long run.  Also maintainability and repairability simply don't matter to them.

   I had one car that I had a lot of trouble with the window switch contacts tarnishing and not making contact. I had given some thought to buying some gold leaf and trying to burnish that onto the contact surfaces.  But in the end, I bought a used part from E-bay and used parts from it to fix mine.

[lpc32]

The iron and solder are fine (why use lead-free at home), and this is just a double-sided board, but I do wonder if the plastic parts are meant to withstand more than the initial soldering.

The top of the switches reads DECO but I wasn't able to find much.

Yeah, eBay could be a fallback. But for now I can switch switches around, and there's also the extra reserve in the passenger doors.

[sokoloff]

Don't overlook just buying the part from a junkyard. Unless it's a very uncommon car (or a first year of a new model), chances are good lots of them have made it to the junkyard^Wauto recycling center.

[lpc32]

I don't dismiss the option, but I'd rather fix what's already here.

I did found a few on eBay that would probably work. They're missing one connector pin, though it's one with no purpose I could figure out.

[schmitt trigger]

This board is a piece of cake compared to what I've seen. A similar construction with hard epoxy potting.

The reason, from the manufacturer's point of view, is because repairability IS NEVER a design objective. If the tinyest of components fails, they expect you to replace not only the board, but the complete module.

I know, I know...it sucks big time.
But this is the reality with most, if not all automotive electronic modules nowadays.

[lpc32]

Turns out I'm unable to desolder it.

Two leads on each side of the switch makes it difficult to pull the switch out one side at a time.
Even if it was one lead per side, the switch sits flush on the PCB so not much lead flexiblity.

Cleaning up holes with the leads still inside doesn't look very doable either, especially if the plan is to reuse the switches.

[soldar]

Generally I manage to work them out with patience. Use a "sucker pump" to try to get out the solder. If you are very luck you might free a terminal that way. Then insert tiny screwdriver under the part and move it out ever so slightly. This way you can gradually work your way around the four terminals moving each one out a bit at a time.

Depending on how invested you are in this repair another option might be to just build a new PCB but it might be easier to just buy a replacement.

[ebastler]

Turns out I'm unable to desolder it.

Two leads on each side of the switch makes it difficult to pull the switch out one side at a time.
Even if it was one lead per side, the switch sits flush on the PCB so not much lead flexiblity.

Cleaning up holes with the leads still inside doesn't look very doable either, especially if the plan is to reuse the switches.

Try to find someone with a proper de-soldering station. (I.e. a motorized suction pump with a heated tip.) Especially when you are worried about melting the plastics, I would rather avoid extended manual fiddling with soldering iron, solder sucker, and wick.

Maybe there is a local hacker space or amateur radio club which has desoldering equipment? Or, better yet, use this project as an excuse to buy your own! 

[lpc32]

work them out with patience.

Works for you also when there's plastic involved?

Use a "sucker pump" to try to get out the solder.

That's mainly what I tried. The result was mostly sucking out the extra solder I added for heat transfer.
I even wasted a bit of my precious Chemtronics wick, to no avail (too bad the fake stuff from eBay does absolutely nothing, at least for me).

just build a new PCB

Not so easy at all, and then, it's back to the core problem of no switches available.

someone with a proper de-soldering station. (I.e. a motorized suction pump with a heated tip.) ... local hacker space or amateur radio club

Could indeed be handy, but there's nothing nearby I know of.

Or, better yet, use this project as an excuse to buy your own! 

It's going to be a very poor excuse. If to clutch at straws, the more dignified thing to do would be to just buy the gear without excuses.

[soldar]

Works for you also when there's plastic involved?

There are many different kinds of plastics. Some will melt if you leave them in the sun and some will stand temperatures higher than the melting point of solder. Some are thermostable, meaning they will not melt again with heat. So the first thing I would do is touch the hot point to somewhere where it will not cause damage and get an idea of what kind of plastic we're dealing with.

If the plastic melts easily then you need to be extra careful. Apply heat with an extra hot tip for just a split second so the solder at the terminal can melt but the plastic does not have time to heat up. You should be applying pressure on a tiny screwdriver wedged in place so that as soon as the solder melts the screwdriver moves in a bit. Then you move to the next terminal. You need to wait for each terminal to cool down before you return to it but if you are working on four this should not be a problem as by the time you return to one plenty of time will have passed.

If you can apply heat to two at the same time then it becomes exponentially easier because you can work out two terminals at the same time.

But test the plastic first. It could well be you are dealing with a hard thermostable plastic. You can poke it with a tiny screwdriver and see how soft it is. Soft plastics tend to melt easily. Hard plastics tend to be more stable. My guess is they would not use a very soft plastic inside the door of a car where it can get very hot.

I find the main thing is to go very slowly and patiently. Losing patience and rushing are recipes for disaster.

Good luck!

[lpc32]

You should be applying pressure on a tiny screwdriver wedged in place so that as soon as the solder melts the screwdriver moves in a bit.

Thanks for the extra detail. I might try that when the stakes are low, after I get the replacement from eBay.