Blower Final Stage (sword) multiple times defective 30 secs after repair

**SatchNL** · 03-29-2019, 02:32 PM

Hi Guys,

My dad has a '91 850iA here in Holland and we have the well known problems with the blower final stage/sword. I bought new mosfets (BUZ11 instead of the standard ones) and tried to fix it myself. I connected the sword to the car again, but didn't put it in the place where it normally sits because I wanted to see if it wouldn't smoke. At first it worked fine, but after about 30 secs I heard a small noise/smell coming from it and the sword went faulty again. The fans won't turn off anymore, it's on max or a little bit less. I think one or multiple mosfets shorted right? I know that the fans need to blow over it for cooling, but the cooling body would be big enough to use it for a short while without I assume?

After this I took the sword with new replacement mosfets of the same type to a friend who really knows what he is doing with a solder iron and the same thing happened.

I tried to search online but couldn't find a clear answer what could cause this. We have another sword which we didn't try to fix yet that is only working on max and can be turned off (which means one or more broken mosfets?). So the car is driveable with this one.

Is it possible that the blowermotor is not working properly and therefore puts a lot more stress on the sword which makes it fail after such a short amount of time?
Is it possible that I bought 'fake' mosfets who fail super easy ?
Any other ideas to test/fix?

**Timm** · 03-29-2019, 04:26 PM

Yes, it is possible that the blower motor is demanding more current, but more likely that there is something else wrong with the Sword - possibly something that means all the current is going through one FET rather than shared between them all. There are many things that can go wrong - but most likely the failure of R13 - R15 as seen on the diagram below:

http://www.meeknet.co.uk/e32/BMW_E32...Unit_Sword.pdf

**SatchNL** · 03-29-2019, 08:10 PM

Thanks for the quick reply and the clear diagram! We have checked the resistors you mentioned, they seem to be oke. They measure the right amount of resistance. However one of the mosfets shorted again. We have replaced the mosfet and now put a 30 amp fuse on the board. If that fuse blows we know that the blower is drawing more current then the 25A it should.

**shogun** · 03-29-2019, 08:16 PM

To install a larger fuse is not a good idea. Better you use an ampmeter.

**SatchNL** · 03-30-2019, 04:26 AM

It's an extra fuse on the final stage board, haven't touched any of the normal fuses in the car. I hope this way we can test the sword without blowing up anymore mosfets because the fuse should break before the mosfets. Will try to find an ampmeter too, thanks!

**Timm** · 03-30-2019, 04:40 PM

Unfortunately a standard automotive fuse will take twice the rated current for 30 seconds before it pops - whereas the BUZ11 will give out a lot quicker (in the order of 1/10th of a second). Are you using heatsink compound when they are fitted? Or even a heatsink pad (insulating)?

**SatchNL** · 03-30-2019, 05:23 PM

Thank you for thinking with me! That makes me rethink my plan with the automotive fuse.

Here are some photo's of the current state of the sword. Plastic spacers to push the mosfets against the metal and a heatsinkpad. A small bolt and nut tightens it.

Here you can see the 30A fuse we soldered in.

This is the mosfet that broke with the last testing. The plastic spacer to push the mosfet down melted onto the mosfet apparently. My friend tested the circuit with a multimeter and this mosfet shorted.
Tested it without any air flowing over it hanging in the footwell for less then 30 sec.

**Timm** · 03-31-2019, 03:17 PM

The body of the FET is sitting above the heatsink pad - although the tab is squashed down it looks like the legs are too short and that is lifting the bodies. As it is difficult to get the FETs flat, I would use heatsink compound as well.

**shogun** · 03-31-2019, 07:06 PM

here some more pics which might help http://www.bimmerboard.com/forums/posts/355700/

**SatchNL** · 04-20-2019, 05:32 AM

Thanks for the help/tips guys. We have replaced the mosfets with even heavier ones:
https://www.reichelt.nl/power-mosfet...0z-p90339.html

We also made sure that they made good contact to the metal heatsink and put some thermal paste in between. It was a bit more work because the mosfets are bigger, but now it works perfect again.