Removal of Power Transistor Rivets

I do not know how to remove the transistor from its original position. I avoided doing so by abandoning the dead transistor and its grossly undersize heat sink and instead bought a replacement heatsink + transistor fo about $100, which provided me with the needed wiring harness to the transistor. I discarded that heatsink and mounted the transistor on a reasonable one mounted elsewhere and easy to access. This is based on my experience and belief the original heatsink and mounting location are almost guaranteed to fail repeatedly.

 

Do you have the AC thumbwheel vent assembly removed? The hinged panel below it allows access to push up the clips and withdraw the AC thumbwheel vent assembly out the front.
Image Unavailable, Please Login
Image Unavailable, Please Login

Sounds like you are trying to do heart surgery by going up the patients rectum.

 

It's "Style 1" in the drawing on page 6 of 7 of the pdf (the page itself is labeled "1" at the bottom). The outer metal case is the collector. The pattern of the leads is asymmetric so just hook up the new transistor same as your old transistor.

 

Not sure if relevant at this point but I have used a similar flexible drill adaptor with the hex drill bits (to remove the rivets on the bonnet slats on my BBi) and, although it did work out, it was far from a precision job. You want you to know the drill bits wiggle a lot since they are only held in with a magnet and not really tight, and the flex shaft is multi strand wire (like a speedo cable) that works very unpredictable, specially in reverse as it is wound up in the opposite direction. So the chances of breaking something are high with this drill setup.

 

There are a couple issues here. For the unit to be a fire hazard would exceed the transistor’s heat tolerance (i.e. it would self-destruct and either fuse into a short circuit which would simply turn the fan all the way on, or open completely which is the usual failure mode). However, with an inadequate heat sink like the existing one I am pretty sure it would get to plastic-melting, leather-discoloring or passenger-startling temperature. That all depends on the fan control-knob position. IIRC max power dissipation is right before it clicks into the max-speed position.

Also, with power transistors like this there is always the choice of whether to electrically isolate the heatsink from the transistor case which reduces the chances of the heatsink touching some other circuit with harmful results. The downside is that the isolation reduces the effectiveness of the heatsink somewhat. So in putting it on some new heatsink and mounting it elsewhere you need to account for potential unwanted contact. The case is connected to the wire to the fan which through the fan leads to +12V. My recollection is the fan motor at full speed conducts 12-15 amps so if that touches ground it will at least produce a healthy spark.

There is space forward of the passenger footwell closing plate (big piece of maybe .060” thick steel.). If I wanted to hide it away I would put it forward of that plate but would also mount a small fan on or near the new heatsink to guarantee air circulation, which unfortunately also requires a +12V wire to run the fan. (My current rig is too big for that so at the moment it is loose on the floor of the passenger footwell. No concours points for me. Someday I will mount it properly perhaps under or behind the passenger seat.)

Another option would be to mount it, isolated, to a large thick sheet of aluminum which itself is mounted to the forward side of the above steel plate. If the aluminum were thick enough (say, 1/4”) I think the heat would be distributed over its surface enough not to unduly heat the overlaying carpet.

And BTW regarding your polarity question, yes it matters very much which wire goes to which pin on the transistor so to be safe I recommend taking a clear picture of how it’s currently wired and then being careful to always reproduce that exactly.

 

Looks like a good solution. I’d show you mine, which you would laugh at, but am traveling for the next couple weeks,

 

Good solution, I had this same problem when I bought my first TR back in 2007. My solution was to replace the rivets with a stud and nuts arrangement on a modified heat sink to facilitate easy removal to repair if needed. This is a very clean solution. I always wondered that the original location was a little crazy, as heat is what kills the transistors, and yet when you turn on the heat in the car, you are bathing the transistor in heat. seems counterintuitive?

 

The hottest air coming out of the center air vent is still cooler than the transistor outer case temperature (so it still provides some cooling), but agree that it's just under-designed (as we've had so many failures). One contributing factor IMO is a psychological one where a human will usually set a variable speed control to 3/4 or 7/8 thinking it's being less harmful than selecting full speed, but it's just opposite in this situation (i.e., using 3/4~7/8 speed cooks the transistor most; whereas, using full speed doesn't stress the transistor at all).

 

To get the rivets out I used a 90 degree driver head with a counter sink hex bit. I did it from underneath and got it out. Worst thing was getting my body under the dash.