I wonder if I am doing something wrong.

On Konstantin's 90 S4 I replaced the AC clutch due to a bad coil. I got a re-manufactured clutch directly from Denso. It was a work of art, but within 100 miles it seized.

So, then I got a used clutch from 928 Intl. I'm not sure it ever worked right. We had problems with it immediately but they were strange. Sometimes it would work, other times it would not come on and I would find no voltage coming from the headunit. Then it would work fine for while. Finally the AC blower fuse started blowing immediately on pushing the AC button.

Yesterday I got back to work on it, and the blower no longer blew fuses. It only ran on high speed/defrost only. I swapped its relay and the fan worked normally again. However, I found no voltage coming from pin 6 on the AC headunit to the compressor. This is the signal that passes through the freeze switch, then the low pressure switch and then to the compressor clutch coil. So, I swapped in another headunit, which showed 12V on pin 6, but that began smoking a minute or so after the AC button was pushed - the relay inside melted.

I eventually found that the AC clutch was shorted to ground.

Here is what I found on disassembling the clutch.

1. The epoxy that encases the coil is melted and spread around throughout the clutch parts.
2. The coil looks shaved down. AFAIK, there is no adjustment for the spacing of the pulley over the coil. The coil is held in-place by a circlip and the pulley fits over it with the bearing resting against the circlip. IS THAT CORRECT? There are shims to adjust the air space between the outer clutch and the pulley, but nothing for the pulley to coil clearance.
3. THE BEARING IS SEIZED, at least partially. Funny thing is the pulley was still able to freewheel on the car. The first Denso was seized solid and burned the belt.

The first failure was similar but the coil was bonded to the pulley by the melted epoxy.

Is this all just from bearing failure? The coil is held in-place by a big circlip. The pulley with bearing is slipped over it with no spacers. Then the outer clutch plates slips on spaced by shims to a little less than a mm airspace. The Denso had a new bearing. Could this be from something wrong causing the coil to heat up and melt and that causing the bearing to seize? I have another Denso clutch on order and I'd like it to last more than 100 miles.

 

Bill,
the air gap spec is .4mm - .7mm (IIRC, check WSM). If the air gap is over by .1mm the clutch plate will slip and cause heat buildup, smoking, melting, seizure.

That written, I've seen two melted/seized clutch bearing/pulley assemblies and in both cases the compressor was seized. Are you sure the compressor is OK?

Also, when I am in doubt as to the health of an A/C clutch, compressor I power the clutch from the pressure switch connector to avoid the possibility of melting the HVAC Head. I will also check clutch current draw from there. All the working systems (~5) I've tested had <3.2A draw on the clutch when engaging.

 

On another note: this is the same car as the "Window Car"? Right? Did you ever pull the fuse panel?

I ask this because of your observations of weird HVAC head and blower behavior. I had similar issues with my '89; it has a giant pile of melted slag for an A/C clutch, pulley, compressor. Most of the blower/HVAC head issues were due to "melty-ness" at the fuse panel. In one case there was a behind-the-fuse-panel wire that was showing copper (and black burn marks) and there were several cases of melted insulation on adjacent harness wires just before the "A-Z" connectors to the panel. I replaced the various sections of wires and the HVAC Head/Blower started working normally.

 

Dave:

Thanks for the excellent ideas. I haven't pulled the CE panel, but the screws are out and I'll do that next. I will be careful about the air gap and I will check the compressor draw (err, wait a minute...the compressor is mechanical and has no electrical draw, and I'm not sure how a seized compressor would affect coil draw, other than heating things up), and other evidence of seizure.

THANKS!

 

Bill, something else you might try to check for slippage: make a number of distinctive marks on both the pulley and the clutch -- can be masking tape with different colored pen on each piece, etc. Then shine a timing light on it while running. Usually a coil wire works best to trigger the timing light, because this way you get lots of strobes. Then run the car with the AC clutch activated. The results will not be as clear as watching the timing mark on the damper because the AC pulley does not sync to anything, but you should be able to get a sense for whether the marks are moving relative to one another. I've done this on a couple of American V8s and the revlimiter's Saturn. It works better on a V8 with a single coil due to more strobes per rev, but I was able to confirm solid clutch engagement on the Saturn -- that car's two strobes per rev is more comparable to what you'll see on the 928.

 

You should be able to turn the shaft of the compressor by hand. With the clutch and pulley off you should be able to turn the shaft with your fingers. (Depending upon finger strength.)

 

Excellent idea!!

 

The only things that could cause a new clutch to look like that after a short time is:

- Incorrect clutch gap (too small and it drags, too much and it slips when engaged)
- Compressor too difficult to turn causing enough stress on the clutch to make it slip while engaged
- Defective new clutch

Dan
'91 928GT S/C 475hp/460lb.ft

 

Bill - I see nothing electrical that could possibly cause this. When the clutch is engaged it sees full system voltage - it can't ever see more than this - so unless your alternator is generating high voltage output - no electrical failure can cause the clutch to burn out.

I'd suspect mechanical issues - clutch slippage, compressor over pressure?, pulley misalignment etc... Particularly as you've had several - look at the systematic possibilities.

When you fix the head unit relay - here is an idea - install a replacement small relay on the HVAC board - direct replacement. Then take the clutch feed and run it to a new easily replaced and much heavier duty standard SPST relay (..53) on the CE panel (just put it in series). You will have spare relay locations on CE - every year does - just use one of those. If you like PM me an address and I'll send you the exact terminals you will need. This is the best possible fix for the head unit relay weakness - its the one Porsche should have done (minus the small relay).

Alan

 

Bill-


I cooked a 10PA20C clutch pretty quickly after my compressor 'rebuild' (shaft seal and gaskets) in May '08 - The clutch was 10 years old (at least) but it looked fine when I rebulilt the compressor. I thought I got the air gap correct (measured at 0.508 IIRC) but I guess I was wrong. The assembly went together exactly as you describe, with 2 small shims between the outer clutch face and the pulley. The A/C worked for a few weeks then went warm. After 1 month, maybe 300 miles of driving, there were all sorts of nasty insulation curlies sticking out from under the outer clutch face before I shot this pic. The compressor itself still turned fine with finger pressure. I gave up and installed a 928Intl re-man compressor, all is well with it.

EDIT: Come to think of it, maybe it wasn't the bad air gap at all- When I looked at the system in June, I actually had a broken front A/C compressor bracket mount, perhaps the (badly) misaligned pulley that resulted ground off the face of the clutch......

 

Rob: That's looks like what happened here. I've not had any problems with the 6E171 compressors I have rebuilt, but this 10PA20C has been nightmare. I didn't rebuild it but just replaced the bad clutch twice. Like yours, it spins free and seemed to work fine for the short while it did and I can't see anything that I did incorrectly. I don't see a rebuilt for sale with a clutch from 928Intl.

Dave A, Dan and Alan: I appreciate your perspectives. Alan, I understand how the source of the problem cannot be electrical. It could be heat distorting the coil from a partially seized bearing or some other mechancial issues, but it's not clear what. The pulleys are perfectly aligned and the air gap is correct.

After 2 clutches burning out (well 3 if you count the original one), I probably should install a properly rebuilt compressor with clutch.

 

Actually, I recently replaced the clutch on my wife's 968 (10P15E). It is very similar to the 10PA20C. The clutch failed after 15 years and it was because the rubber section separated from the outer metal ring. I replaced just the clutch with a new unit and it has been fine. I set the air gap to two washers greater than touching. I always installed clutches this way regardless of what distance is specified. I'm not sure what is up with these 10PA20C's though, never done one yet...

Dan
'91 928GT S/C 475hp/460lb.ft

 

Bill,
If it was an automatic you would not be having clutch problems 8>) Sorry wrong thread.
Roger

 

Good one, Roger.

 

Roger: