PDA

View Full Version : Compressor clutch problem



Colin Stamp
03-09-2007, 12:02 AM
Hi all,

Complete refrigeration newbie, I'm afraid...

The compressor clutch in my car sometimes fails to engage when the under-bonnet temp starts to get high. I've verified that it pulls normal current whilst failing.
The service manual tells me the compressor is a Sanden TRS105R 3211 (The car is a 2001 Saab 9-3 Aero).
Unfortunately, the service manual has no info on adjusting the clutch.
Does anyone know if the clutch on this compressor can be adjusted for wear, and if so, what special tools I might need to do it?

Cheers,

Colin.

Dr._Fleck
03-09-2007, 12:22 AM
Do you mean it pulls normal current whilst working? I'm not an expert in car A/C but the clutch will have high pressure/low pressure switches that control it. If things are getting hot the pressure could be too high.

Having said that, usually the clutch knocks off on LP because there is low refrigerant level.

Check the switches anyway

Colin Stamp
03-09-2007, 08:52 AM
Thanks for the reply.
The clutch pulls normal current when both when working and when failed, so it isn't a control system or electrial problem.
I was under the impression that the clutch must be faulty if it's energized, but fails to engage?

Cheers,

Colin.

Dr._Fleck
03-09-2007, 10:14 AM
Somebody will be along soon with more knowledge than me;)

Electrocoolman
04-09-2007, 09:49 AM
Colin, .....a few thoughts regarding your problem...

Does the clutch / compressor have its own drive belt or does it also share the drive belt with other devices (power steering, alternator, etc.)?
Is it slipping under excessive load which is worse when belt gets hot?

You indicate that power (12v) is being applied to clutch (and clutch is drawing current) all the time but the compressor is not always turning / clutch engaging.... [The normal signs that the drive is engaging being that the bolts on the end of the drive spin....] This would tend to point away from the pressure switch(es) (which would remove the 12v) and towards the mechanics of the clutch itself .....which I believe is the way you are thinking yourself.

As temperatures (and hence pressures) increase, the load on the compressor drive increases and this might be causing the clutch to slip.

As to if the clutch can be adjusted..(in situ?)..I don't honestly know,.... but would imagine that it might be a compressor off engine job, in which case it would possibly be worth considering a new clutch assy. (clutch contaminated with oil?)

Others with experience of these compressors should be able to advise better.

Regards, ECM.

Colin Stamp
04-09-2007, 11:40 AM
Thanks for the reply.

There's only one drive belt on the car and it doesn't seem to be slipping. When failed, the clutch pulley turns normally, but the driven plate is completely stationary. In fact, the air gap between the driven plate and the pulley doesn't even close up.

From trawling around on the web, this seems to be the normal failure mode for compressor clutches - wear makes the air-gap too large so the driven plate is too far away from the magnet, which then can't overcome the return springs at that distance. High temperatures apparently make it worse by reducing the power of the magnet.

I gather that some clutches have shims to make the air-gap adjustable, and some can be adjusted by other means, often requiring special tools.
I've ordered a set of "universal" compressor clutch tools from eBay, so I'll have a go with them when they arrive.

Cheers,

Colin.

paul_h
04-09-2007, 06:21 PM
You need to verify that there is in fact 12v to the clutch when it fails. If there is, then the clutch is faulty. If there's not then you have a bad electrical connection or the HP switch is turning the system off. Check the voltage rather than the current if it's not pulling the clutch in.
I'm not an automotive A/C specialist, I've just done a few, plus my own old cars and some aircraft. And while clutch failures probably do happen, they're pretty rare in my experience, even with 15 year old systems such as the one's I've worked on, the compressor normally goes first.

Colin Stamp
04-09-2007, 08:06 PM
Interesting point...
The voltage right at the clutch is one thing I haven't measured yet, mainly because it's a bit of a pain to do whilst leaving the clutch connected, and it's very difficult to get the system to fail with the car parked and the bonnet open.
Wat I did instead, was to use the compressor fuse as a makeshift current-shunt. I get 15mV across the fuse both when the compressor is running properly, and when the system is failed, so the current is the same in both cases. The clutch is the only thing served by that fuse. All the sensors are out of the picture since they would prevent the compressor relay pulling in.
I guess there could be a very consistent, high resistance connection in the between the relay and the clutch, but my money is still on the clutch needing adjustment. I'll find out when I actually set out to fix it and I've gained enough access to measure the voltage at the clutch. If all is OK with the volts, I'll wade in with the clutch tools.

Cheers,

Colin.

clivemtk
04-09-2007, 09:44 PM
if the clutch is a magnetic coil and you get 12volts either side your fuse closer to the clutch coil will be a butt connection otherwise it sounds like mag coil breaking down as unit warms up

Colin Stamp
09-09-2007, 03:04 PM
Well, the clutch is now re-shimmed, or rather de-shimmed - there was only one shim. The job turned out to be slightly more interesting than I expected.

It started out annoyingly. The only access is from under the wheel-arch, so loads of plastic liner bits needed to come off, with the usual broken clips etc. After that, I couldn't work out a way of preventing the driven plate from spinning so I could undo the centre nut.

Then I had a brain wave. I bypassed the compressor relay so the clutch pulled in. Then I could stop it turning at the crankshaft pulley bolt. Soon it was all back together, minus the shim.
Encouraged by my self-proclaimed stroke of genius, I had a stupidly bad idea. The air vent temperature has always been a bit disappointing at 5-7 degrees, so why not fire it up with the compressor relay still bypassed and see how cold I could get it? What could possibly go wrong?

Well, I started it up and sat in the car looking at the thermometer, not noticing that my compressor bypass had also prevented the engine cooling fan from running. I now had no condenser cooling at-all and a compressor that would stop for absolutely nothing.:eek: I was just wondering why the vent temperature was still disappointing when all hell broke loose under the bonnet. It sounded like a fire extinguisher on speed. So, the emergency over-pressure vent valve works fine then. :(

50 quids-worth of re-charge later and it all seems to be working nicely, though I need some hot weather to test it properly - that'll be next "summer" then.

Thanks to all who helped.

Colin.

kpntreal
09-09-2007, 03:23 PM
Colin
You must measure the voltage at the clutch or as close as possible to it, after any switches. Some power steering cars use a pressure switch on the power steering pump discharge, that disables the AC compressor while making turns. It keeps the engine from overloading and perhaps stalling. This switch needs just a bit of grease, oil, and road grime to foul its contacts.

On some vehicles, misdiagnosis of the clutch is common, as the switching is done thru the EARTH connection, and not the HOT connection.

The error is made when the technician measures the EARTH for the clutch at any convenient earth connection on the chassis.

Could it be that on this vehicle the switching is actually done through earth/ground?

Regards

kpntreal

Colin Stamp
09-09-2007, 10:40 PM
I'm afraid the voltage test isn't going to happen. The only place I can measure the voltage to get a worthwhile result is right at the clutch coil (Anywhere else will give me no more info than I can already glean from the current readings). That point is very inaccesible and the system generally can't be made to fail with the bonnet open anyway.

There are no switches on this car which can cut out the compressor if the compressor relay is pulled in. All of the things which need to turn off the compressor do it via that relay.

What I have measured though, is the current flowing through the coil. It's just over 3A both when the system is working and when it's failing. I don't have the spec for this particular clutch, but it's in the right ballpark.

Other than a clutch problem, the only possible cause would be a high resistance connection between the relay and the clutch. It may yet turn out to be that, but my money is still on the clutch. Time will tell...

Cheers,

Colin.