hwyengr

Where's the best place to start for an compressor clutch that's kicking on, apparently randomly, even when the A/C button is off and the slider isn't in defrost? Pushing the A/C button does not engage the clutch full-time, either.

Is my relay stuck on, and the freeze switch or low-pressure switch is cycling the compressor on and off? I'm leaning away from this diagnosis, though, because the compressor doesn't engage immediately on startup from cold, but rather 5-10 minutes later.

FredR

Clearly you have an issue with the control circuits just a question of whether it is the relay, the switches, the wiring or a combination thereof.

I have seen the relay kick in/out spuriously when switched on but not the other way round.

Regards

griffiths

Jeremy,

I'd suggest its the controller board or faulty
AC switch.
Until another Shark owner with a like problem
gives you the diagnostic procedure, here is
some easy reading:
928 AC Tech Tips

The Patman

I hate intermittent problems. It's much easier to shoot something broken than something you need to "see" fail once in a while as it operates!!

Not gonna claim to know what your problem is, but maybe I can spur some brain activity in one direction or the other.


Sounds like you already know the system pretty well. And I assume you have wire diagrams.

First, and maybe easiest, (but not entirely likely, power would have to short to the line) is an external short in the power line that can be culled out quickly and easily by disconnecting the line at the evap switch (obviously, if it still comes on with the line cut at the evap switch, the issue is in the engine compartment). You can drive with it like this (make sure the power line cant short to the fuselage) but no defrost will be available of course.

Certainly there are a couple of things that could be happening, but in the end, power has to be delivered (as you already mentioned) through the evap and psi switches, and it always comes through the relay (except for a short to power in the signal circuit). Both of the first two are fairly simple devices, and easy to bypass / disable.

Anyway, Second, I think I'd try to pull the side panel on the center console, locate the 4 pin AC push button and unplug it (from the back of the panel, not the inline 4 pin connector). Not always the problem, but a fairly easy way to be sure the spring / contacts in the push button aren't sacked and engaging the relay through vibration. It won't hurt to drive around with it unplugged. (when it happens does the light in the AC button come on?) Disconnecting it at the back of the panel would eliminate the switch, since it's a random failure.
As I think of it though, this may require pulling the front trim piece and pulling the two screws and sliding the AC button panel out. I got to mine from the drivers side with curved hemostats, but plugging it back in is stupid without panel removal!

EDIT: Not sure, but I don't think the setting motor can call for the AC to actually come on. I think it just moves to the full cold flap position when the MAX AC button is hit by the slider. The max AC switch is inside the control head and once again, would be a little more difficult to shoot. But a head swap out could help ...but may disguise something else. Again, not sure the max AC switch can actually call the compressor to be on. The defrost switch can, but would call the fan to, that may be a clue for ya.

A failing (or randomly shorted through vibration) defrost or max AC switch may be difficult to shoot without pulling the HVAC control head, and of course a spare head to swap out would be nice, but swapping it wouldn't eliminate the defrost or max AC switch from the relay. (swapping in a spare may fix it but it could be a few things in there)

The defrost and max AC switches can be easily tested to verify continuity with a DVM, but finding a random short in them may be more difficult. I'll set it on the bench, hook up the DVM and tap the switch with a small plastic screwdriver to see if the DVM shows an open or short through vibration. Sometimes heat or cold need to be applied to a component to induce the random failure. Same with a diode. The only thing here is that the defrost switch is a simple switch and if the fan isn't coming on to the high position when the AC compressor comes on, then this may not be your issue, since the switch doesn't control the fan and the AC separately.
Does it call for the defrost fan and the AC compressor when commanded? Or is that intermittent too?



Jeesh, I talk too much. What I meant to say is

Try to eliminate a short in the engine bay just in case but it's easy, and a short in the push button second but a little more difficult, then control head (defrost switch (maybe but check for fan on high) max AC switch (not sure how it could), relay (probably), or something weird...shorted diode etc.)

hwyengr

Luckily the center console was already out for a vacuum actuator refresh. It was so sweet seeing the flaps work for the first time since getting the car. We can also remove the A/C pushbutton from the diagnostic, the problem happens regardless of whether it's connected or not.

Can someone walk through the circuit theory of the HVAC head? I assume the image below is the standard operating circuit for the compressor to be turned on from the defrost position, yes?



What I don't understand is all of the diodes that connect what seems to be ground bus to everything. For example, why doesn't turning on the lights engage the compressor, a la this scenario?



I'm guessing that my problem is a burned out diode somewhere, but I'd like to understand why all those diodes are there in the first place? Some make sense to me, like the one shown directly left of the A/C relay. But why does the ground bus connect via diodes to the hot side?

Also, on the wiring diagram, the magic blank box that's labeled "Control Unit" just below and left of the HVAC controller; what goes on in there?

hwyengr

Defrost seems to work normally (except for not always engaging the compressor), but the fans don't kick up to high when the compressor does its random cycle.

The Patman

Well...let's see. I'm kinda waiting for someone smarter than me (that would be most folks!) to come and say they know exactly what this is, but in the meantime, the reason I asked about the fan is because the defrost switch being broken would call for the AC and the fan at the same time. If it works normally when called for, but the fan doesn't come on during the random compressor cycle, that would eliminate the defrost circuit from your "random compressor cycling" issue...because it would call the fan too...which it doesn't.

that would leave the relay, a common failure (though almost always failing to "inoperative") and the AC push button, (which you indicate is now out of the circuit and the problem continues), or a control circuit fault (wires and or circuit board issues).

at this point, I think you may be right about the relay being stuck on, but not sure why it doesn't happen right away when the car is switched on, and I might put a test on the incoming side of the evap switch to see if it's "hot" all the time as you first thought. In other words the since it's not the push button, and isn't acting like the defrost circuit, the relay may be on as you first surmised, and the cycling being caused by low freon pressure. But now instead of just unplugging stuff, ya gotta move to step 2...get a meter. A quick check as I mentioned above to see if the power line is hot all the time would tell the tale...I think. Or at least tell ya where to look next.

As for your diode questions...these are strange and magical things and I try not to make them part of my life!

the top diagram is correct the way I understand it. though power seems to come from the opposite direction ( in on pin 1 of the 4 pin connector, and the AC signal back out on pin 2. by the way, an intermittent short between pin 1 and 2 would cause this issue as well...I'm just sayin')

In the bottom diagram, the drawing doesn't seem to be quite the way I would interpret it. In the parallel circuit of the AC switch light and the HVAC control head light, (in from pin 3 of the 4 pin connector) the lazy little electrons see the easiest path back to ground (they actually go the other way, but forget about that) and so as soon as the HVAC control head light is lit, the path is not through the diode, but directly back to the V1 connection point below. That particular diode seems to be there to block the power from the ground line when the relay is closed...and because they had too many diodes at the factory! Seriously, I asked myself the same question on that same diode. I probably would have saved a nickle and left that one out! I can only surmise it's for coil field dissipation when the ac clutch coil shuts down. Maybe. I dunno.

On the "control unit", that is what we all seem to refer to as the "setting motor". That's the box that sums the resistance of the two temp sensors and the slider, and uses that to modify a voltage sent to a pair of transistors that drive a motor to flip-flop a flap or vane between the AC evaporator and the heater core. this allows your cabin blower to blow through the heater core, or the AC evaporator, or some combination of both. When the motor drives all the way cold directing air only through the evap and not the heater core, it also sends a signal to shut off the hot water to the core...I dont have an internal diagram of that box, so I know little more than that. It knows when the max ac switch is on, but I dont think it has provisions to "turn" the ac on. This unit can use that signal to turn on (open and close) the recirculation flap driving cabin air (already cooled) back through the evaporator a second time...for max cooling. When the AC itself is on,(when the relay is closed) the signal for that recirculation flap seems to go through the second pair of relay contacts.

Alright, my brain is fried. I hope I got that right for ya.

Anyway, I'd do a quick power check on the outgoing line to see if the relay is stuck on.

The Patman

Wait one more thing. Did ya unplug the evap switch to see if it still has the issue? I ask because lots of folks have problems with the famous (infamous) "14 pin connector". In the engine bay. If there is an intermittent short in there over to pin 9 (i think) you could also see this issue without any signal from the relay at all...I'm just sayin'.

MFranke

If unit is out, take opportunity to do the $4 radio shack relay upgrade. See if it fixes prob

hwyengr

Tonight while poking around I got battery voltage at the freeze switch and could hear the clutch clicking when pulling the wire on and off. So for a starting point, I'm just going to replace the relay. If it's not bad now, it's probably just a matter of time until it is anyway.

The head unit is out, and there are couple of interesting things: it's not the original head unit (build date 8/87 on a 4/85 car) and all the plastic release tabs were broken off. Someone has been in here before.

If all the screws are there, the lack of plastic tabs isn't going to create a rattle, is it? I've been tracking a few down in the center console...

The Patman

cool. so you're indicating that the relay is on all the time.

You could determine whether something is turning it on, or if it's just welded shut, (since ya got the head unit out) by checking continuity across board pins 5 and 6, and pins 8 & 9. that would tell ya if it's closed all the time due to a bad relay (stuck closed even with no power applied).

you could get fancy and apply a 9 volt battery to pins 4(gnd) and 7( relay power) and listen for a click, as well as check for continuity on the two aforementioned pin sets while power is applied. All of this can be done CAREFULLY while out of the car on the bench.

strategy being that if relay is actually operating OK, you can still replace it if ya want...but the new one may do the same thing...that is, if the relay is OK and something else is turning it on.

(like Scarlett Johansson for instance)




The screws will hold it pretty damned tight.

hwyengr

Good news/bad news. The relay still clicks, but the throw between 5 and 6 is fused closed (continuity regardless of power to the relay). 8/9 works as it's supposed to. Not sure where the good news is in that, but it seems that nothing else on the board is fried out.

I picked up the relay from Dwayne's relay repair write up, but I've been thinking of doing the remote compressor relay in the CEB with a mini SPST on the board itself. A lot more complicated, but it would be nice to keep the compressor amperage out of the head unit.

But this brings us around to the original problem. The compressor should be running all the time, but it's still intermittent. The freeze switch seems okay, can the system pressure fluctuate enough to trigger the pressure switch on and off? Maybe we're back to dodgy wiring to the compressor?

The Patman

Merry Christmas to all!
5 years from now when someone reads this, they'll realize that Jeremy...and...well...myself, had nothing better to do on Christmas.

Again, I assumed Alan or Mr. Merlin or Sean or one of a million others would jump in and tell ya what's wrong with your car in a single sentence, but until then...

So, let's see. The relay is faulty so that's that. Good troubleshooting. If there is nothing else in the way, it looks like ya found the electrical part of the issue.

Now as for the "on and off again" issue. The answers is yes, refrigerant pressures can fluctuate enough to cause the compressor to turn on and off when it would normally just be on. It's actually a quite common symptom of low refrigerant level.

You already know that you have a "low side" and a "high side" to the vapor cycle cooling system. The system also has a third pressure. Static pressure. Obviously static pressure is the equalized pressure between the two sides of the system when the compressor isn't running. For instance, static, the system may be a bit low on freon, say both sides (hi and lo) are showing 50 psi. Enough to "make" the lo pressure switch be closed. Then The compressor comes on by your command (or perhaps by a faulty relay that stays "on" all the time). As the compressor sucks freon into the intake side (the low side) and compresses it and shoots it out the high side, a pressure differential is formed whereby the low pressure safety switch will now cut off the compressor by breaking the contact to the clutch (but you know all of that) a minute or two later, after the compressor sits static, the pressures will equalize, allowing the 50 psi to be felt on both sides again, and the low pressure switch will once again "make" and the compressor will once again come on...starting the cycle all over.

Now that's the theory. and in fact, that's exactly how it happens in the real world everyday. why is it odd in your case? Because in this scenario, the compressor always goes "on" first, then off. Unless you missed the first "on" cycle, you seem to describe your situation as starting out being "off", then coming on shortly thereafter.
(That could also be a function of the design of the Porsche automatic Climate Control...a slight delay before the sensors get situated. Seems like I've noticed that before.)

But. If you know the relay is welded shut and keeping the compressor "on" all the time, then I'd start there. Later, A quick boost of freon would answer the other question in about five minutes after the relay is replaced.

As a side note: I used the radio shack relay trick on mine and am perfectly happy with the results. And by the way, only two wires need to be heavy gauge in there...just sayin'.

I love helping out with long distance troubleshooting...that is, when ya can't see, hear, feel or smell the problem! But having a smart fella on the other end makes it actually fun! Ya get to help with problem solving, but don't have to get dirty!!

The Patman

Oh. you could easily check the freon static psi with just one gauge hooked up to the closest port. just to see what static psi is,

hwyengr

You know, it's very possible that I'm missing the first cycle between turning the key and running back to the engine bay, but I don't hear or feel the characteristic load change. When checking voltage at the evaporator with the engine off and cold, the clutch clicked on so theoretically it should have turned at the first start. The Mrs. is unavailable for garage help (having a car as low as the 928 is a complication for a person recovering from spinal surgery...), so I'll have to see if I can recruit the extended family when they show up for Christmas this weekend.

I was looking for a reason to pick up a set of R12 gauges...

I got 3 semesters of EE under my belt before switching (HA!) over to civil, so I know just enough to get into trouble!