AC Compressor cycles

[~MoMo~]

1998 GS-t.

I need some pointers figuring out why my ac comp. is going on/off every 3-5 seconds. I vacuumed the system, drained oil and refilled to spec.

I measured the dual pressure switch, even with the connector hotwired the compressor would still cycle. Are there other switches/sensors for the AC control unit that can cause this?

 

[uummm]

From what I know, the compressor cycles on and off based on the high pressure switch. When I filled the ac on my old blazer last summer it seemed like the compressor would stay on when the ac was on. Have you checked to make sure you have the right ammount of charge in the system?

 

[JusMX141]

The pressure on the low side will vary based on the ambient air temperature. In the hot summer weather, low-side pressure can range as high as 50-55psi but if it's below 50*f your line pressure will be 30psi or under with the same amount of refrigerant in the system....hence opening and closing the low pressure switch and causing the compressor to cycle.

 

[~MoMo~]

Correct amount of charge, yes. Was done professionally.
What I don't get is that if I hotwire the dual pressure connector the switch isn't in the loop at all and the AC-ECU shouldn't get any sensor feedback telling it to shut down the compressor. I was under the impression that hotwiring the pressure switch would force the compressor to run?

Something is telling the ac-ecu to shut it down, the ac fan also shuts off with the compressor cycles.

Does the engine ecu provide any signal or support for the ac-ecu? Like ground or something?

 

[luv2rallye]

The FIN thermo sensor is probably bad which unfortunately on the turbo model, comes together with the AC-ECM (NT it has a separate plugin). I seem to remember having this problem. I just got the new FIN sensor, cut the AC-ECM wires to the old one, replaced it with the new sensor, and soldered the new ones wires to the AC-ECM. I probably cut and resoldered the wires inside the AC-ECM itself if I could get it open.

 

[~MoMo~]

Ok, is there a way to test the FIN sensor? Or is it integrated in the ac ecm board? Where did you buy a new sensor? Mitsubishi?

 

[luv2rallye]

Check resistance of FIN sensor - should be around 5k ohms @ 68*F and higher as the temp decreases (12k @ 32*, 19k @ 14*). The sensor is inside the evaporator (cooling unit) with wire(s) that come out and go into the AC-ECM (small box mounted outside the evaporator that you can easily see). [Long thread about FIN and ac-ecm here: http://www.dsmtuners.com/forums/new...ac-ac-brain-controller-ecu.html#post151137545].

It's been years but I think I got a new one at Mitsu. Probably asked for just the turbo FIN sensor and if they wouldn't sell me that one then got the non turbo one (since it's sold separately from the ecm - turbo one isn't) and soldered the wire(s) into the ecm where the old one went.

 

[~MoMo~]

Great post, very informative!
I'll measure it and see where it's at.

Ok, back with some results.

Car running, AC button ON:
Measured input (blue/blk) 13,21V
Measured output (grn/orange) 12,91V
Ground is good.

BUT, when the compressor shuts off the output drops to 0v, input is still 13,21v.
I couldn't get a good resistance reading for the FIN sensor, but it's got to be either the FIN or the ECM is bad.
Is it possible to measure the FIN through the ECM pins, or do I need to probe the yellow wires?

Some more; I pulled the ECM board out of the housing and measured the FIN resistance from the soldering points on the board. At 65F it was around 5,1K so seems to be within spec.

Bypassing the entire ac ecm made the compressor and fans run like they should, the passenger fan even kicked in which it never did previously.
I also measured the FIN resistance as the AC was running and it was steadily creeping upwards as the temp dropped. Seems to be fully functional.

Bottom line, I believe I need a new AC control unit.

Thanks to luv2rallye for all help and info, very much appreciated!

 

[luv2rallye]

Car running, AC button ON:
Measured input (blue/blk) 13,21V
Measured output (grn/orange) 12,91V
Ground is good.

BUT, when the compressor shuts off the output drops to 0v, input is still 13,21v.

Ok now wait a minute. These are the CORRECT voltages for a functional AC-ECM!
I assume you have the blower switch in a non-off position along with the AC button pressed.

Bypassing the entire ac ecm made the compressor and fans run like they should, the passenger fan even kicked in which it never did previously.
I also measured the FIN resistance as the AC was running and it was steadily creeping upwards as the temp dropped. Seems to be fully functional.

Bottom line, I believe I need a new AC control unit.

Bypassing the AC-ECM will apply voltage to the rest of the circuitry so that doesn't prove it's bad. Does the compressor still excessively cycle in this condition when left that way? Excessive compressor cycling (other than bad AC-ECM or FIN sensor - both of which you just bypassed) could be:
1) bad dual pressure switch
2) bad AC refrigerant temp switch
3) improper refrigerant or refrigerant pressure
4) bad AC compressor clutch relay
5) poor condition condensor (radiator looking thing in front of water radiator)
OR
6) It's cycling because of the following safety issues: The refrigerant temp gets above 311*F so AC refrigerant temp switch turns it off. Or low side pressure goes below 28psi so dual pressure switch turns it off. Or high side pressure goes above 455psi so dual pressure switch turns it off.

 

[~MoMo~]

Well yes, but the AC ECM should not drop voltage out at the ecm connector pin when the FIN sensor is working correctly? How does the ecm *know* what the other sensors are "reporting"? Aren't they just blocking the current with their increased resistance whenever such conditions are met? Let say too high pressure in the compressor LO side, wouldn't you still have voltage at one side of the dual pressure switch?

The compressor does not cycle when looping the ecm connector (removing the ecm from the circuit), so unless I have completely misunderstood something this should point to a faulty ac ecm? As far as I'm concerned, the AC works flawlessly with the ac ecm out of the loop?

I had a quick look at the ecm board and nothing stood out, only one capasitor that I could see and it didn't show any sign of leakage.
Something is not working right within the ecm, maybe a transistor failure?

 

[luv2rallye]

Well yes, but the AC ECM should not drop voltage out at the ecm connector pin when the FIN sensor is working correctly? How does the ecm *know* what the other sensors are "reporting"? Aren't they just blocking the current with their increased resistance whenever such conditions are met? Let say too high pressure in the compressor LO side, wouldn't you still have voltage at one side of the dual pressure switch?

First I'm not an AC expert but more of an electrical guy. So I look at the electrical diagrams to answer many questions in this area. Here is the wire diagrams: http://www.dsmtuners.com/forums/attachment.php?attachmentid=111540&d=1279076129.
The FIN sensor measures air temp at the evaporator and inputs it to the ac-ecm. So the ac-ecm can turn off it's output voltage (green-orange wire) based on it. Now I don't know if it continuously cycles the voltage based on this or if the FIN sensor is really there to just sense if the evaporator is freezing in which case the ac-ecm just turns off the voltage (no cycling) for safety.

The ac-ecm does not know what the other sensors are reporting. Each sensor, including the ac-ecm, is just a switch - all in series to the ac compressor clutch relay. So if any don't switch on, the ac compressor will not go on. See wire diagram.

The compressor does not cycle when looping the ecm connector (removing the ecm from the circuit), so unless I have completely misunderstood something this should point to a faulty ac ecm? As far as I'm concerned, the AC works flawlessly with the ac ecm out of the loop?

When you say "looping the ecm connector (removing the ecm from the circuit)", I'm assuming you remove the ac-ecm but then jumper the blue-black to the green-orange - right?
By "not cycling" do you then mean it stays on or stays off?

I had a quick look at the ecm board and nothing stood out, only one capasitor that I could see and it didn't show any sign of leakage.
Something is not working right within the ecm, maybe a transistor failure?

You may be right, the ac-ecm may be bad.

 

[~MoMo~]

When you say "looping the ecm connector (removing the ecm from the circuit)", I'm assuming you remove the ac-ecm but then jumper the blue-black to the green-orange - right?
By "not cycling" do you then mean it stays on or stays off?


You may be right, the ac-ecm may be bad.

Yes, jumping the blue-black to the green-orange.
The compressor stays ON when doing this, to me that tells me something is up with the ac control unit. FIN thermo sensor resistance steadily increased with AC operation, though I had no way to measure actual core temp it looked like it was performing as it should.

I don't think the ac ecm is designed to cycle it's output that rapidly unless you are close to freezing core temps, which after only 5 seconds of A/C operation is impossible to achieve.
You seem to really know your way around schematics and electrical systems, after looking at the way my ac ecm operates wouldn't you tend to agree that it's performance is not as designed?

 

[luv2rallye]

Yes, jumping the blue-black to the green-orange.
The compressor stays ON when doing this, to me that tells me something is up with the ac control unit.

Yes according to the wire diagrams I agree that jumping those wires would keep the compressor on (as long as the dual pressure switch and ac refrig temp switch don't trip).

I don't think the ac ecm is designed to cycle it's output that rapidly unless you are close to freezing core temps, which after only 5 seconds of A/C operation is impossible to achieve.
You seem to really know your way around schematics and electrical systems, after looking at the way my ac ecm operates wouldn't you tend to agree that it's performance is not as designed?

Yep, you convinced me, it would seem so.

 

[hunkajunk]

It helps to know the high and low side pressures.

I think your expansion valve could be stuck.
This would cause high pressure on the liquid side and possibly freezing the coil

 

[mcurtiss]

hunkajunk is right, before i begin checking components, i always put gauges on the system. High & low side pressures are crucial in diagnosing ac problems. Correct oil & charge is key, but if the system has a restriction, the compressor will still cycle fast. I would also make sure the cooling fans are operating, & one more quick thing to check is to make sure the condensor is not clogged with debris. If airflow cannot get through the condensor, pressures can run abnormally high & cause quick cycling as well.