Hesitation under throttle and wandering idle with A/C

[Crashsector]

For common record, here is a scan of the FSM page discussing the thermo amp:

http://picasaweb.google.com/lh/photo/ph ... directlink

Pin 1 on B71 supplies the "comparator" with 12V. The comparator uses the signal from the diode on the evaporator to open or close the circuit for pins 2 and 4.

So, jumping pins 2 and 4 in the harness will bypass the amp and keep the compressor on continuously. This is not recommended for a long period of time. But it worked for me.

Bitch is that it looks like I have to discharge the system just to get to the sensor. Good to know I know the source of the problem though.

--Andy

[Crashsector]

Wow, just a few updates.

Got the evaporator box apart without discharging the system. Cleaned it out and picked up another thermostat from a junk yard ($3!). Put in... with the same problems.

In a fit of desperation I cut the connector off the wiring harness that was going to the thermostat and put female spade terminals. I plugged the terminals right on to the thermostat and *BAM* it works perfectly. The compressor still cycles on and off but the duty cycle is more like on for 2-3 minutes and off for about 30 seconds (at idle). Given the humidity lately and the probability of the evaporator icing up, this seems normal.

Geeze am I glad to have this put behind me once and for all.

--Andy

[Legacy777]

Good deal. Glad you got it taken care of.

[Crashsector]

Of course this thread isn't dead yet. Symptoms returned shortly after my last post.

I'm running out of things to check. So far, this is what I've done:

- Replaced all 4 A/C relays in the under-hood fuse box
- Replaced the A/C cutoff relay behind the glove box
- Replaced the thermostatic amplifier on the evaporator core (along with the plug going to it)
- Verified proper charge and pressures
- Verified clutch engagement by manually engaging A/C clutch with 12V from the battery (reading 5 ohms of resistance)
- Bypassed trinary pressure switch in receiver/drier

That, to me, seems to be every part in the entire system. The only thing I'm researching now is how the ECU controls the WOT cut off.

Aside from that, does anyone have any suggestions as to what to test?

--Andy

[Legacy777]

Refresh my memory on whether your primary issue is idle related or other right now.

Regarding the AC cut, ECU will cut AC when throttle is at 100%.

[Crashsector]

Idle and at speed. Some times everything works fine, other times the A/C compressor/system clicks on and off. Some times slow (switches every 1-2 minutes) some times fast (up to 5-6 times per second).

More often than not, the clicking on and off is faster at speed than it is at idle.

The problem is so intermittent it is frustrating. I'm out of ideas.

--Andy

[Legacy777]

Your hesitation and different feel in power is likely due to the AC compressor engaging/disengaging. There's really not much to the AC system, so I'm not sure why you are continuing to see issues.

What type of vacuum pump did you use to pull a vacuum on the system? You really need to pull a strong vacuum for 30-45 minutes to completely boil off any moisture in the system. If you have moisture in the system it could be causing ice crystals to form and could clog the expansion valve slightly.

I'm kind of out of ideas other than to start fresh with the charge.

[Crashsector]

Your hesitation and different feel in power is likely due to the AC compressor engaging/disengaging. There's really not much to the AC system, so I'm not sure why you are continuing to see issues.

What type of vacuum pump did you use to pull a vacuum on the system? You really need to pull a strong vacuum for 30-45 minutes to completely boil off any moisture in the system. If you have moisture in the system it could be causing ice crystals to form and could clog the expansion valve slightly.

I'm kind of out of ideas other than to start fresh with the charge.

Thanks for your help Josh. I think I should recap since the thread has gone around the block a few times.

That is indeed the problem I'm experiencing. For some reason the circuit that controls power to the compressor is being broken.

The pressures for the system are spot-on. I also have the same symptoms if I bypass the pressure switch.

I have replaced all 4 relays under the hood.

I have replaced the cut-off relay behind the glove box.

I have replaced the thermostatic amplifier that controls de-icing of the evaporator.

I am 100% convinced this is an electrical issue. If I switch the system on, and connect the compressor directly to the battery, it works fine (so the problem is not the compressor).

It's not the pressure switch, since I bypass it and it still happens.

It's not any of the relays or the thermostatic amp.

I can't see anything else in the system besides the ECU lead that controls the A/C system. My next step is to find that lead, see what its operating voltage is, and wire the ECU out of the system. I don't go WOT and even if I did, I don't need the A/C compressor to cut off.

Any other input as to what I can do?

--Andy

[Legacy777]

You eluded to my next suggestion. Try replacing the HVAC unit. They are prone to having bad solder joints.

[mcbrat]

any further updates on this? I'm having the exact same symptoms.
I have a spare HVAC control panel I can try, but I had one other odd thing occur. the passenger front window would not go back up after putting it down to help get the heat out... it eventually went back up...

[Legacy777]

Have you checked the body ground points where the chassis wiring grounds?

[Crashsector]

Up to this point I've replaced the HVAC control box and also bypassed the ECU's involvement with the cutoff process and it is still happening. It's maddening to say the least.

I'm familiar with two ground points in the engine bay... one on the intake manifold (ECU) and the other on the passenger (?) side frame rail. Any others?

Thinking of selling the car pretty soon so it's not a huge deal.

--Andy

[Crashsector]

Well just for the sake of argument I added some grounds to no effect. I went from the intake manifold, a/c compressor, and driver side strut to the negative terminal and the problem still exists.

The only thing I can think of is some problem with the coil in the compressor, but I don't know why that would be an intermittent thing. I figured it would either engage or it wouldn't.

I don't really feel like dumping $160 for a reman compressor, $15 for an expansion valve, $15 for a drier, and what ever my shop at work wants to charge me for a evac/charge just to have this keep happening.

--Andy

[subydaddy]

I feel your pain, been dealing with this same problem for the past year. I haven't put in the effort you have but close to it, and man is it frustrating. Hope we can figure this problem out, for the sake of everyone in my car pool

Do your radiator fans cycle?

[Crashsector]

I'm happy to say, legitimately, that I believe I've taken care of this problem once and for all.

As the last-ditch effort to get this thing working right, I replaced the compressor (eBay reman, we'll see how long it lasts), expansion valve, and drier today. I was surprised how easily the evaporator came out of the car. Flushed it with AC Flush and brake cleaner and replaced the expansion valve.

The reman compressor had 5.5 ounces of oil so I added another 2.5 ounces and charged with two 12 oz. cans of 134a. Pressures read exactly where they should be, and the system seemed to work perfectly. Plus the clutch bearing on the reman sounds a whole lot better than the one on the old compressor.

The biggest difference is the air coming out of the vents stays cold when the compressor cycles off, whereas before it would get puke warm.

I'm attributing all of my symptoms to a failed expansion valve. The valve is a one-way restriction which is supposed to allow a pressure differential. I think mine was operating erratically, working correctly some times, some times closing completely (causing the pressure to spike and the system to turn off), or opening completely and getting really cold, and wildly flipping between the two. There was literally a small pile of compressor clutch dust on my timing cover from the compressor cutting on and off so much.

I never read that the system is supposed to retain a bit of a differential when the compressor kicks off, which mine didn't originally, but it makes sense. That's why I still have cold air when the compressor turns off.

So the only down side of the whole thing is I swore I flushed all the cleaner out of the evaporator with air and I also let the evaporator sit in the sun for a good two hours on a 90 degree day to try to make sure there is nothing left in it. After charging I noticed the little amount of oil that came out of the service port has the same orange smell as the flush I used to clean out the evaporator I hope this doesn't come back to bite me, but I really only need one more year out of the car max before I'm ready to move on.

Hopefully no more updates to this thread, and maybe the other posters with similar problems can corroborate my findings.

--Andy

[Legacy777]

Thanks for the update Andy!

That's good info to know. Based on your findings, I think the expansion valve on my Impreza is questionable. I do get a bit of cycling of the compressor and do hear a little bit of noise coming from it. I'd have to hook up the gauges to see what the pressure is doing when the compressor cycles off, but I do believe it drops a little. When I turn off the car I can hear a faint hissing coming from the evap box, which is probably an indication that the expansion valve is slightly leaking.

Also, just an FYI, after using AC flush, I've just used brake cleaner to flush my AC parts. That AC flush is slimey and a bitch to clean.