I originally took a 30-minute drive datalogging the entire thing to figure out why the A/C kept shutting off and not coming on again unless I restarted the PCM.
I also noticed a startling trend with my knock sensor. It does have an LT4 KM btw.

startup, closed loop, go for a drive, no knocking. Drive for 20 minutes, most of the time spent at cruising speeds, slowing to a stop, then accelerating, no knock counts occur, engine operates as usual.

After some WOT runs, in which the engine temp quickly goes back to its normal 80-deg celsius temp from a high of 82, I resumed cruising and light acceleration, only this time knock module goes ape with nearly every press of the throttle. car runs rougher at this point.



A/C...
After a brief stint of what I considered normal operation, it dropped out again, so let me explain what I've learned and think I know, and someone can chime in and direct me to the problem.
Engaging the A/C at idle does nothing... pressing the throttle and raising the RPM's above idle truly engages the A/C clutch regardless of what the PCM says it is doing. Should A/C temp (the thermister in the air-stream after the evaporator coil) drop to 2.0 degree's celsius or lower the A/C compressor turns off until it rises above 2-4 degrees celsius. Sounds good to me, keeps the coil from freezing and cycles the compressor based on low pressure effectively, low-temp application A/C units have been doing that for decades.

Problem is, should the A/C pressure (based on a transducer on the liquid line) drops below around 85 psi, the unit will not come on again. I assume this is a "low refrigerant charge" safety, and it makes sense. My problem is my system is doing this during off-cycle, the A/C pressure continues to slowly drop, and if I happen to be decelerating at the time it will certainly precipitate the scenario.
I have had luck getting it to restart only if the pressure happens to be above around 82 psi and I cycle the thermostat (switch from just air to A/C mode on the dash) a few times, but this is rare. Rather the system locks out and instead of equalizing as a TXV should the pressure from that transducer continues to drop into the 70's regardless of whether I route hot air with the duct damper control, or just drive with air on.

Now, refrigerant flows to the coldest point in the system, so I can perhaps...maybe... see that after the evap coil is around 2-degree's and the system shuts off, the TXV is in effect as closed as it can possibly get, but most of the refrigerant is logged in liquid state in the evaporator coil and the suction-line accumulator, and the compressor is not allowing it to quickly equalize with the discharge. But that's what the equalizer tube is for on a TXV (I'm assuming there actually is one?).
But, why the liquid line pressure sensor continues to drop after the system is shut-off is beyond me, the condensing coil is getting cooler which should precipitate the pressure to initially drop, yes. but 20 minutes of driving and it remained at 78.5ish psi give or take a few without budging until after I shut-down the vehicle.
This leads me to believe that perhaps the equalization is controlled by a solenoid that doesn't let the system equalize until it is shut down?

Possible problems that would cause this (just thinking out loud through most of this post):
A. TXV needs to be replaced due to clogged equalizer tube/orifice slightly plugged.
B. PCM has serious issues that are not apparent and I need a new one.

I will note that I've been working on and charging A/C systems for years, every kind of refrigerant you can think of in most systems including ammonia and R-5xx. So I'm pretty sure I've got a decent charge, I just want a second opinion and better understanding on the A/C side of this before I go ripping out the TXV.
The knock business is a mystery to me, and it does seem as though the car runs rough given the same exact temperatures and conditions after its been romped on, even after its back to the same temperatures and conditions it was acting fine at before the romping.

 

What did you datalog with, if you used Datamaster can you send me the .uni file. Were the pressure readings you mentioned in your post from an a/c gauge set or from the PCM readings. Tapping the throttle to pick the clutch sounds strange, wonder if you have a low voltage condition on the a/c relay or clutch. Yes the settings are in the tuner constant table, stock settings are off at 36 F (2 C) back on at 39 F (4 C). I only have a gauge capable of reading the low side I notice the clutch will cut off at 25lb on the low and restart at about 35 with low refigerant. Are you saying your high drops below 81 after the evap sensor kills the clutch, then you throw a code and have to reset the PCM. It should only throw a code 61 if the clutch is picked and then the PSI is low. Can you see in the trace if a/c clutch relay is on (= 1) when the pressure drops below 81. Or at idle if the clutch relay pick is on w/o the clutch actually picked.

 

With regard to the other part of your question..... what exactly does this mean:
Do you mean you see a lot of knock retard?

 

I use tunercats and datamaster, so yes, if you PM me with your email I will send you all the log files in question.
The readings are from the pressure transducer on the liquid line, and I did verify that they matched a gauge tied in at the same point and it was within 1-2 psi, so I'm certain I'm getting good readings.
you will notice if you look at the datalogs the actual clutch does not engage when datamaster says it does, its obvious from the high side pressure increasing when it truly starts, I have verified this visually watching the front clutch plate engage.
A low voltage condition may be causing some problems, the datamaster logs show I'm running a steady 12.5-12.9 VDC throughout every log. And gazing at old log files going back 3 years it appears 12.7-13.2 is what I've always had.

Since there is no pressure reading of the low side, you must be seeing a P/T relationship of the suction line in reference to the evaporator coil temp as a coincidence.

My biggest problem is that I can't find the "code" its throwing, I don't see one, it simply refuses to start the clutch again once the event in question has occured until I shut down the car and start it again.

 

Yes, I think the quickest way is for me to host the datalogs with comments and pointers to show everything on one page.

 

 

Files in order


http://www.filefactory.com/file/ag0ga95/n/1_uni

http://www.filefactory.com/file/ag0gbac/n/2_uni

http://www.filefactory.com/file/ag0gbaf/n/3_uni

http://www.filefactory.com/file/ag0gba3/n/4_uni

http://www.filefactory.com/file/ag0gba4/n/5_uni

http://www.filefactory.com/file/ag0gba8/n/6_uni

 

If you need someone's e-mail address, all you have to do is left click on the user's screen name, and it will offer you the option to send them an e-mail. Works only if the user permits it to work, and bobdec permits that (system default).

 

aifilaw, just sent you a .uni of my a/c operation to compare to yours. In your 1.uni you are cycling on/off between 135-150 PSI and putting out 7* C at the evap coil. BUT the PCM clutch pick and Clutch Status are always =1. That means with +12 at the clutch coil your compressor is cutting out, or the clutch is slipping. The compressor s/b able to put out well over 200 psi without breathing heavy. In a well charged system like yours a compressor can hit 248 PSI (noticable by fans runng at high speed) w/o any problems... My system at 70* F ambient temp puts out 2-4* C at the evap .What confuses me is in the 2.uni file your a/c request from the dash is present but the PCM is not picking the clutch. Then magically after you play with the dash switch the clutch picks . Yet there are no DTC codes set. Some how the pCM is seeing something not apparent in the trace and it is holding off the a/c. By any chance do you have DTC codes 61, 66-72 or 73 turned off in the switch table.

 

When the computer calls for the clutch, its not actually calling for the clutch, I verified this earlier with a voltmeter, but something that does come to mind is perhaps the relay is starting to fail.

DTC codes you mentioned are all turned on. If this persists perhaps the manipulation of these will enable the A/C to run properly?

I would say this all started when I had the A/C compressor replaced end of last years summer season, but fact is that was the first time I've had the A/C charged and functional in 6+ years. The shop that did the job put in the new compressor and made sure it worked. (First time other than an alignment I've let a mechanic touch the car, historically throughout my life its always been a painful mistake, this time was no different). I picked it up and drove it home with the A/C going full blast planning to set the charge properly when I got home. It made it halfway and then did what you saw here, after restarting the car, it did it once more and then wouldn't operate again even after restarting the car.... nothing. I took it back to them and they said "there was a code, we reset it, its working now." After that day it was so cold I never ran the A/C until this year...

So, I presume that if I let it do this a few more times it will actually throw a code I can see.

Also, I did not receive anything, but would like to see someone's operating normally, preferably on for a while, and then switched off to see what the pressures do over time and normal driving after being run, and then turned back on again after a while.

 

Could be the relay, unless something is wired differently in your setup.. here is a schematic from shoebox.. http://shbox.com/1/hvac_compressor_control.jpg Note the PCM grounds pin D8 and picks the relay, once picked +12 v is applied to the clutch coil from fuse 12, The dark green wire at junction S105 goes back to PCM pin B21 as the clutch status. You can easily swap the relay (the fan relays, fog lite and air pump relays are all the same p/n). Assuming you are familiar with electrical wiring it's also possible to attached very lite guage 5-6 inch long wires to the relay connectors then plug the relay back in to test out the circuits. EG: attach a wire to relay pin B5 and monitor the 0V pick from the PCM, attach a wire to B4 to monitor +12 v to clutch.

As far as pressure bleed down, never timed it but I know it takes a long time for the hi/lo sides to equalize after shutdown. I've watched it up, gave up walked away and came back after 1/2 hour and it was even.

[IMG][/IMG]

 

I noted something today that I didn't see in the log file, on A/C startup the VDC (voltage) gauge on the dash took a 50% nose dive for half a second, "check gauges" light flickered.
Guessing that clutch coil is putting a pretty heavy taxing on the system, either that or battery/alternator needs replacing.
compressor/clutch were replaced with a brand new unit, not a refurb.
the main problem could indeed be caused by the order of:
faulty/dying A/C relay coil
faulty/dying battery/alternator
faulty/dying A/C clutch coil

considering the unit might not be coming on, like a locked-rotar situation, except at 12VDC. or the A/C relay might simply not be engaging.

All of this still does not address the question of how the liquid line pressure drops below 80 psi, and then with the compressor not coming on, stays below 80psi instead of equalizing after 15 minutes

 

aifilaw, This is a guess not based on experience.. Our systems use an "Expansion Valve System" verses an "Orifice Tube System". The suction at the low side of the evaporator coil controls how much the expansion valve opens to allow liquid flow from the high side into the evaporator. If this valve plugs or closes completely then the only way a system would balance after shutting off would be leakage through the compressor. I do not think the rotary compressor design leaks high pressure back to the low side. If not then the system will not balance through a defective/plugged/closed expansion valve. Usually when a compressor is changed it's a good time to remove and clean the expansion valve, similar to replacing the orifice tube while the system is opened up.

 

I finally got it to throw a code, MALF 67: Press Sensor Degraded
But this code did not stop the A/C from coming on.
I had a 50 minute drive ahead of me, mostly highway with some stop and go, logged the entire thing, ran just like it was supposed to until I ran into traffic and came to a near dead stop. On the off-cycle the compressor turned off, pressure was at 108psi this time, A/C clutch status came on, but clutch did not physically engage. A/C pressure slowly dropped to 90 psi over the course of 8 minutes, after about 3-4 times of cycling the control it finally did re-engage and worked fine for the rest of the highway and 6 long traffic lights, cycling on and off as it should.

Still rather perplexed...pressure being too low to start isn't the issue, so I'm no longer blaming the expansion valve and equalizer tube. I would expect the clutch to fail at starting some other times if it was a weak coil, and the A/C relay is brand new. Unless someone else can think of something I believe I'm going to run a manual jumper to the relay and force it on and off while logging to try and narrow it down.

 

Code 67 says the a/c clutch was turned off for more than 10 seconds (pressure was then recorded) and then after turning it back on the pressure did not increase 4 lbs higher than the recorded off pressure within 10 seconds. Same symptom as what you are seeing. The code does disable the a/c system, however I could not find what test or set of events will re-enable it. Something in the 8 mins and cycling the a/c switch on/off could be re-enabling to system. It's all pointing to the wire from the relay to the clutch, and the splice where the wire back to the PCM (clutch status) is tied in. You can get a false 67 if the clutch status line signals the clutch is off or on when the clutch is actually in one state. You can try your switch, forcing the a/c relay pick to see if it's a clutch circuit or PCM to relay fault. Remember you are bypassing all a/c safety cutoffs, low charge, low evap sensor temp (freezeup), over pressure cutoff.