mjdiaz89

Hi, all.

It's my first time posting on CF, so hi, everyone

I have a bit of a mystery with the air conditioner on my 2009 Honda Civic. After Googling for many hours on end, I haven't found anyone with a similar issue. Check out this mystery and let me know what you think.

• The AC system works fine in the morning for about 20min, then hot air all day long (thermometer in the vent recorded 120F).
• I took it into the shop. They say it was low on R134a. They filled it up.
• Next morning, it worked for a whole three minutes longer, but, alas, the same thing happens.
• Thinking I had a leak, I bought a can of ACPro. I waited for the air to blow hot again before connecting the can to the L pressure side. It read 110psi on a 98deg day. I noticed the compressor clutch was not engaged.
  I take it back to the shop and ask them to reduce the amount of R134a in the system, thinking that the system shut-off because of overpressure.
• Their machine reads that it's actually low on refrigerant. They vacuum it and replenish it with dye this time. I was told to go drive it for an hour and bring it back.
• A lot of dye was found around the expansion valve. $500 later, that and the pressure transducer were replaced. I was told that should fix my problem, since, yes, the expansion valve was not seating well, and if it doesn't, it means that I need a new compressor ($800 for the compressor and $400 in labor, which, yes, its ridiculous )!! Dismissing that quote as ludicrous, I start looking into the problem myself. The problem is still there, but the air temp is now ~80F when warm air blows, an improvement, I guess.
• I start to really get into it. The compressor clutch stops engaging, thus the hot air. I start to think the problem is electrical, as why would it work for 20min when it's cool outside and then it quits?
• In the morning, when the system is working well, I hooked up ACPro to the L side again to get a pressure reading. At 90F ambient, 54psi is recorded. Not bad. It's a bit on the high side, but it's still acceptable. 20min later, when the clutch stops engaging, it reads 110psi again on the L side.
• Thinking that the issue was happening because some electrical part is getting too warm and then the circuit somewhere opens, I tested every relay with a continuity tester. They all have continuity.
• I verify that the fans and blower are all operating and the cabin air filter is clean. Unfortunately, I cannot reach the compressor from above (I do not have ramps/jacks), so I cannot check if the Compressor Relief Valve is causing a cut-off or if the compressor is even receiving electricity, for that matter.

Any ideas? Does this sound electrical? I do not have schematics, but, electrically-related, I think I only have the evap temp sensor, the in-cabin thermostat, and the clutch circuit left to check, and they are all pretty hard for me to put a continuity test on them without either disassembling half the dash (please correct me if I'm wrong) or going from under the car, respectively.

FWIW, I also tried performing a DTC test as per http://www.pontiacpower.org/pdf/ACtestUS.pdf, but, assuming I did it right, no DTCs blinked. I also tried a self-diagnostic and got an "Error Code 1" message, which I have yet to find out what that means...

What doesn't make sense to me is the clutch not engaging and then having a crazy high reading on the L side. Unfortunately, I don't have an easy way of getting a read on the H side.

I would not like to throw more money at a new compressor unless I'm sure that's the problem. Yes, I see that a new compressor is really more like $200 on Amazon, but I doubt I'm qualified to remove the old one and install the new one myself, so, unless it's actually a lot simpler than I think it is, I'd still have to pay the shop to install it for me. If I do have to take it in, I'll take it to a Honda dealership shop.

Any ideas? I'm pretty frustrated with this thing.

ezone

Welcome,

Sounds electrical in nature to me so far.
Don't worry about pressures YET, your complaint began with the compressor shutting off and that's electrical.
Your pressures may be a problem in and of itself, but it's separate from quitting completely.

You have a test light and/or volt-ohm meter?


When the AC quits and blows warm, you need to check if the
--interior blower fan is still running (probably)
--radiator fans are still running (and cycle on and off when you turn the AC button on and off).

If all 3 fans are still running, go to the compressor clutch relay and tap on it.
If the radiator fans do NOT run with the AC button turned on, don't bother with the above because this will go a different direction.


You can manually jump the clutch relay and note if the clutch clicks on
If you know how to backprobe a connection, you can check for power at the compressor connector easily. (The 3 wire connector on top of the alternator is for the compressor.)
You can ohm test the clutch coil easily from there, but it's gotta still be inoperative at the time you test it.

mjdiaz89

All i have is a test light at the moment. Sounds like I need a ohm-meter, tho?

Yep, blower is running and both radiator fans cycle on and off like they are supposed to. I do not hear any "ticking" or other weird noises coming from them.

I'll have to try tapping, and if needed, jump, the relay. I'll move onto backprobing the connection, then ohm test the clutch. I'll report back with what I find.

Thanks!!

ezone

I usually begin with a low watt incandescent test light, used as a simple continuity tester.
The 3-4 ohms resistance of the field coil should still let the test light bulb illuminate during the check in the relay socket.
That tells me you have request from the HVAC panel to the PCM ok, and the PCM is enabling the AC system to operate (evidenced by fan operation)

So the issue is most likely somewhere between the clutch relay and clutch.

Relays stick on or off, tap on the clutch relay and see if the compressor comes to life (replace relay with updated part)

I pull the relay and use the test light to see if the socket has 2 powers and 2 grounds that will light up the bulb

Test for power going to the clutch,

or jumper the 2 large terminals in the relay socket to see if the clutch will click on

and if those above have no results I can ohm test from the relay socket first as it's easy and I already had the relay out.

if the coil has correct resistance yet will not click on when power is applied to it, then check for excessive gap between the plate and pulley. (high mileage and wear) Apply 12v power to the coil and then push on the clutch plate to see if it will pull into the pulley.

mjdiaz89

I had to wait until the weekend to be able to do the stuff you mentioned. Here's what I found:

I bought a multimeter to get more resolution as to what's going on. I let the car and compressor run until it the clutch wasn't engaged anymore. I removed the relay (I used this guide http://www.odyclub.com/forums/attach...odyacrelay.jpg) to test the voltage off the dash button and the relay. It was reading 13.9V, so that verified the dash button was working as it should. (We already knew that the PCM was getting the signal, but maybe the compressor wasn't for some reason. Nope. It's fine.)

I kept testing the voltage across the "always hot" and the a/c stator and, yes, it was reading 13.9V.

It must have been a while, say, some 30min, before I put the relay back on and, behold, the CLUTCH ENGAGED! I was really surprised. Thinking then that my relay opens when it gets hot (it almost burned me of how hot it was when I removed it the first time), I went to the auto parts store and bought a new one ($6). That really didn't do anything, as the same behavior was observed. I kept performing diagnostics.

Leaving the engine running, I removed the new relay again and waited some 30min. I put it back in and, yes, the clutch engaged. I immediately removed the relay and tested the clutch coil's resistance through the a/c stator pin and using the engine block as ground. It was reading 5.4ohms. I don't know if that's a tad high, but at least there's continuity. Thinking that then the clutch coil circuit was opening when it gets hot, I placed the relay back in and waited for the clutch to disengage again. I removed the relay and read the resistance again: 5.4ohms. No dice there. I put the relay back on and backprobed the clutch connector (the one on top of the alternator). It was also reading 13.9V. So, yes, the clutch is most definitely receiving power when it's not engaged.

For giggles, I jumpered the clutch from the relay when it disengaged. As expected, it did nothing.

This makes me think that something electrical is making it disengage when something gets too hot. Am I right in thinking that? What else can I probe that's downstream from the relay?

Also, you were right about the "high" pressure on the low side when the compressor stops working. That "high" pressure is my equilibrium pressure! I checked this by putting the pressure gauge on it when cold and watched the readings throughout the entire compressor cycle (cold engine on, engine warm-up, operating temp, etc.) So nothing to diagnose there. It was reading a nice 45psi when running and spitting out a nice 38deg air from the center vent (until it stopped, of course).

Thanks for your time and help -- I really appreciate it!

ezone

How many miles are on this car?

Get it to quit working and shut the engine off.
Use a jumper wire across the 2 large female terminals in the clutch relay socket,
Compressor does NOT engage? (No loud click)

Then go to the compressor and push on the outer plate, see if it will pop in and engage the pulley.

See how much clearance there is between the plate and pulley
Spec is 0.020" (0.5mm) and I bet you probably have a full millimeter or more of movement? (You can't really measure a worn clutch with feeler gauges as the pulley is probably dished)



Some people have removed the outer plate and removed the shims to reduce the excess clearance. IDK what you have though.

mjdiaz89

112k miles.

Sounds good. I'll do that and report back.

There's not much space to do that from up top, but I'll try that too with a broomstick or something else that's long enough. I'll see what I can do about getting a good spacing measurement.

Thanks again, ezone!

ezone

Measure the clutch coil resistance when it's cooled off too. I wonder if your 5+ ohms will come down to like 3.2-3.5 ish ohms @ 68*F.


If you can power it up and push on the plate to get it to engage, then either the field coil is weak or the gap is too big. If resistance is OK cold then the gap must be too big.
Or it isn't getting full battery voltage to the coil.

mjdiaz89

Here's what I got:
-4.1ohms when cold (85F ambient).
-You're correct, the gap is a full mm.
-When it quit, I turned off the engine, jumpered it, and started it again like you said and the compressor did nothing; I verified it was receiving a solid 14V at the harness.
-I pushed on the plate when it quit AND IT WORKED!!

I did notice something: when it was cold and I was waiting for the clutch to disengage, above 3,000rpm, the clutch would disengage. Letting go of the pedal made it reengage. This happened three times before it quit completely. I don't know if the compressor is tied to an inhibitor or not, but I figured I'd mention it.

Looks like I need to adjust the air gap on the clutch?

ezone

Yes..... it would seem the gap between plate and pulley is far too wide.
Spec (new) is 0.5mm +/- .15mm BUT
You cannot really use feeler gauges to measure the gap on a worn plate and pulley, you almost need a dial indicator.

No matter, there isn't much room for adjustment......
The shim(s) under the clutch plate are probably not thick enough to attain the spec anyway..... most are much thinner than a half mm.
You can remove the plate, remove the shim(s), and reinstall the plate and see how it goes.


^This is a hack repair.^
Proper fix would be to replace the clutch set, plate and pulley. And maybe the field coil.


The PCM can inhibit AC operation as it sees fit.

Opening the throttle more than (some certain % amount) will turn off the compressor for a short period of time so you can have full engine power for acceleration. I think rapidly opening the throttle may have the same effect.

Not sure if it should do that above 3000 though. I don't remember ever noticing that. See what happens after you get the AC working consistently again.



HTH

Eric05

Did you ever figure out solution? I'm having same problem.

mbray

Check high side pressure. Hondas are notorious for this. Any blockage or restriction will cause ac to cut off at about 2500 - 3000 rpms. Also, too much freon will do this as well. Removing about 3 ozs at a time until it continues to work at higher rpms will often cure the symptoms, but a blockage is usually the disease ( unless it was just overfilled)

Eric05

thanks mbray,
Maybe it is a blockage, but it seems to only stop working when it gets hot outside. It works fine in the morning and evening no matter how long i've been driving.

My mechanic has recharged it and told me that if that didn't solve it, it is probably something else.
I forgot what he called it, but he pointed to a diagram of it. I think it was by the condenser. I should have paid more attention. He said it would be $200-300 to fix. I'm thinking expansion valve?
Looks hard to get to. If that's what it is, is that a fair price? Should he replace anything else while he's in there?

09ellig

[https://www.google.ca/url?sa=t&rct=j...23664746,d.amc

This worked for me

mjdiaz89

Hi, guys.

Sorry I didn't post what my results were. It turns out that ezone was dead right . My clutch was spaced a full 1mm. It was a pain to get a good measurement without getting under the car, but cellphone camera,a flashlight, tape measure, and patience did the trick.



I passed by my local auto shop to ask the them tighten it, but they said they didn't have the clutch puller. So, I went to Pep Boys and loaned a "Deluxe A/C Clutch Hub Puller," (https://www.pepboys.com/parts/tools/) gave it to the guys, they put it to the 0.5mm spec, and voila, my clutch started engaging consistently. They mentioned it's at the stop now, so it can't be tightened any more. It's been working for almost a full year now despite the continuous ~100F days.

You guys rock!

Hope it helps, Eric05!