Automotive Air Conditioning Information Forum (Archives)

Provided by

We've updated our forums!
Click here to visit the new forum

Archive Home

Search Auto AC Forum Archives

How to test clutch on compressor

BartR on Sun March 31, 2013 9:05 PM User is offline

Year: 2006
Make: Honda
Model: Civic EX
Engine Size: 1.8L-4cy
Refrigerant Type: 134a
Ambient Temp: 77 F
Pressure Low: 80
Pressure High: 80
Country of Origin: United States

I'm kinda new at AC work, somewhat of a diy kinda guy, but need help. Here is what I think i have concluded so far.
1) Engin running, turn on AC and all lights on dash turn on.
2) When AC is turned on, Fan turns on and I think RPM bumped a little.
3) No compressor kicking in.
4) Checked fused under hood (20 & 9) and under dash (36). Tested 9 & 36 with a meter and visual check on 20 via sight window.
5) Hooked up manifold gauge and got 80 psi on high and low side at 77 degrees ambient temp. I think with these pressures I should be good on refrigerant and it should kick on the compressor.

I don't see a Low pressure switch to check like in my older vehicles near the firewall or a dryer (I'm old school, not use to all the new computer stuff) , although I don't think that is the issue. I think it lies with the clutch on the compressor. There are 3 red wires that have a connector that attaches to the alternator that lead down the the compressor. These wires are all wrapped together until they reach the compressor then 2 of the wires go to the back of the compressor, to a Thermal Switch I believe. and the other to the front of the compressor at the clutch where a ground also ties in.

I think that I need to somehow check the Thermal Switch? and the Clutch. Not sure whitch wires are which at the 3 wire connector and not sure how to check the clutch.
How can I check these two things so see if they might be my problem. I have read on these forms in places about jumping the clutch to see if it engages with the car off but I'm not sure how to test this with the 3 wire connector or what the voltage should be.

I'm also not sure about when this thing gets to working, what my pressures should be on the high and low side at a given temperature. Are there some charts or specs that someone can refer me to.

Thanks for any help and any other suggestions to look for. This is my daughters car and I would like to keep her cool in the upcomming heat of the summer.

GM Tech on Sun March 31, 2013 9:32 PM User is offline

I always start at the a/c relay- in underhood fuse block- this cuts the circuit in half- does it have an a/c request? ( a negative input on neg relay coil?) Does relay click (can you feel it?) as you gently plug/unplug it? If so, your issue is the relay or a problem downstream towards the compressor - if not, then it is usually an input issue, like a pressure switch or transducer, or control head.

Next, I'll force 12v (by using a jumper from battery) to the leg of the relay that is switched (that feeds the compressor coil) (this leg is usually a wider blade and matches the one that is NOT battery hot when plugged in). Does coil click in? If so, your compressor is fine, if not- then trace towards the compressor- keep jumper connected with 12v, and wiggle all wires, and then tap on clutch front disc- does compressor snap in? if so you have an air gap issue...

Try these first then report back.

The number one A/C diagnostic tool there is- is to know how much refrigerant is in the system- this can only be done by recovering and weighing the refrigerant!!
Just a thought.... 65% of A/C failures in my 3200 car diagnostic database (GM vehicles) are due to loss of refrigerant due to a leak......

WyrTwister on Mon April 01, 2013 3:36 AM User is offlineView users profile

I know nothing about Hondas , but on GM cars , I do as described in the previous post .

In fact , I jumpered around the relay , Saturday , on my Wifey's car . I wanted to make sure the clutch would work , before continueing to charge the system with freon .

This is a 1996 Lumina , that had been parked for 3 - 4 years .

The clutch kicked in fine , I proceeded to replace the low side Schrader (sp) valve , vacuum it down , check for leaks and gassed it up .

Get a 12 volt test lite at the parts store or Harbor Freight . It will help in tracing things down .

God bless

BartR on Tue April 02, 2013 8:45 AM User is offline

Thanks for the tips GM and Wyr. Wednesday is my day off and I will be able to work on my Daughters car then. I will try some of the tips offered and see what happens. I'll reply back then.
Thanks and God Bless!

NickD on Tue April 02, 2013 12:24 PM User is offline

Been awhile since I worked on a Honda, back then the coil wire with just two leads, one was grounded, other was a single lead connected to the vehicle harness. With all the corrosive aluminum, ran into problems getting a good ground. Had to disconnect the compressor and use a wire brush, plus added silicone to get a good ground.

Like using a lab type current limited power supply for just testing out the coil. Vehicle connector must be disconnected first because that may have a six cent flyback diode someplace, and reversing the polarity could fry it. With a variable power supply can increase the voltage and expect the clutch to pull in at the 9-10 volt range.

Good to check the clutch gap, these normally have shims, with clutch wear, if that gap increases, it just won't pull in. I prefer a 20 mil gap. Then I apply a full 14.5 volts and note the ammeter reading, leaving it on for around 10-15 minutes, that initial current should decrease. These are random wound coils, never seen one yet that was varnish impregnated, and with heat, can actually short out those coils. Noted by an increase rather than a decrease in current at the rated voltage. Its nice to get the clutch out of the way.

Back then, they were using a dual function high pressure switch that would give closed contacts if the pressure was between 40 and around 415 PSI, as long as that static pressure was within this range, an ohmmeter should show 0 ohms of resistance. Today most vehicles switch to thermistors that can feed either the BCM or the PCM, good to have a circuit diagram.

Key cycling switch was a capillary tube in the evaporator operating a real switch, would close if the evaporator temperature was above 39*F, and open again if the temperature dropped below 33*F. Not to be confused with the TXV capillary tube that controls the flow rates of that variable orifice dependent on temperature again. Perhaps they are also using cheaper thermistors for this also, again to see a circuit diagram. These two switches directly controlled the coil relay that also require the blower switch to be in any position other than off. Back then open blower motor resistors would prevent compressor operation. Easy enough to test, just make sure the blower motor runs at different speeds in all positions except off.

These were cycling systems and as such would get both clutch and contact wear, but not quite as bad as fixed orifice systems. With these new systems, may even need a scanner to help you troubleshoot, and with this stupid storing firmware in flashram, may even have to have your vehicle reflashed that can be the cause of your problems.

First approach is to get your hands on a manual to learn what they are doing. can be of a big help.

Edited: Tue April 02, 2013 at 12:24 PM by NickD

AutoCool on Wed April 03, 2013 6:25 AM User is offline

BartR, I believe your compressor is a Sanden TRS090 which is a scroll compressor. The coil has 2 wires, one is grounded to the compressor body and the other one has a quick disconnect (bullet style) where it meets the thermal switch on the compressor body.

The car has a simple binary cut-out pressure switch that closes at 28 psi, opens if system pressure reaches 454 psi, and closes again when pressure drops down again to 385 PSI. 2 wires, in and out. The thermal switch on the compressor body is a bit of overkill, if it's bad we usually just bypass it. Not sure if you can even purchase the thermal switch separately as a part.

Though it looks like more wires, you should have only one wire to the compressor itself. If you can reach the plug (grey or "greasy icky colored) then you can test resistance between that wire and the compressor body with a standard ohmmeter. Something between 2.8 ohms and 4.4 ohms should be fine. Zero ohms is definitely not OK nor is much over 4.4 ohms.

If you get a reading of zero ohms, that can be due to a bad coil, bad grounding, or a bad thermal switch. An easy test of the thermal switch is to just disconnect the "bullet plug" going to the coil. If you suddenly show resistance on your meter, then the thermal switch is bad. If still no resistance, then it can still be a problem with ground. The magnet coil frame on your compressor is not grounded.

Just as NickD wrote, the clutch gap is very important. If the gap is too wide then the clutch may work when you first start your AC, but the gap widens when things get warmed up. Less gap = faster engagement and less heat from slippage. Good luck, enjoy your day off.

mk378 on Wed April 03, 2013 8:53 AM User is offline

The Honda compressors with 3 wires have the thermal switch in the low-currrent control circuit rather than directly in series with the coil. Unplugging the compressor to test it, the two wires to the thermal switch should be closed circuit and the coil wire about 3 or 4 ohms to ground.

BartR on Thu April 04, 2013 11:45 PM User is offline

Thanks for the advice! With your help this is what I have to report.

I ran the car on ramps and removed plastic under the car and then I got a clear view of the compressor. I could see the bullet style connector that AutoCool mentioned and disconnected it. I then checked the coil and got 3.7 ohms which fell the the range mentioned (coil=good).

Then I rigged a jumper from the battery with a momentary switch in line so i could apply power when needed and hooked it to the bullet connector. Applied power and the clutch activated (good ground and clutch gap). So next I started car and hit switch on jumper for a minute and low pressure line quickly got chilly so things are looking up on having expensive repairs, I think.

After disconnecting the bullet connector, I could figure out which wire went to the clutch coil and the other 2 wires at the 3 wire connector on the Alternator went to the thermal switch as mentioned by mk378. I tested the other 2 wires and got a closed circuit (thermal switch=good).

I very much appreciate the advice from everyone here!!! I guess my next step would be to check the relay and see if it is good and then try to work my wiring in what direction that leads. Also the pressure switch is on the high pressure line and has 3 wires going into it. Any advice on checking it. Would 2 of the 3 wires have a closed circuit if working correctly?

Thanks everyone for your advice and God Bless!

mk378 on Fri April 05, 2013 12:43 AM User is offline

Three wire "switches" on modern cars are actually sensors which feed a variable signal to the ECU based on pressure. It's not useful, and potentially damaging, to try to probe them directly.

If the fans come on and the rpm bumps up, the ECU seems to have decided everything is OK to engage the compressor. That really just leaves the relay, fuse, coil (already tested) and associated wiring.

Edited: Fri April 05, 2013 at 12:45 AM by mk378

BartR on Fri April 05, 2013 8:33 AM User is offline

Thanks mk378, you've been a big help. I'm old school and with these modern (computer controlled everything) cars I'm afraid of damaging the computer by doing some probing, jumping, etc. . Will it damage anything to just jump the relay with battery power to see if everything down stream of the relay is OK.

Also, my last post was a bit late for me last night, since I have to be at work by 6:00 AM, and I forgot to ask if there were any references or if anyone on this site could provide a wiring diagram for the AC.

Thanks everyone for your advice and God Bless!

Edited: Fri April 05, 2013 at 8:38 AM by BartR

BartR on Sun April 07, 2013 3:27 PM User is offline

Thanks to everyone who offered their advice. I have learned a lot from this DIY challenge. I have found my AC culprit to be the A/C compressor clutch relay in the under hood fuse/relay box.

It was suggested early on that I check the fuse and relays first, honestly I was able to check fuses because the car manual had them listed, but, it did not list what the relays were and I really didn't understand how relays worked or could be tested. Through a little internet research, I was able to locate a diagram of the relays and find out how to test them and to test the base that they plug into.

Just a little information for any other DIY newbie out there testing a relay. My relay did click when applying power to it but it did not close the switch, therefore do a complete check of the relay, don't just listen for a click.

Any way, a new relay fixed my problem. Still don't understand what would cause a relay to fail but maybe I'll try to understand that another day.

Thanks! again for everyones input.
God Bless!

mk378 on Sun April 07, 2013 6:09 PM User is offline

Congrats on the successful repair. You could bust the bad relay open just for curiosity, you will probably find the contact points are burned away.

NickD on Sun April 07, 2013 7:27 PM User is offline

Point contact relays have been a continuous problem, just take your thumb and index finger and lightly touch the tips together and that basically emulates what is happening on the inside. Try the same action again with a mustard seed or a grain of salt in the tip of one finger, would never make contact.

Relays are rated for either 1,000 or 10,000 operation, typically use silver plated contact that oxidized, tungsten contacts are far superior. Back in the old days, we use to use mercury for reliability, or sliding contacts that offered rubbing the contacts together. Old telephone exchanges used thousands of relays, had a 24 hour crew to constantly burnish relay contacts. Those guys had to find new jobs when solid state took over, a transistor could be switch trillions of times without damage.

Honeywell made a very reliable thermostat using mercury wetted contacts but mercury is now outlawed. Many Fords used transistors for energizing the clutch coil, but would have been far more reliable is they spend a penny more for a transistor with a decent rating.

I am from the old school, have an electric burnishing tool, just pop off the cap, clean the contacts and good for at least another thousand operations. Your relay did make it past the warranty period, that is all that counts.

Back to Automotive Air Conditioning Forum

We've updated our forums!
Click here to visit the new forum

Archive Home

Copyright © 2016 Arizona Mobile Air Inc.