I have a complex AC problem in my Civic. I need to tell you the story so that you might get a potential diagnosis of the issue:

  1. Car started having problems with AC in 2015. After driving for a while, the evaporator would freeze (blowing ice pieces to your face), then no longer would get cool air (after ice clogged evaporator completely).
  2. Once that happened, I would turn off the AC and keep fan on to melt ice, then turn AC back on... I kept doing this for most of 2015 until...
  3. On 2016 it finally died completely. It would not engage, the clutch would not kick in (compressor not spinning), and obviously I would not get any cool air.
  4. Took it to a mechanic, they said it was a bad compressor, replaced the compressor and recharged the system, and $1300.00 later I was happy with my AC working and cooling fine, but...
  5. I didn't use car for 3 days, on 4th day I start the car and guess what? No AC, no clutch engaging, no cool air. I took it to the same mechanic (they have a 1-year warranty on the work...) and they told me they were not able to figure out the problem... and to bring it back other day and leave it. Fine, I haven't brought it back until I get more information myself. Somehow I suspect they didn't do a right diagnostic the first time, and perhaps I paid $1300 to change a compressor that was not really the root cause!
  6. I take it to a HONDA dealership, they say is the ECM/computer... they come up with some weird story that there is oil in the ECM connector and the computer is leaking oil? Seriously? I don't buy that crap. So I start doing my investigation and reading about AC myself. Of course, the dealership wanted to charge me $900 to replace the ECM which scared me away immediately.

Here's my investigation after reading about Civic 2007 electrical circuits, AC, sensors and so forth:

  1. All fuses are OK and working. I tested every single one in the AC circuits.
  2. The relay that kicks in the compressor/clutch is fine. I tested it outside the car, using power-supply and multimeter.
  3. The clutch is not the problem, I manually verified that it engages. Now, when I jumper the high-current terminals to force the clutch/compressor, it almost makes the engine stall. It sounds like the compressor cannot add any more pressure into the system, or there is some obstruction (at least that's my theory). Another possibility is that the new compressor is broken as well? Who knows.
  4. FYI, the relay that kicks in the AC is controlled by a ground enabled by the ECM/computer. That signal is never received as the clutch never engages, the relay is good, and I verified continuity from the relay terminal to the ECM pin. Apparently the ECM is simply protecting the system due to input from some other sensor, perhaps that the pressure is too hight in the system?
  5. I also need to mention that when you start the AC in the dashboard panel, the cabin blower starts as expected. However, the AC condenser and radiator fans are not engaged (both should come on), instead they remain off (again, this should also be that the ECM is not sending the signal).
  6. With the car completely off, the pressure readings I get are 113 PSI(low side), 110 PSI (high side) which I really don't understand much. I can't read any dynamic pressure as the compressor is not kicking in as mentioned.

Update 8/23: The cabin self-diagnostic function (tests control unit, blower, air mix and mode motors, evaporator temp sensor) detected no issues. So I guess that rules out any problem with the control unit or evaporator temperature sensor.

Now, having said all this, what is the most likely root cause of the problem? Bad pressure sensor? Bad evaporator temperature sensor? Problem in the evaporator valve or orifice tube? Any way to check the previous components?

Now, there is a remote possibility that indeed it is the computer. Is there any way to check if that's the case?

  • 2
    if the ecm isnt sending the ground signal to the relay it could be a bad switch, sensor, multiplex unit or ac control unit. if you have the wiring diagram work your way from the multiplex unit back. iirc the switches feed a ground signal to the multiplex unit and the multiplex sends a signal to the ecm via a dataline. – Ben Aug 23 '16 at 11:35

In short: It was the ECM/PCM.

As much as I diagnosed the electrical circuits, I think I was just trying to overlook the elephant in the room (broken ECM) for fear of spending $$$.

Based on the troubleshooting steps described in the Honda Civic Service guide, I convinced myself that clearly the ECM ALWAYS HAS TO GROUND PIN A14 (ECM connector A, pin 14) going to the compressor clutch relay AS SOON AS YOU TURN ON THE AC. Period.

The troubleshooting guide even shortcuts the diagnosis steps and tells you right after not seeing ground in A14 to "Replace your ECM" !!!

I took it to the dealership and they replaced the ECM and voila! AC is working again without issues.

So, this was made out of love for Civics, really... I have spent in this car $2000, first b/c the compressor broke, then for mysterious reasons the ECM broke right after... bad luck eh? Regardless, car is up and working again.


One of the symptoms you mentioned is evaporator coil freeze up is caused by restricted air flow. If your vehicle has a cabin filter change it. When the air flow is restricted the temperature around the evaporator drops and ice forms.

Another cause of freeze up is low refrigerant which would explain your high low pressures being almost equal. When the refrigerant in the system is low due to leakage over time, there will be a decrease in the saturation point temperature of the system. When this happens, it will take a lower temperature for all the refrigerant liquid that has been compressed by the compressor to become vapor in the evaporator. You probably have a leak in the joints in the a/c tubing.

This means that the operating temperature of the cooling coils has now dropped compared to the original design. This lower temperature will cause moisture collected from the air to quickly becomes ice on the evaporator when the temperature goes below below 0 C (32 F).

One of the reasons I am speculating regarding the fan not cycling is related to both temperature and system pressure being insufficient to be triggered.

The ECM and compressor are probably fine. Rotary compressors are for the most part are virtually indestructible. If you want to defrost the evaporator coil turn the heat on and make sure the condensate tube is clear so the water can drain out.

EDIT: Additional info

The source of your problem is either the orifice tube (located at the inlet of the evaporator coil or the thermal expansion valve. Both of these devices do the same thing a different way. If the system uses an orifice tube, there will be an accumulator between the evaporator and the compressor. An orifice tube sometimes lets too much refrigerant into the evaporator and it doesn't all boil. Since the compressor cannot compress liquid, only gas, the accumulator traps any excess liquid before it can get into the compressor. With an orifice tube, the flow of refrigerant through the valve (tube) is constant. Overall flow throughout the A/C system is controlled by cycling the A/C system compressor on and off as required. The end result remains the same; there is a controlled amount of refrigerant entering into the evaporator.
The biggest problem with orifice tubes is that they tend to easily get plugged with debris that may be traveling throughout the A/C system. As soon as the flow of refrigerant through an orifice tube is restricted, the flow of oil (which travels throughout the A/C system in the refrigerant) is also restricted. That becomes the start of catastrophic compressor failure.

Or your ECM is toast :)

  • "One of the reasons I am speculating regarding the fan not cycling is related to both temperature and system pressure being insufficient to be triggered." Yes, but the static pressure (car turned off) is around 110 PSI, and that should be sufficient for the ECM to trigger the compressor. Based on the troubleshooting guide, the ECM turns off compressor if the pressure sensor has readings less than 28 psi or above 455 psi. So I would think 110 PSI is a fine value to have when the car is off. – Joseph On Cars Aug 23 '16 at 15:06
  • Furthermore, I did the cabin self-diagnostic function which should detect any problem with the evaporator temperature sensor. It found no problems, so that's one thing less. Remaining steps are to work my way little by little from the MICU and HVAC controls to the ECM. I will follow all electrical troubleshooting items before claiming is the ECM. However, based on all I've seen so far everything points to a defective ECM not sending any HVAC signal to the relay. – Joseph On Cars Aug 23 '16 at 15:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.