COMPLETE A/C Diagnosis

Old 05-10-2007, 04:09 PM
Rennlist Member

Thread Starter
BS911's Avatar
Join Date: May 2001
Location: Williamsport, MD
Posts: 828
Default COMPLETE A/C Diagnosis

Ok, I'm almost done with my complete write up on diagnosing the 993 A/C system from top to bottom. I think I have covered everything, but before it gets pushed to p-car, I'm running it by here first in case anyone has any suggestions for changes. Sorry if I jump all over the place... this is a tough one to write sequentially. Also, there will be some better formatting in the final version. Later, I will add a heating specific section, but this is primarily about A/C. Hope this information is helpful as we start into the A/C season...


An important note before reading. This is not intended to be a complete step by step troubleshooting and repair guide. Instead, this is to describe the operation of the system and understand what components perform certain functions within the system. Many components can be easily checked and verified, but not all. You may be able to use this to isolate a very potential component, but not necessarily walk through from a to z and locate the exact problem. Most importantly, Porsche dealers have proven time and time again that they love to tell customers to plunk down $1500 to $2000 on a new CCU. Even according to the Porsche manuals themselves, that is the LAST step in troubleshooting an HVAC problem on a 993 – or any car for that matter. Even if you don’t plan to do the actual work yourself, using this information will hopefully help you know where your problem is before your dealer tries to relieve you of your hard earned money.

Before we start, we need a baseline of understanding on what mode of operation we should be in to properly diagnose the A/C system on a 993. First, obviously, the engine must be running. On the CCU the temp **** must be on the “blue dot,” the footwell and windshield vents closed, the recirc and defrost buttons turned off, the fan on high, and the max A/C button engaged. For further reference, the max A/C button is simply a one stop operation to perform the following: Turns on A/C, close defrost flap, close footwell flap, open center nozzle, set temp max cold, and set blower to speed 4. Even though this is redundant, it is a good idea to make sure you have all settings down to the lowest common denominator first to rule out electrical problems.

One more point is that the proper cooling temperatures and refrigerant system pressures can only be measured with the engine running at 2000 rpm’s. This is because the compressor is designed to work at optimum levels above idle speeds only. With the compressor running at idle speeds, you aren’t getting much of a compression cycle in the refrigerant.

I have broken down the A/C diagnosis into 2 main problem categories. One where the compressor is actually running and one where it does not run at all. With the A/C running for at least 10 seconds, you should see the outer most wheel on the compressor (the clutch) spinning with the belt. If the compressor is running, skip ahead to the Poor A/C – Compressor Running section. If the compressor is cycling on and off, continue to the bottom of this section regarding the 3-Level Pressure Switch.

With all that being said, start your A/C and see what you have to work with.

No A/C – Compressor Not Running or cycling on and off.
There are only a few things that will cause your compressor to not run when you have the Max A/C button engaged. Some are upstream and some are downstream of the CCU. By upstream, I mean inputs to the CCU that cause the CCU to determine to not engage the compressor. Downstream means the CCU thinks it has turned on the compressor, but the signal just hasn’t gotten there.


Compressor power lead – engine compartment box passenger side
This is the easiest potential problem to fix. The compressor gets ground from its connection to the mounting brackets. There is one wire leaving the compressor into a small box right next to it on the side of the engine bay. It is very easy for that wire to come unplugged. Follow that wire and make sure it is connected. For further peace of mind, check for 12V at that wire connection. If 12V is present, then the clutch should engage. If not, keep going.

Compressor fuse – engine compartment fuse box driver side.
There is a 7.5a fuse in the fuse box that passes on to the compressor lead. Verify that fuse is intact.

Compressor relay – engine compartment fuse box driver side.
The relay engages with the lead from the fuse to pass current to the compressor relay.

Fog Light relay – front trunk fuse box
It has been reported that the fog light relay is in a daisy chain config to the compressor relay. If that relay fails, no power will pass to the compressor relay. To verify your relays, you can switch with other relays that have the same part number.


Evaporator Temperature Sensor – 3.3 to 7 kOhms – G22 wt/bkThe evaporator temperature sensor is there to tell the CCU when the evaporator has frozen. When the evaporator freezes, the condensation will block airflow and reduce cooling. When this condition is reached, the CCU shuts of the compressor to allow the ice to melt off. Once the temperature is back to normal levels, the compressor will engage. During a compressor off scenario due to an evaporator sensor signal, all other A/C functions continue to operate. (Fans blowing, condenser fan running, etc.) To verify the evaporator sensor, remove the CCU from the dash. Disconnect the “G” cable from the back of the CCU (the longer, top connector) and read the resistance between G18 and G22. (Use the imprints on the cable and back of CCU connector for pin reference.) You should see a reading within the range of 3.3 kOhms to 7 kOhms. Any reading outside this range, especially open or short to ground readings, indicate a failed evaporator sensor. Replace the sensor. For a temporary fix, cut out the sensor from the wire lead under the front hood. Splice in a resistor that has a resistance within the range listed above. This will fool the CCU into thinking that the system is within tolerance.

Oil Cooler Temperature Sensor – 22.6 to 3.8 KOhms – G12 gn/bkExcessive oil temperatures read at the oil cooler will also cause the CCU to disengage the compressor to reduce load on the engine. Test resistance from G12 to G18 as above, looking for faulty readings like an open circuit.

Sensor Ground – G18br/bl to ground
All sensors get ground from G18. Verify that G18 has a strong connection to ground. A bad ground on G18 can cause a host of problems.

3-Level Pressure Switch
This switch, inline to the evaporator under the front hood on the driver’s side near the windshield, tells the CCU when the refrigerant high side pressure is either too high or too low. Between 2.5 and 27 bar (37 to 390 psi) the high/low switch passes current to the compressor. Anything outside this reading will disengage the compressor. As noted, this switch is downstream of the CCU, meaning the CCU has no real knowledge of the shutoff.

Additionally, the switch controls the condenser fan motor. When the A/C is turned on, the low speed condenser fan is engaged. You should feel and hear the fan running in front of the driver’s front wheel. Air exits under the bumper. When the mean effective pressure in the high side of the refrigerant system reaches 17.5 bar (255 psi) the system bypasses the condenser fan ballast resistor, causing the fan to run at higher speeds to help reduce system pressure.

To diagnose this switch, for the purposes of testing to locate a failure of the compressor to engage, use a manifold gauge set to check your high side refrigerant pressure. It is highly unlikely that with the compressor off, your high side pressure would be too high as the system would equalize without the compressor running. You would be looking for excessively low pressure in the system. Don’t just add refrigerant if you suspect this to be the case, verify your pressure first. See below for checking refrigerant levels for more information.

What happens on a cycling compressor. At rest, the refrigerant in the system is not being compressed and pushed through the lines so the system pressure equalizes at … let’s say, 80 psi. (Just for illustration here, not exact.) The A/C tells the compressor to engage and the output side (high side) of the compressor raises in pressure while the low side pulls down to a low pressure. This pressure change is what causes the coolant to go through its liquid and gas stages – getting that wonderful cooling effect. If the pressure climbs too high – or drops too low, the pressure switch interrupts the current to the compressor.

With all of the above met, hopefully the compressor is engaging at this point. Next comes determining if the output (air at the cabin vents) is cool enough.

Poor A/C – Compressor Running

Assuming at this point that the compressor is running, there are still several factors that could affect the output temperature at the vent. Once again though, we need to reach a baseline for what is “good” performance on a 993 system. Using the chart below, under the proper running cycle (Max A/C, engine at 2000 rpm’s, at least a couple minutes) your vent temperature at the center vent should reach the temperature in the chart correlating to the ambient temperatures. The system won’t cool as well on a 100 degree day as it would on a 60 degree day.

For example, at 86 degree ambient temperatures, your center vent nozzles should blow between 44 and 50 degrees. Anything within this spec is considered good. If you aren’t happy with those temperatures, it may be possible to improve slightly but not by much.

There are several things that affect the vent temperature once the compressor is engaged.

Faulty hot air flap valve
Hot air from the engine is routed to the front of the car via ducts that get shut off by electric servo motors in the footwells in the passenger cabin. If a motor fails (which is common) then too much hot air will be fed into the system, causing the ambient air that the system is trying to cool to be artificially raised. Under the kick panels by the driver’s and passenger’s side foot well, (remove a couple screws in the carpeted panel) check the operation of the motor attached to the paper duct to verify it closes completely.

Poor condenser function / condenser blower or ballast failure
If the condenser is not effectively cooling the liquid, then it will have a hard time converting it back to gas in the evaporator. Things to check on your condenser…

First, are leaves and debris blocking airflow? If you shine a flashlight into the front bumper looking back towards the condenser by the front driver’s side wheel you can inspect the inlet side of the condenser. (It looks like a radiator.) Poor airflow is a killer of condenser function – so much that even following other cars closely on the highway will reduce your airflow and cause vent temps to go up. Clean the condenser with a hose if necessary.

Is the condenser fan running? With the A/C on, you should hear an electric fan running by the condenser and also feel air blowing out underneath the car. If you have no condenser fan running, then you have either a blown fuse, relay, or most likely, ballast resistor. Check the fuses and relay in the front fuse panel and swap the relay with the oil cooler relay to verify operation. If the ballast resistor that is mounted next to the condenser inside the front driver’s side wheel well is blown, then you will have no low speed condenser, but it will kick on high speed when the high pressure level is reached. See other DIY’s on checking/replacing ballast resistors.

Once your condenser fan is running, you shouldn’t have a low speed ok/high speed failed problem. It is only the low speed that usually will fail. If low speed works, then high speed should too. When the refrigerant pressure gets too high, the CCU bypasses the resistor and activates the “un-resisted” power lead to the fan, causing it to run in full speed.

There are other problems leading to poor condenser function covered below in “other refrigerant system problems.”

Poor airflow over evaporator
Just like the condenser needs airflow to cool the liquid refrigerant, the air must pass over the evaporator to get the heat pulled out of it. This usually isn’t a huge problem because if the air flows too slow over the evaporator, then that air just cools off even more than if it blasted right by. This is where many aftermarket A/C parts suppliers exaggerate their claims. You see places showing 32 degree vent temperatures, but what they don’t show you is that they have prolonged the airflow cooling cycle by … turning the interior fan speed down. That is cheating. Sure, you can get colder air, but at smaller volumes therefore taking longer to cool your car.

The biggest culprit of poor airflow over the evaporator is a failure in the fresh air and recirculated air control flaps. Either the outside vent that allows fresh air in does not close properly (malfunctioning servo motor) or the inside cabin air intake does not open properly. (Malfunctioning vacuum supply.) See the section below on “Troubleshooting inside airflow issues.”

Incorrect refrigerant levels
This is the biggie… incorrect refrigerant levels. A 993 uses very low system pressures at the compressor compared to “normal” cars. So the standard rule of thumb to simply “throw in a can of r-134” is very bad advice in a 993. Even if you are a little low on refrigerant. A full can – even a small can – will put your system way too high. You would be surprised how much even a little bit too much refrigerant can hurt your vent temperatures. This makes using the cheap gauges that come with refill kits poor for 993’s. If you do use those single/low side gauges, you can get a good idea of what your level is, but just use the chart above to make sure you have the right levels, and not the chart on the gauge itself. Again, working off of an 86 degree ambient temperature, you should only read 10 to 17 psi of low side pressure at 2000 rpm’s.

I’m not going to get into rights and wrongs of adding refrigerant to your car. I will say this though… make sure you follow the warnings and directions on the can, refill kit, and car. You only ever fill on the LOW side port of the compressor. A 993 system also says to only be filled by weight. That would require a complete vacuum purge and refill by weight quantity to do it properly. While that may be the recommended way, I have had good results adjusting pressure with the refill cans found at the local auto parts store.

Again, you really should use a manifold gauge set like the ones sold at or your local auto parts store. With those, you can read both high and low side pressures at the same time. See the next section for what else to do with those numbers. But for now, we’ll proceed with the proper pressure levels.

At 2000 rpm’s with the compressor running, observe your low side (and high side if you’ve got it) pressure. Add or remove refrigerant until you get your levels back within spec according to the chart above. It also helps if you have a thermometer hanging in your center vent taking temperatures at the same time. Ideally, you should be able to get both the low and high side to match the specifications in the chart. The next section will tell you what to do if they aren’t both in range, and illustrate why you really need both sides measured. Assuming you have no other system problems, then proper pressures should yield proper vent temperatures

Other refrigerant system problems
Once the compressor is running, the condenser is getting its clean airflow, the evaporator is getting air over it, then we are down to just the internal guts of the refrigerant system if things still aren’t working up to spec. This is where you must know both the high and low side pressure. Generally, the whole loop of the refrigerant cycle has only one “gate” at the expansion valve. With the compressor off, everything equalizes. The expansion valve will not allow the liquid to flow fast enough, therefore a pressure builds up against the compressor. That is the only place you should have resistance in the lines. Any other blockages will cause poor performance. Following is a general rule to determine where within a refrigerant system you may have a faulty component.

If your pressure is the same on both high and low sides, your compressor is bad. It should be pushing up the high side and drawing down the low side. Incidentally, the compressor, when off, will let both sides almost equalize at about 75-80 psi on both high and low sides so failure to compress would show little change in those numbers.

If the low side reads negative pressure – a vacuum – and the high side is normal, the expansion valve is stuck in the closed position or clogged or there could be a restriction in the liquid line. This could be what your mechanic unclogged Scott.

If your low pressure side reads high and the high side reads normal, then the expansion valve is stuck open.

If both the high and low side are high, you are overcharged or the condenser isn't effective or there is a restriction.

If the low side is normal but the high side too high, then there also could be an overcharge or restriction in the condenser or a clogged receiver drier.

There are a lot of and/or’s in those troubleshooting steps. You may get a good idea from your readings on what to look at next, but if you have high and low side readings that don’t properly correlate, then you may very well be in for a part replacement that requires opening the refrigerant system. This also shows why not knowing what the high side pressure is can hamper your troubleshooting efforts. It is ok if the system is working properly, but if something is amiss, then you don’t have the information you need. In that event, most everyone will be taking their car to a shop for service.


It is highly likely that you just read all of this and feel no closer to solving your problem. In many cases, I think this information can lead you to a faulty component. In others it merely gives you an idea of what type of problem you may be looking at , but not exactly where. At least with this information you can hopefully save your shop some labor hours, or prove/disprove their recommendations.

Further Information

Basic r-134 refrigerant cycle.

The refrigerant leaves the compressor as a high pressure, high temperature gas as a result of the compression. The condenser cools the refrigerant and causes it to turn into a liquid again, but still under high pressure. The liquid now has to pass through the expansion valve which has a tiny port hole. This causes the pressure on the output side to drop. That low pressure liquid enters the evaporator where it begins turning into a gas again at low pressure. The conversion to gas extracts heat from the air surrounding the molecules. The fins on the evaporator help this process by allowing the air passing over the fins donate their heat to the liquid to gas conversion process, effectively cooling the air passing over the fins. Once it leaves the evaporator, it is a low pressure, low temperature gas. It then routes back to the compressor to start all over again.

Troubleshooting inside airflow issues.

Normal operation:
- Recirc: Vacuum pressure pulls opens inside air source / Servo Closes fresh air source
- Fresh: Vacuum releases & spring loaded inside air source closes / Servo Opens fresh air source
The problem is that the electric servo could die anywhere - fresh air source open, closed, or somewhere in between. So the only real thing we are guaranteed to know is if the vacuum fails.
If vacuum fails, and servo still operates:
- Fresh air selection: servo opens fresh source, inside air source closed. NORMAL
- Recirc selection: servo closes fresh source and inside air remains closed. NO AIR FLOW.
If vacuum works and servo fails with fresh air closed:
- Fresh air selection: servo remains closed, inside air closed. NO AIR FLOW.
- Recirc selection: servo closed, inside air opens. NORMAL.
If vacuum works and servo fails with fresh air open:
- Fresh air selection: servo remains open, inside air closed. NORMAL.
- Recirc selection: servo remains open, inside air opens. MAY APPEAR NORMAL, but can't shut off outside air.
So to sum up:
- No air flow/Wheeze on recirc - you have a vacuum leak somewhere:
- No air flow/Wheeze on fresh - your fresh air electric servo has died in the closed position.
To verify electric servo:
- Remove black plastic trim under front hood nearest to the base of windshield and locate center electric servo. (There are 3 servos on a 993, one to open/close footwell output, one to open close windshield output, one to open close fresh air input.) 964's have 2 more to control temp mixers. 993 temp mixer servos are in cabin under door sill trim area.
- Operating the corresponding a/c control should show servos moving back and forth.
To verify vacuum -
- Main vacuum "source" from engine routes to same location as the 3 servos, into an electric solenoid. When the recirc button is pressed, the electric solenoid opens the valve, passing vacuum pressure to output. With engine idling:
- Disconnect vacuum source on solenoid to verify vacuum pressure. If not present, problem is vacuum leak in engine area. (Good luck)
- If vacuum present, press recirc, disconnect output side of solenoid and verify vacuum pressure.
- If not, solenoid not getting signal or failed.
- If present, look under dash, under ccu for rubber elbow connecting vacuum line to flap mechanism - it would most likely be unplugged due to stereo installer or other inside work bumping it free.
BS911 is offline  
Old 05-10-2007, 05:51 PM
Rennlist Member
epj993's Avatar
Join Date: Aug 2005
Location: Huntington Beach, CA
Posts: 1,352

Awesome Brian!
epj993 is offline  
Old 05-10-2007, 06:08 PM
993_Pilot's Avatar
Join Date: Apr 2006
Location: somewhere on a twisty road...
Posts: 1,593

Perfect timing... Mine isn't blowing very cold when I tried it yesterday. Thanks!
993_Pilot is offline  
Old 05-10-2007, 06:20 PM
Don Magee
Rennlist Member

Join Date: Jun 2001
Location: Austin, Texas
Posts: 621

Outstanding effort.

Maybe I missed it in the "inside airflow issues" section, but I have the fan speed vary from low to high randomly occasionly. For example if the ac is on and the fan speed is at the second mark, all of a sudden the fan goes on high for a few moments without me touching anything. Curiously, it does not do it if the recirc button is activated. Any thoughts?
In any event, this is a very impressive post.
Don Magee is offline  
Old 05-10-2007, 07:16 PM
Rennlist Member

vuugti's Avatar
Join Date: Sep 2003
Location: Home of Little League!
Posts: 413

Thanks for wrapping this up into one page. Saves a ton of trouble.

Nicely done as well.

vuugti is offline  
Old 05-10-2007, 07:19 PM
Rennlist Member
macanswer's Avatar
Join Date: Dec 2004
Location: Chambersburg, PA
Posts: 89


Well, this is great! Guess that puts the keebosh on the draft I was writing!

I took quite a few pictures as I was troubleshooting my system last weekend. I would be happy to contribute those to your effort if you think they would be useful. I have pictures of removing the CCU, the back of the CCU, measuring resistance across pins 18 and 22 to check the operation of the temp sensor at the evaporator (I think I'm stating that correctly) and several others.

Thanks so much for taking this on and let me know if I can help!
macanswer is offline  
Old 05-10-2007, 07:26 PM
Air Kuul TT
Super User
Air Kuul TT's Avatar
Join Date: Sep 2004
Location: McKinney, TEXAS
Posts: 3,753


Thanks for taking the time and effort on the write up... nice work.
Air Kuul TT is offline  
Old 05-10-2007, 09:22 PM
Rennlist Member

Thread Starter
BS911's Avatar
Join Date: May 2001
Location: Williamsport, MD
Posts: 828

Thanks all.

Scott- that would be great. We could co-contribute this one! If you'd like to add/revise/whatever that would be just fine with me. Also, pics would be outstanding. I am in the process of moving in the next few weeks so taking pictures (or taking time for pictures) are out of the question for me. But this is something I've been collecting and working on for the past year or so so wanted to get it out before it got too hot still.
BS911 is offline  
Old 05-10-2007, 09:49 PM
Rennlist Member
AOW162435's Avatar
Join Date: Jul 2005
Location: Ellicott City, MD
Posts: 15,791

That is a wonderful write-up and will undoubtedly prove very useful for many of us in the future. Thank you!

AOW162435 is offline  
Old 05-10-2007, 09:51 PM
Land Jet
Rennlist Member
Land Jet's Avatar
Join Date: Mar 2007
Location: Maryland
Posts: 4,210


Thanks so very much for this resource. I am tracking an A/C gremlin myself and was even researching aftermarket parts. Any thoughts on that?
macanswer, IMO, pictures could only be a plus for us more inexperienced Porsche owners.It would be great if you two could coordinate for the optimal results.
Land Jet is offline  
Old 05-10-2007, 10:36 PM
colo993's Avatar
Join Date: Jun 2006
Location: Colorado
Posts: 87


You rock!

All my AC is doing is blowing warm air. When it was working last summer, I could clearly hear the compressor running. I'm not hearing anything now when the AC is turned on. I'll be using your guide this weekend to check it out.


colo993 is offline  
Old 05-10-2007, 11:15 PM
baron95993's Avatar
Join Date: Feb 2007
Location: Tyler Texas
Posts: 525

Nice work my air has been out for a few months but it is getting hot here in Texas and need to get it fixed. Thanks for putting this together you are a life savor.
baron95993 is offline  
Old 05-11-2007, 01:45 PM
Join Date: Mar 2007
Location: Bristol, UK
Posts: 205

Great work Brian!

"Oil Cooler Temperature Sensor – 22.6 to 3.8 KOhms – G12 gn/bkExcessive ..."

do you mean 22.6 Ohms to 3.8 kOhms?
leeshephard is offline  
Old 05-11-2007, 02:09 PM
Rennlist Member

Thread Starter
BS911's Avatar
Join Date: May 2001
Location: Williamsport, MD
Posts: 828

The schematics show :
20 degrees C : 22.6 kOhms
60 degrees C : 3.8 kOhms

I'm actually not 100% sure if the oil cooler temp sensor will cause a shutoff of the compressor or not. I seem to recall that info from some time before, but haven't been able to locate my notes on it. Worth a look anyway.
BS911 is offline  
Old 05-13-2007, 01:06 AM
Coleman's Avatar
Join Date: Jul 2002
Location: Cape Neddick Me.
Posts: 1,199

Brian, Obviously you went to ITT Technical school for HVAC and Thermodynamics!!! ;-)

What would you guess as the likely culprit of a slow leak, requiring the system to be refilled every year, on a 95?


Coleman is offline  

Thread Tools
Search this Thread
Quick Reply: COMPLETE A/C Diagnosis

Contact Us - About Us - Advertising - Cookie Policy - Privacy Statement - Terms of Service

© 2019 MH Sub I, LLC dba Internet Brands

We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to and affiliated sites.
  • Ask a Question
    Get answers from community experts
Question Title:
Your question will be posted in: