'06 Z4 N52 A/C exceptionally weird issue

[bmwdirtracer]

Hi friends,

Thought I'd open this up to all the excellent minds here.

Car is an '06 Z4 N52 engine, automatic, ~90k miles. It's owned by a really gorgeous hottie.

Problem is that the A/C compressor will not activate until the engine exceeds 1200 rpm.

The following steps have been taken:
(another tech started here, and didn't realize the issue, so replaced a couple of things that didn't need replacing)

Expansion valve and dryer have been replaced, along with the pressure switch/sensor.
The IHKA control / head unit has been replaced with used.
The replacement IHKA was correctly coded by ISTA.
The DME received the newest software, via ISTA.
The system is perfectly charged with exactly the correct amount of R134a, and when the engine is held at 1200 rpm and above, the vent temps are 40 degrees or less, and the high and low side pressures are perfect.
The electric fan works perfectly, whenever the compressor is on, or when commanded by any of the diag computers.
The compressor activates perfectly, via any of the 3 diag computers, at any rpm.

Note please that none of these steps has had any result, other than to make the vent temps colder when the system was evacuated, dryer changed, and refilled precisely.

There are NO codes, in any computer.

Anyone here experienced anything like this before?

Many thanks,
Chris

[02Pilot]

I haven't been inside one of these A/C compressors, but the first thing that comes to mind is that the intake side reed valve is not sealing until enough pressure builds to push it shut. Of course, being that there may not even be a reed valve in these compressors, I could be completely wrong, but it sure seems like problem is related to the compressor not compressing until there's more activity in there than the spec requires, which suggests some sort of internal sealing problem.

[bmwdirtracer]

I believe that the issue is a computer problem, my friend.

I can activate the compressor through any of 3 diag tools, and the pressures follow suit. The pressures aren't changing until after the compressor activates, which happens at 1200 rpm, when the DME sends the "okay to activate" signal to the IHKA.

Trouble is, I can't figure out why the signal's not being sent. I gave the newest software to the DME, and replaced every item which even MIGHT cause this issue...I think?

Admittedly, I have not replaced the fan, or the compressor, but these are expensive parts to throw, when there doesn't seem to be any diagnostic imperative to do so?

Please correct me if I'm wrong; but I see the compressor and fan working when commanded, and the pressures don't change 'til the compressor activates, so how can a reed valve be the cause? Just trying to avoid throwing more parts....besides, I'm a softie for hot women, what can I say?

[02Pilot]

My brain is stuck in an earlier time, so it immediately defaults to mechanical. I'm probably not correctly understanding exactly what the problem is, but if the signal has to be sent from the compressor that it's OK to go, and it's RPM-dependent, then the first question is what sort of signals are possible. I assumed that pressure was the only monitored variable at that location, but perhaps there are others? Got a junk compressor lying around that you can take apart?

[bmwdirtracer]

The DME sends the "okay to activate" signal to the IHKA (control head). Then the IHKA sends power to the compressor. The only sensors and switches I know of are (1) the pressure switch at the dryer (swapped), (2) the evaporator temp sensor (reading correctly), and (3) the ambient temp sensor (also reading correctly).

The sensors which report to the DME (ECT and outlet temp) are reading correctly.

https://screenshots.firefox.com/xpUY....prodemand.com

https://screenshots.firefox.com/SvsQ....prodemand.com

[02Pilot]

OK, slowly I'm wrapping my brain around this. So what are the conditions for the DME to send the Go signal to the IHKA? Is it just those three sensors you mentioned? If so, RPM doesn't seem like it should be a factor. Total long shot, but is the alternator doing its job properly? In other words, could it be voltage related?

Photos of the customer might help to motivate further thought on the matter.

[ross1]

are you able to monitor if the DME is sending the signal, which would implicate the IHKA if so?
Is there supposed to be any RPM threshold for engagement or is it other criteria?
Oh, customer hotness has been mentioned twice. Once more and we'll need pictures.

[bmwdirtracer]

Thanks for your consideration of this issue, guys!

I have not been able to "see" the DME's signal. using any of my 3 devices (ISTA, Autologic, Launch). I only know it's there by looking at the wiring diagrams. However, the IHKA has been swapped, and the replacement coded successfully, and the fault remains unchanged, so I have to believe that the problem isn't with the IHKA control unit.

I also updated the DME software with ISTA, and again, there was no change. And since the system works PERFECTLY, ice cold, whenever the engine passes 1200 rpm, I find it hard to believe it's a mechanical issue. The compressor just doesn't switch on, until that engine speed. But WHY? As far as I know, engine speed is not a parameter for a/c, on any car.

Note, though, that I can use the diag computers to turn on the compressor at idle (eight seconds at a time, because that's the test period). Also: this activation is done via the IHKA, NOT the DME.

Hmmmmm...."Miss hottie customer, I've been asking for help on diagnosing your A/C, and the guys really want to see that you're as gorgeous as I said. Can I take a picture of you and post it?"

[02Pilot]

Is there any variance in the high and low side pressures between "normally activated compressor above 1200RPM" and "test activated compressor below 1200RPM"?

What are the states of the battery and charging system? Given the advanced charging plan in these cars, I wonder if a weak battery could signal the DME not to run the A/C until the alternator is producing sufficient voltage.

[bmwdirtracer]

Regarding your first question, it's impossible to give you an accurate answer, because each given "test sequence" only operates the compressor for 8 seconds max, then it takes at least 4 seconds to activate another test sequence. However, by repeatedly activating the testing as quickly as possible, near-standard pressures were attained on both low and high sides.

The '06 Z4 is a little bit old school, in spite of the N52. It's not really driven to the far edge of charging strategies, like, say, an N63 vehicle. I've never found an N52 engined car that's not charging properly, at idle, but I admit that I have not checked that on this car -- and I WILL check that first thing Monday morning. Thanks for that idea ! Easiest way to test that, besides a multimeter, will be to hook up a charger, and see if the compressor turns on below 1200 rpm. (I'll do both tests, of course.)

[dworthy]

One thing I do know is that on the E-46's if you don't have the correct outside air temp, it can mess with the IHKA. I am sure the Z-4 has one as it is based on the E-46, have a look under the front of the car to see how much damage has been done.

Just a thought.

[bmwdirtracer]

Ambient temp sensor reads perfectly.

[bmwdirtracer]

Just checked voltages, alternator charging just fine at idle....I'm out of ideas....

[ross1]

Seems to me any investigation into pressures is a red herring as the compressor is inactive hence pressures don't change with RPM and the charge is proven correct.
Has to be an electronic fault. I'd like to watch voltage as the RPM crosses 1200.
This car isn't looking for any vacuum signal is it?

Oh, "Miss hottie customer" that's three!

EDIT: All connections checked? I was thinking perhaps some movement in the engine bay causing an intermittent connection coincidental with 1200 RPM.

[bmwdirtracer]

Well...I know what the problem is now, but I haven't a clue how to fix it. Here's a link to the computer data line wiring diagram, to help what I'm saying make sense:

https://screenshots.firefox.com/PY5L....prodemand.com

The problem is that the IHKA (A/C module) thinks that the engine is revving 300 rpm, when it's at normal idle speed of ~700 rpm. All other computers which use engine speed show that speed correctly....including the KOMBI (cluster).

I mention the KOMBI, because it is the Gateway module which joins the CANbus (powertrain stuff) to the Kbus (body stuff...yes, I oversimplify here).

The IHKA is on the Kbus, and gets its engine speed data from the KOMBI. Now, the problem is that this is (as Abel reminded me!) a digital signal, not analog, so the problem can't be faulty wiring with too much resistance. It also can't be a default value, because the reading goes up corresponding to the engine's real rpm. When the REAL rpm reads 1200, the IHKA shows engine rpm as ~650, and the compressor starts!

Therefore, the IHKA isn't starting the compressor until it thinks the engine is turning fast enough to carry the load.

I recoded the entire car with ISTA today, and it made no difference. (The DME was given the latest software, and the replacement IHKA was coded Friday)

The problem can't be the IHKA, because it was swapped, and the issue remained unchanged.

Therefore, the problem, in my estimation, has to be within the KOMBI....and it's going to be difficult to convince the owner to buy a brand new cluster, in order to make their A/C work at idle....


Pretty strange, huh?

[328 Power 04]

I am wanting to not believe that the modules are talking to each other via K-line for live rpm data... K line is diag line for your computer, but not for inter-module communication (**actually partially wrong, modules do send out telegrams to each other). Live constant telegrams would slow down the k-bus line.

It wouldn't explain the half rpm deal though. I could be wrong.

Alternate wiring diagram:

Screenshot (7457).png

- - - Updated - - -

Evaporator control is activated by pressing the A/C button in the operating unit. A/C standby mode is indicated by the LED in the A/C button. 

Actuating the A/C button initiates the following steps:

- The DME is requested to increase the idle speed to ensure the A/C compressor already generates sufficient refrigerating capacity with the engine idling.
 
- Following the ”switch on A/C compressor” request from the control panel/control unit, the DME responds with an enable signal (Pin 15).
 
- Consequently, the control panel/control unit activates the magnetic clutch for the A/C compressor.
 

The following information is transmitted in the form of a telegram on the K-bus to the instrument cluster and from here via the PT-CAN bus to the digital motor electronics DME:

- A/C compressor information to increase the engine idling speed
 
- Information on the load torque of the A/C compressor
 
- Information on the set stage of the electric fan
 

Switching conditions for the magnetic clutch of the A/C compressor
A/C compressor ON                                                                                            A/C compressor OFF
 
Actuating of A/C button and                                                                                  Actuating of A/C button or 
 
Evaporator temperature > 2°C (depending on the outside temperature) and               Evaporator temperature < -0.5°C (depending on the outside temperature) or
 
Outside temperature > -10°C and                                                                           Outside temperature < -10°C or
 
Vehicle voltage ≥ 9.7 V and                                                                                   Vehicle voltage < 9 V or
 
Within permissible pressure range                                                                       Outside permissible pressure range or
 
                                                                                                                     Blower zero position or 
                                                                                                                    Terminal 15 OFF1 
 

1 With terminal 15 OFF, the current function ”A/C system” is stored in the EEPROM. The function light is switched off. Derived from the key memory (personalizing), the function last selected is activated at terminal 15 ON.

Pressure monitoring
The pressure of the refrigerant circuit is monitored by means of a pressure sensor on the dryer. Derived from the signal of the pressure sensor on the dryer: The load torque to be expected when switching on the A/C compressor and during its operation is derived from the signal of the pressure sensor on the dryer: The load torque in transferred via the K-bus and PT-CAN bus to the DME. The DME responds with corresponding engine control (adapted air and fuel volume).

[bmwdirtracer]

Thanks Abel....but you know I'm a bit of a computer dummy, so I'm not sure how that alters my perception of the problem. Now, I HAVE noticed that the IHKA always shows EXACTLY 300 rpm engine speed, until the throttle is pressed, and rpms rise.....while the Canbus data shows the correct engine rpm ~650 to be quite variable, as much as 25 rpm either way. Maybe that's due to the "telegram" part of the KBus equation?

Still, this is the only piece of data which is blatantly incorrect....and the IHKA-shown engine rpm DOES rise (proportionally) to the real engine rpm, until real rpm shows 1200, and IHKA-shown rpm hits~600 or 650. Wouldn't this mean that the K-bus info from the KOMBI is the problem?

[328 Power 04]

Well, here's the thing. K-line either works or it doesn't. So it seems somebody is sending out a wrong message to the IHKA. I blame that somebody.

- - - Updated - - -

There are ways to put in used clusters into the Z4 and have it work fine, but at the moment, with my arm in a brace from elbow surgery, I am no good at soldering.

[bmwdirtracer]

What did you do to your elbow, my friend?

I agree fully with your thoughts: that "Somebody" is sending the IHKA the wrong message....and that this must surely be the KOMBI, yes? I'm really good at soldering....but not circuitboards.

- - - Updated - - -

Further study of that quote you posted says that there's no engine speed requirement for the compressor to turn on? What did I miss?

[328 Power 04]

So, when, how, and why did the AC problem get brought up now? Curious of the timeline and if something was done to cause it. If anything.

- - - Updated - - -

I had surgery to release my ulnar nerve from being compressed by surrounding tissue. Sort of like a "tennis elbow".. but as I understand, a chisel was used as well on the bone.

[ross1]

What did you do to your elbow, my friend?

I agree fully with your thoughts: that "Somebody" is sending the IHKA the wrong message....and that this must surely be the KOMBI, yes? I'm really good at soldering....but not circuitboards.

- - - Updated - - -

Further study of that quote you posted says that there's no engine speed requirement for the compressor to turn on? What did I miss?

Glad you've tracked it down. I wish I had more to offer you here.
My limited experience with circuit boards is that there is often a failed solder joint at fault. Not sure if this could corrupt a digital signal(I'm thinking not) but maybe an exam of the boards in question and a reflow of the solder. I recently bought a soldering "re-flow station" to replace some surface mount devices in a TV and this thing works great, cheap too.
Good luck

[bmwdirtracer]

Abel, I'm not really sure how and when this came about....new customer, and it sounds like she's had the problem for a while....and she....and her husband (crap!) are going to live with it. Turns out they also own a brand new ZR-1 Corvette, too. The car went home today, and I missed it......

Many thanks for all the thought and input on this, my friends. The customer will never appreciate just how much thought and effort went into discovering the underlying issue....but I do!

Chris

[ross1]

Interesting to know that there is apparently a low RPM threshold below which the compressor won't be commanded(by the cluster?) on.

[328 Power 04]

Assumedly there to help engine start and prevent stalls.

Would have loved to see what happens to this Z4 if a used cluster was put in as a test. Oh well.

There has to be some history with this car.

E46 clusters had some issues with doing funny things, but nothing like this. E85 does use a similar era cluster..

[ross1]

@ Chris
was the RPM signal from the cluster halved all the way up or does(did) it correct itself at higher revs?

I wonder how half that signal got lost.
This is one of those you'd like to fix just so you could say you did it.