RPM occasionally fluctuates with AC Compressor on

[ac427]

Just taking a quick look and not having read all 7 pages. How does the ecu determine the effort needed to maintain idle?

I believe there maybe a separate map for when the A/C is engaged. Would the A/C pressure sensors come into play?

[Bry5on]

Just taking a quick look and not having read all 7 pages. How does the ecu determine the effort needed to maintain idle?

I believe there maybe a separate map for when the A/C is engaged. Would the A/C pressure sensors come into play?

The DME estimates a torque request from the AC compressor, it actually translates this as a number over the CAN bus (or was it D bus?) and uses this to set at least the throttle relative opening target as the revs fall to idle. I don’t think it would cause hunting, unless maybe the AC compressor clutch was slipping? Hmm.

[Andratch]

Well folks, I think I’ve reached a stopping point on this for now. I feel a combination of the Mullet Tune and a 950 rpm idle has made this much more manageable.

I’ll keep trying to find a solution, but it’s now cold out so the AC isn’t needed as much. When it warms up again outside, I may try to convince Kassel to let me swap the DME and see what that does.

[0-60motorsports]

Well folks, I think I’ve reached a stopping point on this for now. I feel a combination of the Mullet Tune and a 950 rpm idle has made this much more manageable.

I’ll keep trying to find a solution, but it’s now cold out so the AC isn’t needed as much.

Wish you the best of luck. Been years for me and never found the solution. Bet you its wiring degradation or ECU component failing....

[dukeofchen]

I haven’t recharged it, but I tested the low/high pressures and they were “in spec”

From my understanding our AC systems don't line up with the standard universal charts out there, because BMW uses a variable displacement type compressor. In my observation, the compressor instead regulates to a low side pressure in the 35psi range, and as refrigerant charge increases only the high side pressure will rise unless it is severely overfilled. In contrast, the standard charts you see show both low/high pressures increasing with ambient temperature. The implication being that if going only off pressures, at high ambient temps the low side pressure will be lower than the chart, and the "happy zone" on the high side might actually be higher than what standard charts would suggest.

The only true spec for our cars is the total refrigerant mass, so a total evac and refill does seem like something easy enough and worth checking off the list, especially if the other options are a deep dive into engine tuning or chasing electrical gremlins.

[Andratch]

Hey Folks, looking to get y'all's thoughts on one more test I may want to run...

I'm curious if there's something in my DME (and 0-60motorsports and the others with this specific problem) that's failing that is causing this issue, and want to isolate it. My car is a '05, my DME was a non-HP from a '03, converted to HP and upgraded to CSL by Kassel. Maybe something went wrong in that process.

I have a local shop that is open to letting me 'borrow' a tested, unmodified HP DME from a car they're converting to S62 for a test. Could I get that, save a backup of the factory tune, then load at least temporarily load an alpha-n tune on it to see if the problem goes away? If that fixes the fluctuation, then I buy that DME from the shop, convert that DME to CSL, put the mullet back in, and go about my merry way. If it doesn't, then I flash the borrowed DME back to factory, give it back.

Is that a viable plan? I wouldn't need to modify the borrowed DME to test it if I use a alpha-n tune temporarily, right? Am I missing anything?

NOTE: I'm holding off on replacing the AC compressor for now because 0-60motorsports and others mentioned that they replaced their compressor/dryer or did vacuum-and-recharge with no change, and it'd be odd if all of our AC compressors were 'quiet quitting' at the same time.

[0-60motorsports]

Hey Folks, looking to get y'all's thoughts on one more test I may want to run...

I'm curious if there's something in my DME (and 0-60motorsports and the others with this specific problem) that's failing that is causing this issue, and want to isolate it. My car is a '05, my DME was a non-HP from a '03, converted to HP and upgraded to CSL by Kassel. Maybe something went wrong in that process.

I have a local shop that is open to letting me 'borrow' a tested, unmodified HP DME from a car they're converting to S62 for a test. Could I get that, save a backup of the factory tune, then load at least temporarily load an alpha-n tune on it to see if the problem goes away? If that fixes the fluctuation, then I buy that DME from the shop, convert that DME to CSL, put the mullet back in, and go about my merry way. If it doesn't, then I flash the borrowed DME back to factory, give it back.

Is that a viable plan? I wouldn't need to modify the borrowed DME to test it if I use a alpha-n tune temporarily, right? Am I missing anything?

NOTE: I'm holding off on replacing the AC compressor for now because 0-60motorsports and others mentioned that they replaced their compressor/dryer or did vacuum-and-recharge with no change, and it'd be odd if all of our AC compressors were 'quiet quitting' at the same time.

Good plan, my DME is with a forum member as well for testing but we dont have a timeline when he will get a chance to check it.

I bet we have a wiring degrading issue or dme component degrading issue going on due to the age of our cars.

[Andratch]

Hey Folks,

Update - I did a live tuning session with HTE Performance today, hoping to get a new set of eyes on this and collect some more insights. Actually, it looks like HTE was able to tune out the issue! I did not need to go Alpha-N.

Hassan asked me not to share the exact tweaks he made out of confidentiality (and I imagine he wouldn't want people blowing up S54s by taking stabs in the dark from internet advice) but i will say he changed values in a few different areas to get the result.

The asterisk here is that it's fairly cold in Houston today (50 degrees...cold for us, anyway), so the AC wasn't having to work as hard as a 100 degree day. But as he tuned and flashed updates, I tried all of my normal tricks to get the idle to fluctuate and the fluctuations were progressively smaller until they didn't happen basically at all. When its warm out, I'll do another test.

[Bry5on]

Awesome! Glad he was able to get you sorted

[0-60motorsports]

Hey Folks,

Update - I did a live tuning session with HTE Performance today, hoping to get a new set of eyes on this and collect some more insights. Actually, it looks like HTE was able to tune out the issue! I did not need to go Alpha-N.

Hassan asked me not to share the exact tweaks he made out of confidentiality (and I imagine he wouldn't want people blowing up S54s by taking stabs in the dark from internet advice) but i will say he changed values in a few different areas to get the result.

The asterisk here is that it's fairly cold in Houston today (50 degrees...cold for us, anyway), so the AC wasn't having to work as hard as a 100 degree day. But as he tuned and flashed updates, I tried all of my normal tricks to get the idle to fluctuate and the fluctuations were progressively smaller until they didn't happen basically at all. When its warm out, I'll do another test.

Great news! I had already spoken to him about you and myself and i just need to get the car hooked up for him to sort it out. Enjoy!

[Andratch]

Well folks, unfortunately, I spoke a bit too soon. The temperature warmed up here today, and once I got the car to operating temperature, it was fluctuating again.

Hassan had a look and notes that It gets pluses for load with what looks like a delay it starts playing catchup. He mentioned that perhaps my map sensor location could be the problem. I’m using one of those sensor kits that puts the map sensor up by the brake booster, not on the rail itself, like the true CSL.

He also mentioned it might be oxygen sensor trims. So, Just working through process of elimination right now, then responding to what we find.

[lemoose]

Well folks, unfortunately, I spoke a bit too soon. The temperature warmed up here today, and once I got the car to operating temperature, it was fluctuating again.

Hassan had a look and notes that It gets pluses for load with what looks like a delay it starts playing catchup. He mentioned that perhaps my map sensor location could be the problem. I’m using one of those sensor kits that puts the map sensor up by the brake booster, not on the rail itself, like the true CSL.

He also mentioned it might be oxygen sensor trims. So, Just working through process of elimination right now, then responding to what we find.

Same boat as well and Ive had a hunch its the MAP sensor location too because if i hold the brake down after starting, it will settle down.

[Slideways]

The MAP sensor in that location should not cause an issue.

One other part that is in that area and part of the vacuum system is the brake booster. Maybe there is something there causing a vacuum issue. I had one experience where the ICV port was not fully clipped in under the intake manifold and the vacuum leak would cause the engine revs to fall below 900 RPM and stall when pressing the brakes to slow for a stoplight. Fully clipped the ICV line back in under the airbox and it fixed the issue.

Here is a thread on E46F about the booster - https://www.e46fanatics.com/threads/...-leak.1057387/

[Andratch]

The MAP sensor in that location should not cause an issue.

One other part that is in that area and part of the vacuum system is the brake booster. Maybe there is something there causing a vacuum issue. I had one experience where the ICV port was not fully clipped in under the intake manifold and the vacuum leak would cause the engine revs to fall below 900 RPM and stall when pressing the brakes to slow for a stoplight. Fully clipped the ICV line back in under the airbox and it fixed the issue.

Here is a thread on E46F about the booster - https://www.e46fanatics.com/threads/...-leak.1057387/

I see where you’re coming from. But I don’t know how the whole “mainly happens when the ac compressor engages” bit would be influenced by that.

I tried unplugging the MAP sensor as well, and unfortunately nothing changed with that. Ignoring O2 inputs also changed nothing. There are a few more tricks to try, but I’m bothering the neighbors so I’ll get on that tomorrow.

[Bry5on]

I see where you’re coming from. But I don’t know how the whole “mainly happens when the ac compressor engages” bit would be influenced by that.

I tried unplugging the MAP sensor as well, and unfortunately nothing changed with that. Ignoring O2 inputs also changed nothing. There are a few more tricks to try, but I’m bothering the neighbors so I’ll get on that tomorrow.

Interesting point. You could try disconnecting then plugging the line from the brake booster. Maybe when the revs surge it's pulling more vacuum that causes the leak to be of meaningful value?
edit: remember there's a check valve in front of the brake booster, so a smoke test would not identify this vacuum leak if it's there.

This still sounds like a hardware issue to me, and not a tune issue..

My MAP sensor is next to the brake booster for what it's worth.