Coding out poor Idle

[nyc_dude]

Like everyone else, my M3 idles poorly. 650 rpm down to 500 rpm with a shudder.

When you turn on the A/C the idle fixes itself and jumps to 700-750 rpm and remains steady. That means it isn't a mechanical issue like the fuel tank breather valve, low pressure fuel sensor, vacuum leak, EGR valve, secondary blower motor, idle control valve, throttle value actuators, fuel pump, EPK FP module, etc..

The source of the issue must then be in the ECU software. Is there an idle speed setting with and without the A/C on or off that can be coded to be higher than 650 rpm?

Dreading going to the BMW dealership so they can troubleshoot it for a month along with my gas cap SES emissions light.


Cheers

[jcolley]

There are base idle aim tables vs temp along with a couple of correction tables.

The AC compressor is just one of the items that the DME monitors for countertorque along with the alternator and "reibmoment" (friction torque) in the control of idle. The torque manager has a map of potential torque output of the engine in RPM vs. load. The torque manager then calculates all of the torque consumers to determine total drag and does a reverse lookup to the torque table to determine what "load" value (relative filling %) correlates to that newton meter torque value at the idle aim table RPM. It then sets the RF (by calculating the ideal throttle and idle actuator settings). Then ignition timing is used with a PID control to set the correct ignition advance to hold the target idle RPM at the given idle/throttle actuator positions.


EDIT: Also, there is a point with ignition based idle PID control that an engine can become filling limited. What I mean by that is, when air flow is so low, you can only advance timing so far before the engine reaches MBT and further advance yields no torque gain. This is easy to see at very low filling values where the engine can't advance timing far enough to raise torque enough for produced torque to exceed countertorque this increasing RPM. With the mss60 DME, raising the idle setpoint actually causes the torque manager to choose a higher RF target, so more air is going in allows the ignition based idle control to control RPM as desired.

From the factory, idle is set low with filling values as low as possible and igntion timing as close to MBT as they reasonably can to maximize available torque with minimal fuel consumption and emissions.

[wfdeacon88]

Mine idles at 550-600ish without anything running, but it also does what you described.. Drops slightly to 500 (or a little less) every now and then when warm and does the shudder thing and I can feel it in the wheel. Is this definitely just an ECU issue as you noted? I've always been paranoid when it happens... Would be nice to know its not mechanical (RB, etc) at least.

[EJT86]

How are you sure it’s not the fuel breather? When my fuel breather went bad I constantly ran the heat specifically to raise rpm so it wouldn’t stall

[nyc_dude]

A gold star to our friend from Maine. Yahtzee!! Free Lobsters for life.

I had the idle problem last year after 150k+ miles with the original rod bearings. I didnt notice it since you use the A/C all summer long in Cali (Hell) and get a nice high idle at 700-750 rpms. Then when the Fall comes you stop using the A/C, the low idle comes along and you think the engine will explode at every stop light and on the nightmarish 405 during rush hour. In my case, the motor mounts were collapsed like a bagel with cream cheese. So the shudder at 500rpm was much worst.

Had the rod bearings done and new motor mounts installed. I was optimistic, since the shudder was greatly reduced and the car was running smoothly, except at idle. Fuel pumps and entire DCT were replaced at XMAS, so it is not fuel starvation.

I had most preventive maintenance done right before the RBs, including spark plugs, alternator, belts, and water pump/thermostat, plus a lot more. The shop couldnt get the fuel tank breather valve in time. A month later, the gas cap SES emissions light goes off. Replaced the low pressure fuel sensor and DMTL pump since that was easy, along with a new gas cap. Still get the error code. New Bosch fuel tank breather valve is next. But if it is bad, how come the idle is perfect with the A/C On.

Could be the pair of cam position sensors on top of the valve covers. But then the car wouldn't run properly at different engine rpm, if the cam position sensors were faulty.

I had the ICV service done years ago...but that fixes a rough idle. Third pair of TVAs have only 30-40k miles. Again, the idle wouldn't be smooth with the A/C On if these device were bad.

O2 sensors? Maybe...but with the A/C On they work fine.

BMW screws up multiple design items on most of their cars, and then avoids fixing them through a redesign to save money on mandatory Recalls. But DME SW fixes should be easy, but the testing and worldwide rollout makes them too expensive. DCT M3 GTS SW is much better that the stock DCT SW, but BMW never rolled it out to every M3 customer.

So the rpm and torque tables need a tweak at idle speeds. Will have to see if that can be coded by BPM or AlpineAlex. They might be able to develop a global fix for everyone in the M3 universe. Idle rpm with A/C Off = Idle rpm with A/C On.

A Big Thanks. I am an Aerospace Engineer and understand 90% of the write up on the DME software.

Cheers

PS Any low rpm at idle that causes stalling should be a fuel pump issue. Wild swings in rpms means it is the fuel pump, including stalling at zero. There is also a EKP module in the back seat that is easy to replace that controls the fuel pumps (2). Sometimes a tune or flash will fry the EKP. I understand that with a full tank of gas, sometimes the car stalls leaving the gas station, due to the fuel tank breather valve.

[wfdeacon88]

Originally Posted by nyc_dude

A gold star to our friend from Maine. Yahtzee!! Free Lobsters for life.

I had the idle problem last year after 150k+ miles with the original rod bearings. I didnt notice it since you use the A/C all summer long in Cali (Hell) and get a nice high idle at 700-750 rpms. Then when the Fall comes you stop using the A/C, the low idle comes along and you think the engine will explode at every stop light and on the nightmarish 405 during rush hour. In my case, the motor mounts were collapsed like a bagel with cream cheese. So the shudder at 500rpm was much worst.

Had the rod bearings done and new motor mounts installed. I was optimistic, since the shudder was greatly reduced and the car was running smoothly, except at idle. Fuel pumps and entire DCT were replaced at XMAS, so it is not fuel starvation.

I had most preventive maintenance done right before the RBs, including spark plugs, alternator, belts, and water pump/thermostat, plus a lot more. The shop couldnt get the fuel tank breather valve in time. A month later, the gas cap SES emissions light goes off. Replaced the low pressure fuel sensor and DMTL pump since that was easy, along with a new gas cap. Still get the error code. New Bosch fuel tank breather valve is next. But if it is bad, how come the idle is perfect with the A/C On.

Could be the pair of cam position sensors on top of the valve covers. But then the car wouldn't run properly at different engine rpm, if the cam position sensors were faulty.

I had the ICV service done years ago...but that fixes a rough idle. Third pair of TVAs have only 30-40k miles. Again, the idle wouldn't be smooth with the A/C On if these device were bad.

O2 sensors? Maybe...but with the A/C On they work fine.

BMW screws up multiple design items on most of their cars, and then avoids fixing them through a redesign to save money on mandatory Recalls. But DME SW fixes should be easy, but the testing and worldwide rollout makes them too expensive. DCT M3 GTS SW is much better that the stock DCT SW, but BMW never rolled it out to every M3 customer.

So the rpm and torque tables need a tweak at idle speeds. Will have to see if that can be coded by BPM or AlpineAlex. They might be able to develop a global fix for everyone in the M3 universe. Idle rpm with A/C Off = Idle rpm with A/C On.

A Big Thanks. I am an Aerospace Engineer and understand 90% of the write up on the DME software.

Cheers

PS Any low rpm at idle that causes stalling should be a fuel pump issue. There is also a EKP module in the back seat that is easy to replace that controls the fuel pumps (2).
Sometimes a tune or flash will fry the EKP. I understand that with a full tank of gas, sometimes the car stalls leaving the gas station, due to the fuel tank breather valve.

So is this idle thing a sign of any sort of issues within that could affect performance, or do you suspect it's mostly an issue that stops when not idling? of course, assuming it's an annoyance and never stalls or anything ...

[nyc_dude]

It is a mystery, without a fix, especially if it is a DME SW flaw.

My problem, like many others, is just at idle. If your idle fluctuates wildly, then that is the fuel pump, which is more serious. But it is an easy fix. The fuel pump parts are $700-1000 with minimal labor. Devices are under the bad seat. See youtube for the install.

BMW 3-series e9x models have fuel pump reliability issues. It is a mostly plastic part. There are two half fuel tanks under the back seat (so two fuel pumps), plus an EKP electronic module. Not sure if the new EKP has to be coded to your DME via computer or if a replacement is plug and play. Might be able to replace only the fuel pumps and leave the original EKP module installed. That is what I did.

Symptoms of a bad fuel pump, are erratic idle/stalling and stalling at speed and your DCT goes into neutral.


Cheers

[jcolley]

Another thing to consider is that actual filling may not be the same as what the DME perceives to be "relative filling".

Relative filling is calculated via multiple paths and compared, i.e. via airbox pressure compensated alphaN (since the pressure sensor is upstream of the ITBs, it's technically not speed density / MAP), via the o2 sensor path (reverse calculated from lambda and fuel pulsewidth) and on the mss65 (s85) via MAF.

The downside of all predictive volumetric efficiency models is that if anything in the intake tract changes from the initial calculation (i.e. changes in head losses, throttle body diameters, cam duration/lift, etc), then the actual filling is different than the calculated and the DME relies on filling adaptations and lambda correction (reactive, not proactive) to optimize the lambda.

On the s65 and s85, we see a lot of carbon build up, especially with certain oils that are more susceptible to blow-by and poor fuel qualities. This coking of the combustion chamber alters the filling characteristics and induced air flow is lessened. Especially at low loads when filling is minimal anyway, it can easily drop airflow a few percent to the point that idle ignition advance becomes MBT limiting and there just isn't enough air coming in for any ignition angle to overcome.

I'll have a look in a bit at the idle values in the tables, but most tuners should have the base idle table mapped which establishes the minimum idle target regardless of other compensations.

I'd take a boroscope down the spark plug hole and see if there's obvious carbon build up.

[wfdeacon88]

Mine isn't erratic, just a subtle dip near or below 500.. but then the engine kind of shudders as OP described... at the same time I can feel it in the steering wheel enough to notice. If music is on, it's harder to notice, but I always do feel the subtle wheel vibrations when it happens. Anyone else in the car would never notice.. but being paranoid about my scheduled RB job (as soon as BE restocks) I am extra sensitive to these changes. It doesn't appear to have any performance impact and I've never stalled, but just worried it's a sign of something wrong. I haven't seen any error codes trigger.

[nyc_dude]

Get new motor mounts with your new BE rod bearings service.

Just put the A/C On at the lowest setting to avoid the issue.

Cheers

[nyc_dude]

Had my BMW e90 M3 serviced to replace the Idle Control Valve, Idle Fuel Rails/Grommets, and Hoses. One new idle fuel rail comes with a new MAP sensor attached. The hoses are gray, so you think they are metal, but they are cheap $$$ rubber hoses. This repair included a new Fuel Tank Breather Valve (Bosch). That solved the Idle problem and an Emissions light (gas cap warning). Rock solid idle now. The parts and labor for this Idle Control Valve service is $2000+. It is double the price of just a new ICV part.

Check these SIB 13 05 08 and SIB 01 03 16.

The Master ZTech at AutoTalent indicated that during some prior service (Spark Plug Change or Valve Cover Gasket Change), an ignition coil was damaged with the rubber hat being ripped off.

The idle issue was fixed, but the shudder still existed. Changed all 8 ignition coils (BMW OEM) myself. Just needed a long screwdriver and a trim tool. No puller, coat hanger, or shoe lace. Broke both plastic ignition coil covers (front grommets) and one cabin air filter housing ($275). Prior service will damage both cabin air filter housings on everyone's BMW e9x M3. The front screw tab is super thin plastic, so reinforce it with a strip of thick plastic.

Idle issue solved. The S65 V8 roar is back. $3000.

EPA 8yr/80k miles warranty might cover the ICV. ICV, like the TVAs, are defective as design. Premature wear and failure.

I have the old parts and will cut open the hoses and ICV. SIBs say there is rubber debris that causes the ICV to fail. I doubt that. The original threads on this problem indicated the idle problem appeared in new M3s with only 7k to 15k miles. That is not enough time for the hoses to disintegrate from the inside. ICV was just designed very poorly. Bosch Fuel Tank Breather Valve is also junk, and causes false gas cap and DMTL faults codes.


Cheers

[nyc_dude]

New Turner Motorsports Driveshaft also ROCKS. No more drivetrain klinking noises. Instant Power from the sweet //M3 S65 V8.



Cheers

[cmichaelo]

So does anyone know if it's possible to code the idle speed to something different from the factory setup? And how?

From my carly adapter it appears the "Desired Idle speed" is 720 rpm. However, the idle speed can drop to 500 rmp when really warm and that's a problem for my aftermarket battery which doesn't like getting charged below 13V and at 500rpm, the charge voltage can dip into the 12-13V range.

So I'm trying to find out how can increase idle speed a little, say +100 rpm.

[C-los]

Tried Removing your air box and cleaning the throttle bodies?
Carbon Builds up around the slight opening of the throttle body(ies) and can cause low idle issues. The dme tries to compensate but there are limits, when it's reached, the carbon buildup gets worse causing low idle, hard starts etc.

[cmichaelo]

I've not tried clean the TBs. However, the idling has been as I described since I got the car in 2012. It's just that the 500rpm idling never was an issue with the AGM batteries. But it seems to be an issue with the LiIon battery I have.

Also, the previous posts in here all seem to describe the same idle rpm variation.

So, there must be a way to increase the idle rpm in the ECU. Anyone knows?

[C-los]

There's no 'coding' for RPM to my knowledge.

You would need a 'tuner' to modify the idling tables in the mapping.

You might try changing the battery coding to see if it would be of any help since you think it's an issue with the battery type you've changed to, but i doubt it would be of any 'real' help, unfortunately. I also don't think the battery itself if going to cause any type of 200 rpm worth of load change... but i could be wrong.

[cmichaelo]

I've coded the battery according to the manufactorers recommendation, namely 40AH AGM.

I don't think the battery is causing any load to the engine itself. The idling is behaving the same with the giant stock battery.

What do you mean with "a tuner" ? Is that a special OBD2 device?

[nick3753]

Just had bearings and new motor mounts (121k), rpm still drops, but no more shuddering, I assume from the new mounts.