E46 Fanatics Forum banner

Mystery abnormal throttle response

24K views 37 replies 20 participants last post by  Brittany 
#1 · (Edited)
I bought this car with 130K on it from the original owner. 2002 325 XI with the electronic throttle (no cable). There is one annoying problem that I can not get resolved.

The problem:
When I give it 100% throttle then suddenly let off the car continues to go full power for about a half second, then goes to less power for half a second, then cuts all power. So from full throttle to no throttle there is at least a one second delay in response. This only occurs when the engine is under load, it does not happen when the car is out of gear or the clutch is in (free revving).

Some other possibly related things (that I am not even sure are problems): when the engine is revved briefly, it dips below idle, then goes above idle, then settles at idle. If the engine is revved high is does not do this.

Things I have tried:
I cleaned the TB (throttle body), ICV (idle control valve), disa (?), replaced the intake boots, and reset the throttle memory. I had the car smoked tested for vacuum leaks which came up negative. The car has a new MAF (mass air flow) which is from a hyundai sonata, but this issue was the same before changing the MAF. The TB seemed to function properly, it wasn't sticky or anything. The previous owner replaced an o-ring in the clutch assembly which means bleeding the clutch hydraulics. There is no guarantee this was done right. (may be why this only happens when the car is in gear.)

My indie confirmed this issue but has no idea what to do next. He suggests I take it to the dealership to use their big expensive computer, which I assume means a big expensive diagnostic fee. ANY input would be good, I am getting desperate and completely grasping at straws here.
 
See less See more
#2 ·
I have this exact problem.. however I just cleaned the ICV last night and it made this happen less frequently, but it still happens sometimes. Did you notice any difference at all when you cleaned ICV? I'm thinking I just didn't clean it enough. there was still a little bit of gunk in there when I put it back, since the cleaner I used couldn't seem to get it out.
 
#3 ·
I cleaned the ICV as throughly as an ICV can be cleaned. I used 2/3 of a can of TB cleaner on it in addition to rags and stuff. If you twist the ICV back and forth about its long axis you should hear the valve inside click. It should be a metal to metal click. If its more of a dull thumb or thud, then it needs more cleaning. Make sure that it is perfectly dry before you reinstall it.

I thought I noticed a difference the first time I put the ICV back in, but it was only for a mile or so and I am not sure if it was in my head or not. Besides that there was no difference at all.

I am sorry you have the same problem, but am happy to hear that I don't have some freak E46 and am never going to find a solution. Lets keep in touch, you are the first person I have talked to with this same issue, and I have been trying to solve it for 3 months now.
 
#7 ·
Yes, it's electronic; I don't have a throttle cable. I'll help keep this thread bumped until we can find a solution. I'm sure it's probably bugging me just as much, since I've been dealing with it for years now. Have you done your vanos seals yet? I did them last year and it made the ride a little smoother.
 
#6 ·
Well I am pretty sure I sufficiently cleaned the TB. The butterfly was perfectly clean. But there was still some discoloration on the tube part of the TB, not enough to cause a physical change to the surface of the TB. Holding the TB im my hand and manually opening/closing the butterfly, there was not any abnormal resistance in the butterfly valve action. There was a slight delay because the spring in the TB was moving the electrical actuator but that delay was not even enough to measure. (maybe a 1/10th of a second). Even then, if TB buildup were the culprit, cleaning it would have improved the problem. I am not going to say that the TB is not the issue, but if it is I think it would be the TB motor itself that has gone bad. I really don't think this is caused by any kind of buildup or anything. On the other hand this is the first "modern" car I have tried to trouble shoot, so feel free to argue with me. I am by no means an expert.
 
#13 ·
A slight delay inthe throttle closing may be normal but this definately is not. Like I said, there is aleast a second of hesitation between letting off the accelerator and the engine cutting power. Its enough time to move your foot all the way from the gas to the break and apply the breaks. This is confirmed by my mechanic. But it may be something in emission control malfunctioning. Any idea what that could be?
 
#14 ·
A lot of manual trans cars are programmed to drop the rpms slowly when moving to aid in shifting, BMW may be the same and you have an issue that aggravates this. may have to have it put on a scan tool to see tps, fly by wire, etc settings. Is there some maintance of the throttle body that is recomended? something that could benefit from lube?
 
#15 ·
this sounds like the idle control valve is mechanically sticking. Although, it could be an issue with the actual electronic throttle body. Unfortunately, you really should have the BMW codes pulled from the DME. You may find a fault already exists with one or the other which could help you in finding the problem.
 
#22 ·
Hey guys,
Let me know Ive got a simialar prob,,car sat for 10 months,,,Long story short would idle,,found cracks in the all the tubes and replaced the ccv and tubes,,,,Idles perfect now but in 5th(auto) on the hwy idle bounces up and down, and can feel the car not right,,,Driving me nuts,,,only code was po444 on the cheap machines,,,but looking around on here,Hell it might be the Disa valve,,,a vacuum leak(again),,o2 sensor,cat,, dmu,, now read may be a tranny prob...But if i put my cruise on it kinda holds steady,,, I may take to dealer just to see what codes they pull up,,,but gotta save up the 135 hook up price,,,, tired of pulling my hair out,,,and just cant keep throwing parts at it,,,
 
#25 ·
Hey now, careful with those big words, haha. Bently's says there are two pots in the module and Ill be checking them both over Thanksgiving. Do you know what the range and tolerance for the pots are? Bently's doesn't have that info. Also is there a good resource for more advanced technical info like i/o voltages and resistances of different sensors on this car?

While this seems like this is a good place to start, I am still not very convinced this is the problem. it doesn't occur when the car is not under load, i.e. the car is in neutral or the clutch is pushed in. I would think if it were something in the accelerator pedal, the problem would occur 100% of the time regardless of weather the clutch was in or not. Are there other potentiometer type sensors in the gear box somewhere?

I also want to clarify that this is not the Stick-slip effect explained here:

http://www.meeknet.co.uk/E38/E38_Stick_Throttle_Pedal.pdf
 
#26 · (Edited)
Update

Update:

I finally got around to checking out the accelerator pedal module. Apparently there are two potentiometers behind the pedal that detect pedal position. I took a multimeter and measured the resistances between the six different pins on the module. I was able to find what pins go to which pot and determined that the two pots are not connected. This is confirmed by the wiring diagram in Bentleys.

The thing that has me totally stumped is that the resistances didn't change when the pedal was pressed. This is completely perplexing to me, since if the resistances don't change when the pedal is moved, the computer should not be getting any signal at all, yet the car revs fine (except for the original issue). Any input on the would be helpful. Thanks.

Here are the measured resistances:
Pot 1: pin 1-4 = 57 ohms, pin 4-5 = 427 ohms, pin 1-5 = 477 ohms
Pot 2: pin 2-6 = 57 ohms, pin 3-6 = 427 ohms, pin 2-3 = 477 ohms

Also, I drove the car 8 hours (at the speed limit) in heavy rain a week ago. My destination was 3000 ft lower in elevation than home. After this trip, the response issue was much better, and I had more power across. The problem is slowly coming back as I am driving around town.

Again these did not change when the pedal was pressed :banghead:
 
#27 ·
Hmm, your car does seem to have the symptoms associated with a bad accelerator pedal. The DME will slow responses to sudden changes in input when it detects a fault.

Btw, there are two different types of drive-by-wire systems on E46s. One style uses pedal linkage to actuate two potentiometers, the other style is an enclosed pedal. I believe you have the latter, which uses hall sensors, so you need to apply +5V (as the DME does, but you could probably use a little more/less) across pins 5 and 1 and measure voltage at 4, or across 3 and 2 and measure at 6. One output should measure exactly half of the other.
 
#30 ·
Btw, there are two different types of drive-by-wire systems on E46s. One style uses pedal linkage to actuate two potentiometers, the other style is an enclosed pedal. I believe you have the latter, which uses hall sensors, so you need to apply +5V (as the DME does, but you could probably use a little more/less) across pins 5 and 1 and measure voltage at 4, or across 3 and 2 and measure at 6. One output should measure exactly half of the other.
This was spot on! thanks for the info. Here is what I did. USB is 5V so you can take a USB charger for a phone (iphone / ipod / etc) and plug an old printer cable into it. Then cut the other end of the printer cable off and use the red and black wires. Instant 5V power supply. (Double check the 5V with a multi meter before connecting it to the pedal).

hall sensor 1: Pin 1 = negative, Pin 5 = positive, measure at pin 4
hall sensor 2: Pin 2 = negative, Pin 3 = positive, measure at pin 6
measure each sensor separately, red lead of your multimeter on 4 or 6, black lead on 1 or 2 respectively. Make sure your volt meter is set to voltage. Now depress the accelerator pedal and you will see the voltage change.

Results: I got .72V - 4V on sensor 1 and .37V - 2.0V on sensor 2. From what I have read, sensor 1 should be .5V - 4.5V and sensor 2 should be .5V - 2V. Since the maximum is 4V instead of 4.5V I am only getting 88% of max throttle, I believe this warrants a replacement. If you disagree let me know.
 
#28 ·
PWG modules (aka. accelerator pedals) rarely go bad. the issue people are referring to is the sticky/notchy "stick-slip effect" pedal feel. It is purely mechanical and is related to the lubricant used on the pedal that causes the mechanism inside to bind when it gets hot (parked car in summer or heater on in winter) OP, you would physically feel the pedal sticking if it were to blame. Also, the EML would go to failsafe if there were any issues with the potentiometers in the pedal so I doubt that's it.

Since the condition changes with altitude, I would look to an intake leak like the DISA valve. It seems like they tend to go bad in the winter.
 
#29 · (Edited)
I completely agree. I have read up on the stick slip effect and I know it was not that. I had a vacuum leak test done, and it passed, so I figured it could be the pedal, disa, ICV or TB. I pick the cheapest to rule out and am working my way up to the most expensive.

Also, I found a bunch of info on the Hall sensors. take a look here: http://www.bmwmotorsports.org/BMW_docs/m54x5.pdf
Is there a place that has a lot of technical info on the E46, like the voltage in and out for other sensors/actuators?
 
#32 ·
Unfortunately, I have not resolved this and have decided to give up for now. I know it is not the pedal, the maf, the tb, or the icv. Since the problem is predictable and not constant, its safe to assume its not a mechanical failure. I am 90% sure it is a software issue or a speed sensor issue. My dealer charges $120 for a software update so I decided to let it ride and see if it gets worse. If you really want it solved I would take your car to the dealer and make sure the software is up to date.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top