Idle Control Valve

[Todd]

Rookie on the board here. Just got an 87 e28 535i and having problems with engine killing at both idle as well as high speeds. Then after coasting for a few seconds it will fire right back up again. I have watched the Air Flow Sensor door at idle and watched it ramdomly fly open, killing the motor. I think I've narrowed it down to the ICV. HELP!

[m535is]

Have you checked the resistance of the ICV?

[Todd]

I have not tested that yet, no. Does it require a special gauge or tool? I'll start inquiring around and try to get that done. What should I be looking for in the way it SHOULD be acting? Thank you again m535i!!!

[DJM1986-5]

The ICV won't kill the motor at high speeds (above 1200 RPM). Sounds like other electrical issue.

[Todd]

Thank you DJM, I did replace all the fuses and relays, all the vacuum hoses and cleaned all of the grounds. Checked the ECM, connections look good. Could the ECM just tire out over time and need replacement? Is a bad ECM a somewhat common thing? Both fuel pumps are new as well. Frustrating.

[Todd]

djm1986-5

Why couldn't the ICV cause the engine to kill above 1200rpm? As I watch the Air Flow Sensor door move violently to the open position, and the engine dies, the only thing that could cause that door to move that violently to the open position would be a large flow of air into the intake manifold. The only two things, that I can see, that would allow that type of air movement would be, one the ICV, or two the throttle plate. During the time this is happening I have verified there is no movement of the throttle plate, so that leaves me to believe the only way to get this type of door action would be an influx of air through the piping that goes through the ICV, especially since the piping is so large. I also noticed that if disconnected the ICV is in an open state, allowing full flow of air though the valve and into the manifold. In conclusion I believe what may be happening is either the wiring connection to the ICV has a short, or the ICV itself is malfunctioning allowing full flow of air into the intake. I believe that once the motor shuts down, even at high RPM's the door is allowed to close, the ICV/ECM is reset, and the motor will fire back up.

Please let me know if this thinking is not correct, I'm just tying to put all the pieces together to find the solution.

Thanks for all of your help

[m535is]

Like I said, check it. It takes all of 5 minutes. You need a multimeter to check resistance. The 2 outer pins should be about 40 ohms and the center pin to each of the outside pin should be about 20 ohms a piece. That will let you know if the ICV is bad.

[Todd]

Sounds good, thanks for the specs, I will check that tomorrow.

[Scottinva]

I bet it just needs a good cleaning, they naturally get sort of gummed up. Whenever I have issues with an e28 and I assume it has something to do with the IAC, I take it out and make sure it isn't all nasty and gummy inside. I clean mine with carb clean, and finish it off with a squirt of wd-40 or some other light lubricant. I would bet that's the problem before I would assume electrical.

[DJM1986-5]

I bet it just needs a good cleaning, they naturally get sort of gummed up. Whenever I have issues with an e28 and I assume it has something to do with the IAC, I take it out and make sure it isn't all nasty and gummy inside. I clean mine with carb clean, and finish it off with a squirt of wd-40 or some other light lubricant. I would bet that's the problem before I would assume electrical.

Todd:
As written above, ICVs (technically Idle Air Stablizer Valve or IASV) are fairly robust and are checked easily. They do in fact have a tendency to stick after many miles (say 180k+). When stuck, they stick open and cause a high idle around 900 to 1100, and may cause a surging idle on top of that. If stuck, sometimes taking a plastic hammer and tapping sharply on the silver inlet will help it, but only temporary of course. They should be cleaned with "brakekleen" or similar, with the opening pointing south. Then dried and followed up with a lubricant such as WD40 or similar.

You are only 5 minutes away from ruling out the ICV. High speed stalls are not usually a symptom of the ICV - designed to feed air to the intake on idle (with a closed throttle plate).

[Todd]

Thank you again for the advice, I WILL rule out this ICV thing today!!! Crossing my fingers.

[Todd]

So, I did some checking today, and found that the ICV did check out in a fixed position, but I really needed an analog tester to verify full connectivity along the entire movement of the motor. Anyway, I decided to clean and lube the ICV as suggested in a previous post, which in the end I believe may have fixed the problem, at least temporarily. I took it on an hour test drive, as well as letting it idle for around 15 minutes, and not one stall. In conclusion I believe the ICV was causing the problem, I believe the problem stemmed from a wore electric motor that, in a particular position, would have an open circuit, causing the motor to briefly stop and allowing the full air flow into the intake manifold. I will continue to drive and test this theory, if I find anything different I will post. All this advise and testing is good information, and hopefully it can help someone else on a later date. Thanks again for everyone's advice and help. It takes a good board like MyE28 with people who have the knowledge and desire, and are willing to share to keep these cars on the road.

Thanks

[johnnye23]

[Todd]

The ICV is not a motor ,it is a solenoid . It may have indeed stuck due to build up but not because the motor had a bad spot. I think with the symptoms you stated your problem was not caused by an ICV.

You may be correct, although I disagree. Once the ICV was cleaned and lubed the vehicle did not perform any of the previous faults described in my previous post. I also disagree with the statement that the ICV is a solenoid. The ICV has a rotating shaft that opens and closes an opening allowing air flow at different rates depending on the condition of the motor. As I checked resistance there was current throughout the full movement of the shaft, and would range depending on location, but would remain basically the same. Since a solenoid is basically an on/off switch, using electric current I don't see how the ECM could control the air flow/opening size with a solenoid. With that said, even if it is a solenoid, there are contacts that wear over time and can become intermittent, think of a starter that is going bad. The contacts can become dirty or wore causing intermittent failure of the device. Again, the vehicle is currently running, and has not had the issue as of yet. I will definitely keep this thread going as to let people know the outcome. Oh, and I will disassemble the ICV to verify what type of device it is as soon as I get a new one.

Thanks

[Todd]

The ICV is not a motor ,it is a solenoid . It may have indeed stuck due to build up but not because the motor had a bad spot. I think with the symptoms you stated your problem was not caused by an ICV.

Oh, also if you don't believe the problem was the ICV, please feel free to add suggestions as to what you believe the problem may be. Although critism can be useful, advise on the problem is really the type of answer to the post I was looking for, and will help me the best in troubleshooting.

Thanks

[johnnye23]

You were right I was wrong glad you fixed your car .

[SamSpade]

Yes, the ICV is a solenoid. More specifically, it's a Duty-Controlled Dual-Coil Rotary Solenoid.

Our Bosch ICVs looks something like this, but without the bi-metallic strip:

Diagram taken from this Toyota document but the principle is the same.

The middle pin #2 is connected to B+ battery voltage. Pins 1 and 3 are alternately switched to GND by the ECU on a fixed frequency but with varying duty cycles and will be mirror images of each other:

Ignore the 3rd inconsistent red pulse. That was the result of bad firmware from the standalone ECU I'm using.

As you can see from the ICV diagram, varying the duty cycle would orient the shaft to either open or closed. The alternating opposing fields of the 2 coils 'balances' out to fix the shaft to a certain position, i.e. if pin 1 is pulled to GND longer (shorter duty cycle), the shaft will move to the closed position.

The humming we hear from the ICV is caused by the oscillation of the shaft, though in a very miniscule amount.

Being solenoids and not motors, they're pretty robust. But dirt causes the valve/shaft movement to get restricted in some positions. A lot of ICV have been thrown away when all they needed was some thorough cleaning.

Glad you fixed yours.

[tn535i]

In any case, an ICV going full open is no different than cracking the throttle and the ECM should adjust for this and the engine keep running and not stall.

I think it is far far more likely you still have an intermittent electrical issue and power is dropping out to all or part of the ECU, like a main relay or bad solder joints in the ECU board. Now what really happens is you simultaneaously loose spark and ICV control and see the resulting symptom, not the cause.

Bet it will do it again unless there is a reason a bad ICV connection would also cause and ECM problem, maybe the ICV 12v is shorting out somewhere and dropping ECM voltage enough, that might do it. Can movement of the internal ICV parts cause a short to occur? I wouldn't think so?

[Todd]

Yes, the ICV is a solenoid. More specifically, it's a Duty-Controlled Dual-Coil Rotary Solenoid.

Our Bosch ICVs looks something like this, but without the bi-metallic strip:

Diagram taken from this Toyota document but the principle is the same.

The middle pin #2 is connected to B+ battery voltage. Pins 1 and 3 are alternately switched to GND by the ECU on a fixed frequency but with varying duty cycles and will be mirror images of each other:

Ignore the 3rd inconsistent red pulse. That was the result of bad firmware from the standalone ECU I'm using.

As you can see from the ICV diagram, varying the duty cycle would orient the shaft to either open or closed. The alternating opposing fields of the 2 coils 'balances' out to fix the shaft to a certain position, i.e. if pin 1 is pulled to GND longer (shorter duty cycle), the shaft will move to the closed position.

The humming we hear from the ICV is caused by the oscillation of the shaft, though in a very miniscule amount.

Being solenoids and not motors, they're pretty robust. But dirt causes the valve/shaft movement to get restricted in some positions. A lot of ICV have been thrown away when all they needed was some thorough cleaning.

Glad you fixed yours.

Thanks, that is some great information. Thanks for clarifying, I am not that familiar with this type of ICV/selonoid, I am more familiar with a TPS, so any and all information will help me figure this problem out. I may have to get out my o-scope and verify signals from the ECM if the problem persists.

Thanks for you time

[Todd]

In any case, an ICV going full open is no different than cracking the throttle and the ECM should adjust for this and the engine keep running and not stall.

I think it is far far more likely you still have an intermittent electrical issue and power is dropping out to all or part of the ECU, like a main relay or bad solder joints in the ECU board. Now what really happens is you simultaneaously loose spark and ICV control and see the resulting symptom, not the cause.

Bet it will do it again unless there is a reason a bad ICV connection would also cause and ECM problem, maybe the ICV 12v is shorting out somewhere and dropping ECM voltage enough, that might do it. Can movement of the internal ICV parts cause a short to occur? I wouldn't think so?

I agree with your analysis of the situation. I also figured the ECM would, or should adjust for the change in air flow as if the throttle plate was opened, but as I watched the violent movement of the air meter door I don't believe the ECM could adjust for that quick of an influx of air. If the ICV by-pass lines were only 3/8th's or smaller I would never believe this could happen, but since it is a 3/4 line I believe this could and has been happening.

The relays and fuses have all been replaced, but I havne't checked the solder points on the board. If this persists, and I need to check the solder points, are there any specific locations on the board that are known to have this type of problem? I believe this type of scenerio is definitly still a possiblity, and makes more sence, but I thought I would start with the easiest fix first.

I figured if I had a short in the ICV I would either be blowing a fuse, or burning up the ECM completely, so I don't believe a short would be the issue, but an open circut could potentially cause the ICV to quit working, but would not have any affect on the EMC itself. This could cause the ICV to throw to an open position potentially causing the stall.

Thanks again for the information, it definitely gives me places to continue to look if the problem persists.

Thanks

[tn535i]

Maybe you're right about an ICV going full open causing a stall. But jumping full open would require an electrical stimulus I think, like maybe a full short to ground on the open side solenoid, maybe some ecm interaction. Just unplugging it for example leaves it in the last position and there is more restriction through the valve than the large hose they use. Why they use such a gigantic hose for what it does IDK? I recall that once when my ICV hose came off there was no way the car would run but that was all unmetered air. However, I still can't imagine wide open ICV any different that WOT except for where the air is delivered which is downstream of the throttle.

In any case, for someone who calls himself a rookie you seem very knowledgable and you'll have no trouble figuring it out if the symptoms persist. I'm sure you'll find the help you might need right here.

Welcome,

[Todd]

Maybe you're right about an ICV going full open causing a stall. But jumping full open would require an electrical stimulus I think, like maybe a full short to ground on the open side solenoid, maybe some ecm interaction. Just unplugging it for example leaves it in the last position and there is more restriction through the valve than the large hose they use. Why they use such a gigantic hose for what it does IDK? I recall that once when my ICV hose came off there was no way the car would run but that was all unmetered air. However, I still can't imagine wide open ICV any different that WOT except for where the air is delivered which is downstream of the throttle.

In any case, for someone who calls himself a rookie you seem very knowledgable and you'll have no trouble figuring it out if the symptoms persist. I'm sure you'll find the help you might need right here.

Welcome,

I will definitly keep everyone updated as to the status of the problem, thanks again for your help, and thanks for the welcome. I always consider myself a rookie when I start working on a car that I've never opened the hood on, hopefully I can pick up on the differences of a BMW quickly. I will do what I can to add to the board as much as possible.

Thanks

[SamSpade]

Maybe you're right about an ICV going full open causing a stall. But jumping full open would require an electrical stimulus I think, like maybe a full short to ground on the open side solenoid, maybe some ecm interaction...

An open or intermittent circuit on the closing coil (pin #1, ETM calls it 'retract control') will also leave the ICV fully open, since there will be no magnetic field to counteract the field of the opening coil (pin #3, 'extend control').

Intermittent issues are such a pain to diagnose. Has it happened again? Before cleaning it, did you test the movement of the ICV shaft by moving it with your finger or a screwdriver? It should rotate very freely, as in absolutely no frictional resistance.

[David Hunt]

Geez. Turn on the ignition without starting, and if the ICV is humming, it is good. Move the throttle plate, it stops humming and snaps shut. It works.
The air is bypassed to the ICV, around the AFM, right? The door opening is a response to the regulation of the engine speed. I don't think it is supposed to open that abruptly at idle.
You said you watched the thing at idle for 15 minutes? Then it works.
The reason an open ICV won't kill the motor at speed is that there is a great deal of air coming through the intake, and opening the ICV, with the restriction of the device itself and where the air ports through the body of the AFM, it would only lean out the mixture a bit, not kill the engine.

[Todd]

Maybe you're right about an ICV going full open causing a stall. But jumping full open would require an electrical stimulus I think, like maybe a full short to ground on the open side solenoid, maybe some ecm interaction...

An open or intermittent circuit on the closing coil (pin #1, ETM calls it 'retract control') will also leave the ICV fully open, since there will be no magnetic field to counteract the field of the opening coil (pin #3, 'extend control').

Intermittent issues are such a pain to diagnose. Has it happened again? Before cleaning it, did you test the movement of the ICV shaft by moving it with your finger or a screwdriver? It should rotate very freely, as in absolutely no frictional resistance.

Yea, I definitly agree with what you are saying, I also figured there should be some open circut causing the ICV to open fully, but I wasn't sure where that open was comming from. I figured the open circut may be coming from the operation of the ICV itself hence not necessisarily needing an outside sourse like a bad plug/wiring, or an ECM sinal.

The only good thing about intermitten issues is they can be fun and challenging to figure out, just frustrating when driving. No, the problem has not happened again since cleaning. The shaft did seem to move smoothly, but the shaft itself looked and felt like it had some scaring on the surface. It also had plenty of buildup on the shaft itself. I did notice the shaft felt weighted where it would rotate to the open position easier than the closed position, but I would assume that was the design of the ICV itself.

Thanks for more ideas, if this continues I will start looking at the wiring and ECM.