1984 533i loping idle diagnosis

[richardb]

Hi all, hope your summer is going well. I've been having some idle issues with my 84 533i lately and would appreciate some input. I have read about 5 threads here diagnosing this issue specifically for the 533i. Here's where I'm at, and some recent work:

Just installed new plug wires, spark plugs, oxygen sensor, fan clutch and air filter (maintenance unrelated to idle issue)

The Issue:

The car idles high around 1,200 rpm at cold start until it reaches temp, at which point the idle drops down and becomes loping up and down from roughly 1,000 to 1,100 rpm. Periodically it will stop the loping and remain a solid idle for 30 seconds or so. It will then begin loping again.

Attempts to resolve so far:

1. I cleaned the ICV and inspected the tubes and intake from the AFM, no obvious cracks or leaks.

2. Removed the ICM from glove box and opened it up. Removed board from green plastic housing for inspection. Generally speaking it appears ok, however one resistor appears to have a brown or potentially burnt spot on it. Though I cannot confirm this is not normal. Photos here: https://bimmerzeit.com/2020/08/17/533i- ... ol-module/

While it is common for the ICM (idle control module in glove box) to be the culprit, I haven't seen a way to directly check if it is faulty. I reached out to the company that rebuilds them and they were unable to provide an answer as to how one even confirms the module is faulty through testing.

3. Running the car with the ICM unplugged made no change in loping idle.

4. Running the car with the ICV unplugged made no change in loping idle.

5. Checked resistance on ICV with meter and read 9.4 ohms. In-spec is supposedly 9-10 ohms.

6. Tested voltage at the ICV plug with key on 11v. With car running 12.2v. With car running while loping 12.6v

7. So resistance at ICV is in spec and the plug itself is getting power. I assume since it is getting power that means the ICM is likely good?

8. Throttle position sensor (TPS) has audible click on/off when butterfly is moved.

9. No fuel present on the vacuum line side of the FPR (fuel pressure regulator)


I'm not sure where to look next, possible a proper smoke test. Also have read that the poorly adjusted throttle plate on throttle body has resolved issues for some members. Another possibility is something related to the purge valve allowing unmetered air according to another thread.

Appreciate any input or tips. Thank you.

[Quadrajet]

Did you happen to clean the ICV connector when you probed it? Mine had a little corrosion, which I cleaned off with some electrical contact cleaner and a q-tip, which solved my issue. Initially I didn't notice it.

[richardb]

Did you happen to clean the ICV connector when you probed it? Mine had a little corrosion, which I cleaned off with some electrical contact cleaner and a q-tip, which solved my issue. Initially I didn't notice it.

I will try that tomorrow, thanks.

[richardb]

Update:

Idle Control Valve tests at 9.5 ohms resistance which is in spec I think according to the forum here.

Lubricated/cleaned ICV with WD40, applied 12 volts to it on the bench and can audibly hear it click open/closed.

Cleaned/sanded lightly terminals and plug pins for ICV with 400 grit paper and cleaned. Getting 12 volts with key on.

I think this rules out the ICV, no?

[Quadrajet]

I would think it should. But really once the motor warms up and the ICV actually controls the idle, that's the real test. Also I use a steel brush to clean my pins, just my preference. After I check mine today, I'm going to put a dab of dialectic grease on mine. My part of the country will be getting into our wet season mid to late September.

[tn535i]

My experience is for the most part with e30 eta using the ICM/ICV but I think things work the same. I glanced at your 84 533 schematic and it appears identical to an 86 M20 eta.

Voltage is not the critical thing to measure since this is a time based on/off modulation you need to measure current through the loop. The spec calls for a current that is set with the ICV bypass screw when the engine is warm with no load. I think it should be close to 480 mA per spec if like the e30 eta 2 wire ICV/ICM. If you cannot get to that I think it means the ICM isn't functioning properly and this is what I might suspect. I've seen them fail where they just don't deliver the proper current level anymore leaving the idle high and fluctuating. When it gets really bad it will bounce off the idle speed regulator. Current closes the ICV bringing the idle speed down, more current means more closing time applied to the valve.

I don't recall if it is green or red box that is better ICM but one of the two suffered from both poor solder joints and failing component in the ICV driver circuit that would not allow it to achieve proper current. I had this problem on our old eta and have seen it on others. In fact I re-soldered an old ICM once and got it to work again for a while but was able to replace it with the newer version I got from a wrecked car for $20 so I have one of each still.

You should be able to plug the ICV connector back on with only one connection made (you'll see it's designed like that and has no real polarity) then complete the circuit with your DMM set up in DC Amps. If you can measure that current you can get better diagnostic.

It would be good if someone else can verify the current desired although I expect my memory is correct and the 533 is the same.

[richardb]

My experience is for the most part with e30 eta using the ICM/ICV but I think things work the same. I glanced at your 84 533 schematic and it appears identical to an 86 M20 eta.

Voltage is not the critical thing to measure since this is a time based on/off modulation you need to measure current through the loop. The spec calls for a current that is set with the ICV bypass screw when the engine is warm with no load. I think it should be close to 480 mA per spec if like the e30 eta 2 wire ICV/ICM. If you cannot get to that I think it means the ICM isn't functioning properly and this is what I might suspect. I've seen them fail where they just don't deliver the proper current level anymore leaving the idle high and fluctuating. When it gets really bad it will bounce off the idle speed regulator. Current closes the ICV bringing the idle speed down, more current means more closing time applied to the valve.

I don't recall if it is green or red box that is better ICM but one of the two suffered from both poor solder joints and failing component in the ICV driver circuit that would not allow it to achieve proper current. I had this problem on our old eta and have seen it on others. In fact I re-soldered an old ICM once and got it to work again for a while but was able to replace it with the newer version I got from a wrecked car for $20 so I have one of each still.

You should be able to plug the ICV connector back on with only one connection made (you'll see it's designed like that and has no real polarity) then complete the circuit with your DMM set up in DC Amps. If you can measure that current you can get better diagnostic.

It would be good if someone else can verify the current desired although I expect my memory is correct and the 533 is the same.

thanks for your reply, good call with the mA reading, I will check that now. I did see that the eta m20 stuff might be close or the same, however the ICV looks different than my brother's 86 325es and mine does not have any adjustment screw on it.

[tn535i]

I did know there were some differences but didn't know the 533 had an ICV without bypass screw. I wonder if it could be interchanged? In any case it may mean you could only adjust the current by adjusting the t/b closed position. Adjusting the t/b closed position may be the way it's done on your car and the adjustment is easy on an M30 whereas the M20 they really changed the t/b and made it very difficult to adjust the TPS or the t/b stop.

My guess is that the M20 eta and later cars there was a more deliberate attempt to make any adjustments to the t/b or TPS part of any service so they needed to add a bypass screw in the ICV. That is a far simpler procedure compared to adjusting the t/b and then resetting the TPS.

If you want an extreme idle adjustment procedure/challenge try an S38 motor.

[richardb]

Ok, I placed my meter in between the circuit of the idle control valve (between the valve connections and harness that connects to it) to read the flow of current. With the engine hot and idle loping, the current reading was bouncing roughly from 200-320 mA up and down.

[richardb]

Ok, I placed my meter in between the circuit of the idle control valve (between the valve connections and harness that connects to it) to read the flow of current. With the engine hot and idle loping, the current reading was bouncing roughly from 200-320 mA up and down.

Would this 200-320 mA rating (vs. 480) indicate a bad idle valve or bad idle control module?

[Quadrajet]

If I'm interpreting your description correctly, that you have your meter between the ICV and it's connector, that it would indicate a faulty module.

[richardb]

If I'm interpreting your description correctly, that you have your meter between the ICV and it's connector, that it would indicate a faulty module.

Yes that is correct, I placed the meter between the ICV and it's connector. Ok. So the relatively low power rating indicates the issue is with the ICM itself. Does anyone know how that ICM typically fails? This one resistor looks burnt: (photo 3)

https://bimmerzeit.com/2020/08/17/533i- ... ol-module/

I have ordered a rebuilt ICM. I will report back once it is installed.

[tn535i]

Re-reading things... I'm not 'sure' how the 533 behaves if for example you unplug the ICV. Maybe not like an M20 eta that will cycle up to about 1500 rpm and the DME may not have the same fuel cut. So when you said unplugging either the ICM or ICV made no change to the loping idle it raised a ? I think you need to confirm the TPS is sending the idle signal to the ICM electrically. Voltage at pin 1 should go from hi (maybe 5v) to lo (0) when the switch closes. I've seen several switches that click but have a broken solder connection inside that no longer ground pin 1 when closed. They are not that hard to prise open & repair... and the micro switch itself is really robust. I've repaired 3-4 of these.

So your low current could be the condition the ICM goes to when the TPS is open and there is no idle signal. That's how these things work, even the 3 wire types. An off idle position (lower duty cycle and therefore less current) and an idle control modulation mode that applies more current (if adjusted properly).

Now if the ICV is 9-10 ohms (which seems correct to spec) then ohm's law would suggest about 1.25 amps current if the valve went to 100% closed. You might test that using jumpers from battery to fully close the ICV as one last test of the ICV function. A 9v battery might even work for this test. If you do that while the motor is running it should slow down and probably struggle to keep any idle speed, maybe die.

The pics of your ICM do show what I would assume a power resistor (because it is bigger) that has been very hot. Might be normal for one that has been in service 30+ years. I don't think getting a rebuilt ICM is a bad idea but be prepared it's not the base problem.

[richardb]

Thanks again for your detailed reply. Very helpful.

I did apply 12v power from my battery tender to the valve while running to see what would happen, it immediately stalled out the car. Tried this 2-3 times just to be sure.

I will take a look at the TPS in the mean time for the voltage reading. The car does drive and rev fine however, wouldn't a faulty TPS show more drivability issues?

Thanks for looking at the ICM photo. I would like to see if anyone has an image of the typical failure since it is so common on the early cars. Oh yes, I am prepared for money spent not to be the final repair, lol. It never is easy, is it?

[richardb]

For those that want to follow along in video format, I've documented all of the tests and diagnosis so far here on my YouTube channel:

https://youtu.be/aDzlJtgZUAg

[tn535i]

I will take a look at the TPS in the mean time for the voltage reading. The car does drive and rev fine however, wouldn't a faulty TPS show more drivability issues?

lol. It never is easy, is it?

It may simply be your TPS now that I've thought through it all again. It has an idle input and WOT input. So it simply signals the engine controls when to be in idle mode and on the WOT end of things it may signal things to go open loop and max power regardless of the O2 sensor. No effect on drive-ability between those bookends. Your WOT may very well be working even if the idle input is not. Lack of idle signal may explain everything on your car. I'm used to later cars where loss of that signal generates an even more erratic idle behavior than you describe. Your car may be first generation computer controlled idle and set up closer to a car with no idle control.

The solution is often simple but finding it can be the hard part, especially when not sure where to search.

[richardb]

Ok, I measured the resistance at the pins on the throttle position sensor (TPS) tonight. Here are the results:

Leads of the meter touching together: 00.7 ohms (continuity)

I believe pin 2 and 18 regulate the idle circuit, 3 and 18 regulate the WOT (wide open throttle) circuit.

TPS pin 2 and 18, closed throttle: 00.7 ohms (continuity)
TPS pin 2 and 18, open throttle: goes to 0L

TPS pin 3 and 18, closed throttle: 0L
TPS pin 3 and 18, open throttle: still 0L
TPS pin 3 and 18, wide open throttle: still 0L

Another data point - the car idles at 1250 rpm on cold start with the TPS plugged in. I unplugged it and the car idled at 1750 rpm. For what it's worth.

If anything, it appears that the WOT circuit is not functioning or contact is dirty on the TPS. But according to the readings, the idle circuit is working.

[richardb]

Updates! The rebuilt ProgRama idle control unit (green guy in the glovebox) came in today and I installed it... and....

The idle has stopped surging! To take it a step further, I again placed my meter in between the idle valve and it's wiring to get a reading on the mA of current running to it while the car is running. The reading was 470 mA which appears to be within spec (480 mA spec mentioned in this thread). That is substantially more than the 2-300 it was seeing with the original module.

The bad news is, the idle is still too high. It is currently sitting at approx 1,200 rpm even while warm/hot. I wonder if the two-pronged thermo temp switch in the thermostat housing is the culprit. I decided to take a resistance reading across it's two terminals. Here are the results:

Multimeter leads touching together: 00.3 ohms
Car cool-ish (ran earlier than sat for a few hours) 00.3 ohms
Car at running temp and running: 00.3 ohms

It appears that based on this reading (unless I am doing something wrong) that the thermo temp switch is bad resulting in the car being stuck in open-loop (cold idle all the time). What do you say, guru-san?

[richardb]

Jumped the two wires together from the two-prong temp sensor (idle control sensor I believe) and this made no difference, hot or cold.

[richardb]

Here's the video for part II of this diagnosis series. Installing the ICM and testing the voltage output again after the rebuilt unit is installed.

https://youtu.be/Oltp2uAcJlQ

[tn535i]

There are two temperature sensors for idle control. An air temp sensor that closes (near 0 ohms) below 0C (32F) and a coolant temp sensor that closes above 113F (45C). The first can be hard to spot and not sure where it would be on your car, maybe inside the firewall A/C cavity. Since it should be open (ohms = infinite or OL) when the air is warm you can try simply unplugging it. The coolant temp sensor should be near the t-stat and should be open circuit engine cold but closed when it warms up as you would get by jumping it.

Both are easy to test. I'm not 100% sure what each does to idle speed, but certainly cold outside temp or cold engine would speed up idle a bit. So maybe each condition adds a couple hundred rpm?

Glad the new ICM seems to be doing it's job now. But it would still make me suspicious you are getting as high of an idle as you are with the current close to spec(if 480 is right). This makes me think you now need to close down the throttle body a little bit and see if it brings idle down into proper speed range.

I said earlier I'm not sure how the 533 idle control behaves exactly since it is a little different than an e30 eta. Perhaps since there is no bypass on the ICV you do need to tweak the t/b to get the idle control balance correct. My assumption would be that the ICM/ICV current is about that 480mV with a warm idle and no load. Then when any other condition like low temp or load occurs the current drops off allowing the ICV to open more and boost idle.

[richardb]

Thanks for the reply. I definitely should pick up an e28 Bentley manual.

My brother has an 86 325es and sent me a photo of the e30 Bentley as something I overlooked- apparently the early idle valves (black plastic body) as I have on my car have an idle adjustment screw internally. So I have to pop off one of the hoses and then adjust the screw inside. They later changed the screws to be adjustable without removing the valve (silver metal body valves).

However Turning the bypass screw clockwise will increase control current; turning counterclockwise will decrease current. Since mine is reading 470 mA at warm idle with no load that is within spec. However that is only correct spec for a correct idle speed...

From Bentley:

"With the engine running, the current reading should be 460 +or- 10 mA at 850 + or - 50 rpm on cars with manual transmissions and 470 + or - 10 mA at 850 + or - 50 rpm on cars with automatic transmissions (reminder, my 533i is an auto transmission).

[tn535i]

However Turning the bypass screw clockwise will increase control current; turning counterclockwise will decrease current. Since mine is reading 470 mA at warm idle with no load that is within spec. However that is only correct spec for a correct idle speed...

From Bentley:

"With the engine running, the current reading should be 460 +or- 10 mA at 850 + or - 50 rpm on cars with manual transmissions and 470 + or - 10 mA at 850 + or - 50 rpm on cars with automatic transmissions (reminder, my 533i is an auto transmission).

Interesting and I learned a new thing about that ICV. So IF your idle control is working, meaning it is bringing the speed back to the same point when you adjust that screw it has to be either a sensor telling it to go to higher idle or the tachometer is wrong. IDK what speed the outside air or engine temp sensor will tell the idle to go up to but 1200 seems a little high. However, my M6 idles about 1200 when I first start it up although my 535 and our older e30 not quite that high. So it's possible still I guess it's idling in warm-up speed. The CTS also gives signal to the ICM but if your car is running well it should be good. You can test this one also and easily with a meter.

Their is an A/C input to the ICM but I think that only kicks it up 100-200 rpm. Does the idle go up even more if you turn A/C on?

I think it would be obvious if the tach were wrong. Big difference the way 800 and 1200 rpm behave. Since you have an auto it probably lunges ahead if you do shift into gear. Since the ICM get's speed direct from the sensor on the flywheel that makes me wonder...was there an ICM module for a 4 cylinder car and did you get the wrong rebuilt module? "IF' it thinks it's a 4 cylinder then it would try to hit 1200 rpm. 6/4*800=1200. Yet another strange possibility is your idle really is 800 (behaves normally just reads high) but you have a replacement tach somewhere along the line with coding plug from a 4 cylinder car. I'm not sure 4 cylinder e28's made it to the US so this seams almost absurd suggestion.

[richardb]

Thanks for the further comments. I don't believe this is a tachometer or cluster issue, it definitely sounds too loud and too high of an idle.

Turning on the AC (which works, by the way) reduces the idle to a little over 1000. So the load the AC presents definitely brings down the idle. The idle control module appears to be the correct one, see it here:

https://youtu.be/Oltp2uAcJlQ

[richardb]

Video update, part iii: https://youtu.be/zQ1I-a1vtyU

I have replaced the (yellow/brown) two prong coolant temp switch part number 12631279716 with the updated violet one part number 12631279719. This made no change in warm idle.

I also purchased a Facet brand throttle position sensor/switch (made in Italy) and installed and adjusted that. This made no change in idle.

Furthermore, while allowing the car to warm up for 10 minutes after the coolant sensor was replaced, the car randomly stumbled, idle dropped, and then it went back into the surging for a minute until I shut it off.

Really puzzled. Could a bad coolant temp sensor be causing an issue? I'm grasping for something here...

[Jetblue]

You bought the wrong ICM. Do not buy re-manufactured from Programma. They are garbage. I made the same mistake and had the same results. Go ahead and spend the $400 on a new one from BMW and the problem should go away. My problem was solved with a new ICM from BMW.

[Jetblue]

Oh and don't bother trying to clean the ICV. Every time I have has only made the idle worse. Buy a new one. But that part is no longer available through BMW or anywhere else. I notice the 528e ICV is still available and i'm not sure what the difference is. I was planning on buying one and attaching it to my 533i and see what happens.

Same with the ICM, although the ICM for the 533i is still available through BMW, but per the Bentley, the wires that go into the ICM for both the 528e and 533i are still the same.

I have both a 528e and a 533i. My idle was bad on my 528e first. I bought a $100 piece of garbage from Programma and my idle stopped fluctuating but it was very high. I then changed the ICV to a brand new one and still the same problem. I then bought the dealership ICM and my idle has since been perfect. So buying both parts together is the way to go.

In case anyone else reads this and is considering buying a reman control unit, don't. They are scams. The companies only stay in business by suckering poor saps into thinking its a good deal because it's cheaper and by refusing to accept a return because after it's been installed they can't be returned etc. I have plenty of experience in this regard.

I'm not sure who came up with the idea, but the ICV is a very delicate part and the wrong cleaner or any cleaner improperly applied can do more damage than good. It may be different on a 535i but on the 84 and older cars it's never helped me.

[richardb]

Thanks for the reply. Can anyone else share their experience with a bad rebuilt ICM? Ugh, $400 for a new one though... it just seems on the high side. Maybe a known tested used one... anyone here have a known good used module?

[Jetblue]

Spend the money on the Bentley manual and run the tests It describes for the ICM first. That way you will know for sure.

But from what you described I would bet it’s your ICM. Equally important is the wiring to the ICM, which the tests will also help you rule out. Bad wiring would cause it to act like the ICM is bad.

[tn535i]

Spend the money on the Bentley manual and run the tests It describes for the ICM first. That way you will know for sure.

But from what you described I would bet it’s your ICM. Equally important is the wiring to the ICM, which the tests will also help you rule out. Bad wiring would cause it to act like the ICM is bad.

I think he's already tested things AND he is getting darn close to the proper current through the ICV. With bad wires there would be no current. OP also applied 12v to the ICV and it stalled or about stalled the motor indicating the valve does close. So it is really a mystery at this point why the idle did not come down to proper range.

If the ICM current is correct AND the ICV is moving freely it should have come down in idle speed. Unless something has been altered previously. For example if some sort of restriction (like pennies with hole in it) has been placed inside a hose and the t/b position altered then maybe even though the ICM/V current is at 470 mA and part closed valve it is not having much effect other than stopping the surging. I'm now puzzling on this idea because I think originally the idle did NOT go above the approximate 1500 rpm where fuel is normally cut. On newer engines I know a wide open ICV will do this.

In OP's last post he asked about coolant temp sensors and I don't believe anything makes it run up to as high idle like he has. OP has also tested all those and they check out. But then then OP also said the surging has come back now and that sounds like failure of the ICM again.

If it were mine I would assume at this point anything and everything should be inspected, retested and questioned and try to start fresh. Bentley might help but I think he has all the information needed by now.