E36 Severe Power Loss Under Acceleration - No stalling, No EML, No Rough idle

[ed323i]

Wait.. I now remember a recent, somewhat similar case to yours, also M52B25, from a guy in Sweden. And he finally resolved it by replacing an ignition coil.
Using INPA he could easily see which coil was misfiring (watching live values of engine roughness, per cylinder). Turned out to be a perforated rubber boot of the ignition coil. Here is the (long) thread: E36 323i 1999 m52b25 Misfire and (hissing?) noise when warm .

So, my advice would be to install INPA ( https://www.bimmerforums.com/forum/s...xperimentation ) and watch the engine roughness.. You can also go ahead and inspect all coils.. You can buy a set of new (not-OEM, but they seem fine) coils for $60-$70 if I remember correctly. These include 6 new rubber boots. Eric93 has the ebay link. Here it is: #6 . As you're not in the USA, you could consider getting them from AutoDoc, where I bought my set of 6 coils, which seem to keep on working without any trouble, after 6 months. Brand name: Stark. Total price 90 euro if I remember correctly.

[Thomas525]

Did you get the car jacked up to check underneath?

[CaptainsLog]

I have literally no equipment as I have always just used a mechanic before, so no jacks/stands. Pretty unequipped to do anything, I have a fairly decent socket/ratchet set. I'm going to try reach the crankshaft sensor this morning. I don't have the OBD connector for the INPA software which is a shame, cause I'm handy with a computer so would have been great.

[Thomas525]

Well apart from a multimeter, you pretty much have all the tools that you need to check the sensors and even the guibo. Yes very bad guibo can be responsible for idle roughness and underpowered acceleration.

The multimeter enables you to check sensor resistances that you can compare against the bentley manual specs. You can also test continuity from the ecu harness clamp to the sensor - there may be breaks. You can test the resistance of the fuel pump right from the fuel pump's fuse socket itself. And also check the current running through the pump while it is running, and when you accelerate by flicking the throttle by hand. And anyway, the sensor disconnect test will also rule out the bad sensor if there is a bad sensor. Unplugg it and the car will return to normal immediately.

The guibo - there are two ways to inspect this. First, if your car is not lowered, use the tyre jack to jack up the car, then put the spare tyre under the car next to the jack to protect against jack failure. Then you should be able to shimmy under safely and eyeball the guibo. Best to use glasses or goggles. With a lowered car, you will need to put a stone or something else on top of the tyre to help protect against a collapse.

The second way is far simple. Take a large cardboard piece. Activate your phone's video with flash function and place it on the cardboard with the camera and flash pointing up. Slide this under the car to where you think the guibo will be. Do this slowly to give the camera time to focus on the details. Then take it out and view the video, and shoot again with better positioning you won't get it right the first time. Heck in fact you can do a whatsapp video call to yourself from another phone and look at the second phone while sliding the first one under the car. YOu can even use the forward facing camera to avoid touching the screen but you will need to provide alternative lighting somehow. It would work superbly and get you to quickly look at and screenshot anything suspicious.

If you can't do all that go back to your workshop and have them jack it up. This is of course the very best way to inspect a guibo. You can turn the shaft around etc easily.

You really must do all of the above just to be very sure.

[CaptainsLog]

I started to car up this morning to see how it ran without the MAF connected and not sure if it ran better, but it wasn't running great. Revs still elastic and after a while I smelled something burning, so I turned the car off and opened the bonnet again and there was smoke coming from the left hand side somewhere. What is down there that would cause this? Anyone?

- - - Updated - - -

How it was running this morning. No MAF connected not sure if that made any difference whatsoever

[Thomas525]

Ok my friend. Please listen carefully. There is nothing on the left of the engine that could cause any smoke. Smoke if at all should only come from the right side of the engine due to a valve cover oil leak splashing onto the hot exhaust manifold. I mean this smoke is not a showstopper. But your smoke on the left is. Because there are no oil lines there and nothing hot enough to vaporise oil.

So the smoke has come from an electrical short. This can indeed explain alot of your symptoms. One or more of your sensors is shorting to ground and sending distorted signals to the ecu which then controls the engine badly without realising that something is wrong.

I recently had an obscure problem that was ultimately traced to burnt wiring after much troubleshooting. Here it is :

https://www.bimmerforums.com/forum/s...2#post30193712

The last post on that thread.

You need to identify exactly where the smoke came from. A good flashlight would be essential here. Remove the airbox and maf and start looking and sniffing. And unplugging the maf will not cause this problem. Look at the fuse box as well.

You might have to get the car to the workshop if this is beyond you skills set. It is not something you can train up on easily and is too dangerous to live with while you slowly figure it out. You cannot do a half baked repair either. Unless of course you have another car that you're using, then you can take your time to slowly figure it all out.

But really since you have a workshop nearby - might be a good idea to let them look at it. Don't agree to any repairs, come share what they say and their quotes to repair and people will tell you if it makes sense and is right.

Don't start or drive the car until you have found the source of the short.

[CaptainsLog]

Ok my friend. Please listen carefully. There is nothing on the left of the engine that could cause any smoke. Smoke if at all should only come from the right side of the engine due to a valve cover oil leak splashing onto the hot exhaust manifold. I mean this smoke is not a showstopper. But your smoke on the left is. Because there are no oil lines there and nothing hot enough to vaporise oil.

So the smoke has come from an electrical short. This can indeed explain alot of your symptoms. One or more of your sensors is shorting to ground and sending distorted signals to the ecu which then controls the engine badly without realising that something is wrong.

I recently had an obscure problem that was ultimately traced to burnt wiring after much troubleshooting. Here it is :

https://www.bimmerforums.com/forum/s...2#post30193712

The last post on that thread.

You need to identify exactly where the smoke came from. A good flashlight would be essential here. Remove the airbox and maf and start looking and sniffing. And unplugging the maf will not cause this problem. Look at the fuse box as well.

You might have to get the car to the workshop if this is beyond you skills set. It is not something you can train up on easily and is too dangerous to live with while you slowly figure it out. You cannot do a half baked repair either. Unless of course you have another car that you're using, then you can take your time to slowly figure it all out.

But really since you have a workshop nearby - might be a good idea to let them look at it. Don't agree to any repairs, come share what they say and their quotes to repair and people will tell you if it makes sense and is right.

Don't start or drive the car until you have found the source of the short.

Interesting. What's the course of action here? And what sensor short would cause loss of power etc? The aftermarket CD player in the car died a few weeks ago, started acting weird and juttering sound, I pulled it out to see what was happening behind there and whoever had fitted it did a shit job, instead of using an ISO they had used a chop block and there was wiring everywhere. Anyway, I approached Halfords and asked them if they could sort it out they gave me a quote, I put the CD player back in situ to keep it out the way and the power completely died to it after that. Could the be contributing to a short? No idea what I should do now. I don't have a voltmeter to begin with.

How sure are you of the short?

[Thomas525]

Im pretty sure it is a short. There is nothing hot enough on the left side to vapourise anything. Plastic insulation to some wires are burning. How did the smoke smell ? Did it smell like burning oil ? Or burning plastic? Burning plastic is more acrid and sharp unpleasant. Exactly which spot on the left did it seem to be coming from ? Was it the fuse box area ?

I have no idea which sensor could cause this. My guess is a bunch of sensors are being messed up because their wires are bundled along the way to the ecu. Anyway can you do a quick five second experiment?Unplugg the throttle position sensor. It is found on top of or to the left of the throttle body, adjoining it. The start the engine and see if you can throttle up beyond 2k rpms. It looked like you could not go beyond that.

Within 5 seconds of this shut down the engine. You don't want the short to take control. After that, remove the air box and maf and start looking. You might see burnt wires. Look at the fuse box. Do all that and then reply. No point speculating further here. You have a serious problem more information is needed.

You might have to tow the car to the nearby workshop. Or ask those guys to drive down with some tools after work and pay them part of the tow fee instead let them diagnose the problem at your place. Or if you have craigslist you will find mobile mechanics there, you could describe the problem ask them to come down to help you trace wires etc negotiate a diagnosis fee.

And do get a digital multimeter soon its only ten bucks the kind that we need for the car, nothing more fancy.

[Nanniepoo]

I think before continuing you both should clarify your perspective when saying left hand side. If it's exhaust manifold side your valve cover gasket is leaking. If it's intake side, although there are oil and coolant lines, I don't think they would cause smoke until the car is warm enough or if it's cold enough outside to notice steam from the coolant.

[Nanniepoo]

I realize I kind of reiterated what was already said but think the OP needs to clarify which side is left

[CaptainsLog]

Im pretty sure it is a short. There is nothing hot enough on the left side to vapourise anything. Plastic insulation to some wires are burning. How did the smoke smell ? Did it smell like burning oil ? Or burning plastic? Burning plastic is more acrid and sharp unpleasant. Exactly which spot on the left did it seem to be coming from ? Was it the fuse box area ?

I have no idea which sensor could cause this. My guess is a bunch of sensors are being messed up because their wires are bundled along the way to the ecu. Anyway can you do a quick five second experiment?Unplugg the throttle position sensor. It is found on top of or to the left of the throttle body, adjoining it. The start the engine and see if you can throttle up beyond 2k rpms. It looked like you could not go beyond that.

Within 5 seconds of this shut down the engine. You don't want the short to take control. After that, remove the air box and maf and start looking. You might see burnt wires. Look at the fuse box. Do all that and then reply. No point speculating further here. You have a serious problem more information is needed.

You might have to tow the car to the nearby workshop. Or ask those guys to drive down with some tools after work and pay them part of the tow fee instead let them diagnose the problem at your place. Or if you have craigslist you will find mobile mechanics there, you could describe the problem ask them to come down to help you trace wires etc negotiate a diagnosis fee.

And do get a digital multimeter soon its only ten bucks the kind that we need for the car, nothing more fancy.

Guys, sorry for not clarifying. Left side as in, if you're looking into the engine. I've attached an image (stock image, not my image but same kind) with an arrow as to where the smoke was coming from. The thing is I thought it was just exhaust emissions but they hung around a bit too long to be anything other than smoke and not burned oil, smelled different, a concerning kind of burning smell. The revs can go past 2k but once they do the engine gets weird. After I shut off the video I took for you guys, I gave it some hard revs (4k/5k) and that's when the burning smell was noticed and I was concerned something was on fire.

maxresdefault.jpg

[Thomas525]

Left side should be the side that blinks when you hit the left indicator. Right side, right indicator.

- - - Updated - - -

This is obvious.

[CaptainsLog]

To clarify, it's not coming from the fuse box side it's coming from the side where the OBD sits

[Thomas525]

You have a valve cover leak. When you throttle up, there is more oil splashing around, and leaking out, and dripping on the exhaust. This needs fixing but can wait. You don't have a short. Be very grateful for that trust me. A couple of posts ago, I gave you some suggestions on the sensor disconnect test, fuel system check, and shooting videos of your guibo with your mobile phone. Please do all of that and let us know. And shoot that video with the engine off.

[CaptainsLog]

How would one go about checking the fuel system? Still haven't disconnected the crankshaft sensor. Hard to get to without removing stuff. Could do with the oil filter coming off to be honest.

Quite a bit of white smoke coming out the exhaust when cold starting. Is this normal for cold starts? So much so, that when I'm breathing in the car and the door is open I can see I'm breathing the fumes.

[ed323i]

OK.. MAF doesn't seem to cause your problems..

I think it's ignition related. So, go check all your coils, and especially the rubber boots.. Very easy to do. Remove beauty cover on passenger side of engine, then unscrew and unplug the coils and pull them out. Perhaps best to also replace your spark plugs if you don't know how old they are. If your coils are the original ones, it's a good idea to spend the $60-90 euro and replace them all.

If that's not it, then I still suspect the crankshaft sensor, as I'm 99% sure the engine should run fine without the camshaft sensor connected.

[CaptainsLog]

OK.. MAF doesn't seem to cause your problems..

I think it's ignition related. So, go check all your coils, and especially the rubber boots.. Very easy to do. Remove beauty cover on passenger side of engine, then unscrew and unplug the coils and pull them out. Perhaps best to also replace your spark plugs if you don't know how old they are. If your coils are the original ones, it's a good idea to spend the $60-90 euro and replace them all.

If that's not it, then I still suspect the crankshaft sensor, as I'm 99% sure the engine should run fine without the camshaft sensor connected.

Ok man, thanks. My next POA was to check out the coils. I checked the service history, there's a fair bit but no sensor replacements. The guy I bought it from changed the plugs back in April last year so they should be good, right?

[ed323i]

Plugs should last tens of thousands of kilometers in the M52, so those should be fine.

[Thomas525]

a. Please remove your airbox. Leave the maf in place. Maintain this setup for testing purposes. Replace the airbox if going for a test drive.b. Crank sensor connector is under the intake manifold somewhere. Trace the wire, find it, and unplug that. If you're still not sure, check youtube for illustrative videos for your exact engine - they will be there. With both cam and crank off, your engine should not run. With either one connected, the engine should run normally at all ranges. You can use that to test. c. Unplug the throttle position sensor, it is on the throttle body. Test rpms at idle. If good, go for a test drive. Post results here. d. Do the video of your guibo as I described earlier. Before that, watch some videos so you have some idea of what to expect. Let us know what you find. e. Purchase a simple digital multimeter for $10 somewhere. It is an essential tool for car maintenance. f. After all that, if the problem has not been found, we can test the fuel system.

[Thomas525]

Actually, start with the tps throttle position sensor first. It is easiest. And always plug the tested sensor back in before going onto another sensor.

- - - Updated - - -

When you are smooth with the engine, you will be able to unplug, test the engine, and reconnect all the sensors, including the 2 o2 sensors, within 15 minutes tops assuming the engine can be tested against that problem at idle. Its a good skill to be good at, you will need it in future.

[M3AN]

Stop.

Get the codes read. Anything else is a fool's errand.

[CaptainsLog]

Stop.

Get the codes read. Anything else is a fool's errand.

I got the codes read at a garage. Guy did it for free with his snap-on reader. Popped up camshaft sensor (or crankshaft sensor) but no CEL light. Not sure if CEL is even working, gonna test the cluster this morning, remove crankshaft sensor and see how it fairs after that.

Can I drive it without the crankshaft sensor connected if that is what has failed? I mean, should I be able to test it round the block etc. will it be in limp mode?

Ok, crankshaft sensor is disconnected (quite difficult to reach without removing oil filter cover so no idea how I'm gonna get it back in) but car did not start with it disconnected. Cranked fine, no start. Just like camshaft sensor. What now?

To confirm, ran instrumental cluster test. CEL did not illuminate. MAF pulled CEL was not illuminated.

[Nanniepoo]

Have you checked your coils/plugs? While a leaking valve cover itself shouldn't cause too bad of an issue besides for being stinky, the gaskets around the spark plugs wells can leak, drowning the spark plugs in oil, which can cause issues/misfires/etc. If you notice oil down there, just soak it up with paper towel, get everything as clean as you can, and try again.

I'm sure it feels like you're getting bombarded with suggestions. But personally I feel pretty comfortable saying your valve cover is leaking so I'd start with a known side effect before chasing after sensors.

Side note I don't think E36s in the UK even have CELs, not sure if it's just no bulb or no wiring...

[Nanniepoo]

I'm not convinced a car will run without a camshaft sensor. That's what determines if you're on an intake or exhaust stroke, and therefore injection timing and spark timing. If you're getting codes for the cam/crank sensors those definitely can cause your issues. I could be wrong, but I don't see how unplugging one will help diagnose. There's information somewhere on here on how to test them. I think resistance check works for one, but not the other, best way is with a oscilloscope but most people don't have one just sitting around. I'd start with making sure your plugs are clean, then see if you can bench test the sensors. If it comes to it, I'd just replace both.

I know thomas keeps suggesting look at the guibo but that will only cause issues while driving. Since you're also having problems idling, revving it in park/neutral, the guibo isn't even involved.

[ed323i]

I personally haven't tried to have the car run without camshaft or crankshaft position sensor, but the sensors FAQ on this forum indicates that it should be possible with only one position sensor connected.. Without camshaft sensor there is less details, and the DME will revert to batch injection and wasted spark ignition, which virtually has no impact on performance or drivability (idle will be a bit less fluent without seq. injection). My Megasquirted E28 528i used an M30B35 trigger wheel and crank position sensor, and no cam sensor and ran perfectly fine (wasted spark, batch injection). The other way around (only camshaft sensor connected, crankshaft sensor disconnected), I'm not 100% sure how/if the DME can determine all necessary information to know when to spark and inject at the exact right times.. But, like I wrote, in the sensors FAQ, people say it should work with either one sensor connected (and the other disconnected).

Perhaps someone can let us know if he has tried this (disconnecting one of both sensors and confirming if it still runs fine), so we are 100% certain that the DME can run with only one of both sensors connected.. This is very valuable information, easing the process of diagnosing a bad sensor immensely: Disconnect one sensor -> does it run fine -> you know for sure that the other sensor is 100% fine. Does it stop running or does it run very badly -> you know that the other sensor is bad.

Correct: European E36s don't have a CEL light connected in the instrument panel. So, it will never light.

Update: Looked it up. Engine should (theoretically) be able to run with just the camshaft sensor. See https://mechanics.stackexchange.com/...d-a-cam-sensor . The one disadvantage of running with just the camshaft sensor, is the lack of resolution, as the crankshaft wheel is a lot higher diameter, and has more teeth, so the DME won't be able to detect misfires. Also, the crankshaft wheel is mounted directly on the crank, so there is no play, while the camshaft sensor is mounted on the camshaft which is driven by a chain (or belt on older BMWs) which has some slack, and so, is less precise.

So, all in all, knowing that the MS41 DME is quite advanced, I think it's safe to assume that it should run with only one sensor, doesn't matter which one:
* Only crankshaft sensor: You lose sequential injection, turns into batch injection (idle little less smooth, perhaps <1% drop in fuel economy?). You lose sequential spark, turns into wasted spark (each spark plug fires twice during each combustion cycle, one is wasted, but has no influence)
* Only camshaft sensor: You keep seq. injection and seq. spark, but you lose some diagnostic capabilities of the DME, like misfire detection

Well, that's the theory. Would be nice if someone could confirm it, by disconnecting the sensors on a perfectly running engine and then let us know what happens.

Update 2: It's confirmed. MS41 runs perfectly fine with camshaft sensor disconnected: http://www.romraider.com/forum/viewt...p?f=42&t=14259 . Quote: "The car starts and runs but CPS doesn't work with the new DME. Engine acts like if CPS was bad and it gets better if it's disconnected altogether.".
Now let's see if I can find the info on running without crankshaft sensor (and only camshaft sensor connected)..

Update 3: Another confirmation: MS41 runs perfectly fine with crankshaft sensor disconnected: https://www.bimmerforums.com/forum/showthread.php?2000171-M52-engine-runs-with-CPS-disconnected-WTF . Quote:
"Crankshaft and Camshaft position sensor disconnected, engine only cranks and will NOT start.
Crankshaft sensor connected but Camshaft disconnected, engine starts fine.
Camshaft sensor connected and Crankshaft disconnected, engine STILL starts fine."

Conclusion:
We can now safely apply this rule:
Disconnect one sensor -> does it run fine -> you know for sure that the other sensor is 100% fine.
Does it stop running or does it run very badly -> you know that the other sensor is bad.

Also: It's of extreme importance to only buy the original Siemens OEM part (double check this, many garages just install Delphi or Hella or other brands, and these don't work). There are some big design differences in the original Siemens part compared to copy-parts. From the last link, another quote:
"In my situation, the crank sensor is located in the front of the engine, just below the coolant hose connection and above the timing flywheel. What I recorded was a loss of signal amplitude due to an increase in sensor temperature. Basically, the copy replacement part couldn't withstand the rising engine temperature which in turn would affect the sensor's internal resistance, thereby causing signal corruption and periodic loss of timing to the ECU. From cold, the engine would start perfectly, but after between 15 / 20mins (with bonnet closed) the trouble would set in. The copy sensors 'soak up' the heat as opposed to resisting thermal build up.

How does an OEM part cope with this? Well, having carefully stripped the original part previously, I noted that the main body of the sensor encapsulated the inner sensor component in what can be described as a type of 'ceramic' material. This is what is the difference between the expensive OEM part and cheap knock off part.
So then, by disconnecting a faulty sensor, we give the ECU the 'nod to fully commit to using it's default internal reference values. The engine does labour somewhat in starting, but it does start eventually. Keep the heat-absorbed copy part in situ and you'll not get that engine to full running ignition."