Mystery cold hard start issue Continues – 89 525i M20

[MBman]

If I had hair to pull out! I would!
So here is the situation , car has 70,000 miles on it. Had been sitting with little driving so over the last month or so I have been sorting out the issues and fixing as I go. Many service items needed attention and I have completed the following

all fluids and filters including fuel filter
valves lashed to spec (X2)
R/R O2 sensor
Eliminated all vacuum leaks
R/R Engine temp sensor ( blue top)
R/R plugs
R/R rotor/cap
R/R injectors

she has gone from a rough , low power , gas eating car to a smooth as silk engine, with plenty of power , and much better gas mileage. Once warm or for that fact any time after the first ( cold start) of the day she runs well…..perfect
Here is the scenario

Come out after a cold night , 40 or so in the garage. Car cranks and cranks , them will fire , but I really got to work on it. This will sometimes take a full minute or so of cranking.
The exhaust once started smells pretty rich to me
I not 100% sure it is the cold temp , or just sitting that is the issue
Once it does fire the next 60 seconds or so , perhaps two minutes , if I press on the gas she wants to die, like cars of old that had choke stuck open

After this - she runs perfect. Idle is rock steady at 800, no vibrations, sounds great . I can start and stop it all day long

I am at a loss, I have searched but most items that were thought of as culprits I have adressed.

I would appreciate any ideas or things I should check ???

[bd92]

I would say clean the idle control valve, (it cured a rough idle when cold for me), and if you havent already change the coolant temperature sensor, which i believe is difference than the engine temperature sensor but i'm not ENTIRELY sure.

[MBman]

both of those tasks have been completed - thanks, ICV is clean and functioning, sensor ( s) new

[THEhomelessONE]

I have the same problem in my 89 325i, just changed the plugs to the oem replacement ngks. Once the cars warmed up she runs great but cold starts take a few minutes. Changed the o2 as well. Also played with the AFM and the problem is still there, swapped distributors with my e23 and nothing helped.

[LouieD]

I know you said the blue sensor has been changed but I would check the new one to make sure it's within spec (new parts have been know to fail) and two check the wiring that goes to and from that sensor. Also pull the plugs and check the color of the burn. Wrong heat range or improper gap could also contribute to the issue. All three of my 3 e34's (m20, m30 and m50) have been through this and it was always the blue sensor or plugs. To facilitate cold starts until you get it sorted out, crack the throttle slightly open before the first crank to allow a bit more air and hopefully balance out the mixture.
Good luck.

[MBman]

ok will check spec on new sensor and pull a plug or two to see the tip color, thanks
what about fuel control regulator or CPS??- any thoughts there?

[LouieD]

From my limited experience CPS wouldn't let you start let alone drive ok. Fuel pressure regulator is easy enough to check. See if it holds vacuum or remove vacuum line and see if any fuel comes out. I had a bad FPR on my m30 and I found it because the vacuum line was pissing fuel out of it.:eek: I just can't remember it causing hard start issues. Sorry short term memory burned out in the 80's.

[520 Driver]

I have the same issue and my brother had it as well. I am changing the oil to 5W-30 next chance I get. First start since it got below 15 degrees was today. Are you running 5W-30?
A co-worker from Germany said he had a problem with hard start and rough running when cold. It was from a slightly warped head (losing compression when cold, expanded and stopped when warm) and they skimmed the head and it was fine. I sure hope we do not have those problems. Just my 2 cents.

[MBman]

thanks again, I do not think it is a compression issue, she runs perfect after these hard starts, even when not up to complete temp. I am running 5-30 right now. Yes I think the cps is fine, I dont think it would run at all and just wondered if they ever fail "from time to time" - probably not . bently has a test, just gotta find it ( in another thread) Right now leaning to FPR , or coil. I cant test coil becasue my meter does not measure enoungh ohms ( at least that is what I find in bently)
I am going to remove some plugs to check tip color and check specs on new sensor.
I hope I have time. Work gets in the way of having fun. If you can call sorting out an old car fun...:) I do for the most part. Thanks for the info and I will keep you all posted

plugs look good, sensors ohm correctly, coil -...hmmm well that is another issue see new post??

[sailor24]

CPS test in Bentley's has the wrong values. I replaced one once only to find the values were the same. I forget what the issue was but I think they flipped the values for pre 91 and post 91. Don't remember. When it is warm out does it every crank for a bit? Have you tried unplugging the computer temp sender?

[MBman]

that is interesting, I am leaning to just replacing, temp sender is new and ohmed to specs , what would unplugging tell me?
coil is good and ohmed to specs

[bolloc]

that is interesting, I am leaning to just replacing, temp sender is new and ohmed to specs , what would unplugging tell me?
coil is good and ohmed to specs

Yet in another thread you said you had zero resistance in the coils' primary.... did I misunderstand?

[E34x2]

My M20 will do something similar but mine is completely random. In fact the female lawyer I got the car from got rid of it for this very reason (for $300)! That being said, these M20 motors do run extremely smooth when they run right, they are one of my favorite engines. :) From what I've discovered from my research this appears to be a systemic issue with M20's and no one seems to have a good answer/solution. You've gone through all of the typical causes; tune, air leaks, etc which is the correct way to start. Did you replace the injectors or just the seals? I have gone through everything on mine as well and am beginning to suspect the wiper contacts in the AFM but can't confirm it. I would agree that CPS is likely not the issues. From my experience they either work or they don't and when they do fail while driving it is usually when HOT not COLD.

Is your car an automatic or manual? Reason I ask is the M20 uses two different TPS's based on the trans. 5sp has a variable TPS while the automatic TPS is just an idle switch/WOT switch. Might be with making sure the switch is closed at idle.

Another thought is possible fuel pump, it is after all an electric motor and if the pump doesn't startup, no fuel = no start. My pump was starting to get noisy so I replaced it with a Pull & Save pump which is nice and quiet. Can't say it runs any better but at least it doesn't sound like a rock crusher in the fuel tank. If you have access to a fuel pressure gauge, it would be worth getting a delivery pressure reading when cold. IMHO

And one more consideration, are you checking the sensors at the sensor or at the ECU? The readings are only as good as what the ECU is "seeing" (this is what they taught us at the factory schools I attended). You can have a correct reading at the sensor/component but have an incorrect reading at the ECU. That being said I'm not aware nor have I seen BMW's of this vintage with wiring harness issues. Just something to consider.

My planned solution is a Miller MAF and War Chip when funds allow. For now I just live with it and limit other people (wife and son) from driving it to much.

[MBman]

yea it is frustrating, i am going to do cold fuel pressure tests today and will let you know. The CPS is seems easy to R/R so I may just do that since it looks original and at 21 years seems logical. I always have tested the senders not at the ecu, good point .
It is an automatic and the tps is on or off as you say.

I will let you know what I find

[E34x2]

IIRC running the CPS wire behind the water pump pulley is a major PITA but given that the sensor is 21 years old may be worth the effort. Keep us posted.

Been sitting here thinking about this whole deal and our similar issues and started more searching. Came across this thread from the E23 group.

http://forums.bimmerforums.com/forum/showthread.php?t=1392003

If you scroll down the thread he has some interesting pics. Think I may open my ECU and have a look see....

[MBman]

very interesting, I am an old mercedes guy and deal with vintage ECUS all the time, I can tell you from experience they seldom go bad. OTOH I resolder the temp control units pretty regularly , so he may be onto something,and I will put this in the back of my mind for future reference. Mine seems to be related to ignition , as I have ruled out everthing I know to rule out
OK here is the deal today - I checked all fuel pressure tests. Fuel pump is spot on, residual pressure after shut off is well within specs. Coil is OK. Made sure TPS was OK and it is. ( as I was waiting to see if the fuel pressure dropped I changed out the tranny and engine mounts - badly needed - what a difference - no vibrations any more)
So after all this I fire her up and the thing cranked an cranked, then caught, smelled rich, 120 seconds later , ran perfect
CPS on order , dont know what else to change out, everything is within spec.
I am open of course to opinions

if you open your ecu let me know

when you say ECU , you mean the ICU -part number 12141748264

[Trevor M]

Was reading through your thread and I know your pain. I experienced this myself, not with a 1989 M20 but on a 1981 M30. I know, not the same engine or fuel management system but close enough. I had the exact same symptoms you described. If the car sat overnight and it was cold enough, the engine would refuse to start. When I cranked it over, it sounded like it wanted to fire but wouldn't catch. If by chance I could get it to start and it warmed up, I could start & stop it all day and it would run fine. But if I parked it over night, I knew just by the temperature whether it would start or not. I had the factory manual and checked everything, much like you're doing. I could find anything wrong at all. When the car would start it ran perfect. Hunting for the problem went on for a few years up until the point where I threw my hands up and took it to a shop. That was a waste because they couldn't re-create the problem and changed out a perfectly good starter. Well, maybe it was worn out after all that cranking.
I knew it wasn't fuel related and had to be something with the ignition. I swapped cap & rotor, plugs and nothing helped. Then one day out of frustration I pulled the wire set and checked them with a ohmmeter. I found the spark plug boots on a couple were open! I guess there is a resistor inside the boots of 1.5k ohm and a couple were open circuit. When it was warm enough, the spark would jump the break and the engine would run but if it was cold enough, no deal. Once I replaced the ends (they were replaceable on the M30) the problem went away.
So my advice to you would be pull and check your wire set. I'm pretty sure the M20 uses the embedded resistors in the spark plug boots. Ideally, get a wire set from a wreckers to swap and test. Hope that helps.

[MBman]

Trevor , that is very interesting and will ohm out my wires, although I believe they are new or nearly new, or at least they look that way. I just need to find the values for the resistance.
Today she ran perferct, started perfect , 23 degrees. So strange
I did talk to a local BMW wrench , who seems pretty knowlegeable, and the first words out of his mouth was CPS.
So I will keep everone posted for now , my next move is the CPS
On another note
One thing I do notice and suggest to anyone with the famed E20 engine vibration - get new mounts - tranny and engine, the before and after is amazing. Perhaps more so in mine since the PO actually had the wrong engine mounts installed

Ok so bently says wires should be approximatly 0 ohms, as in continuity??
that cant be right can it ??/

[THEhomelessONE]

I've tried the spark plug wires from my M30 and it didn't help. She usually starts up a lot quicker now if I floor it while I crank. I just have to play with the throttle after that for a few seconds, keep the revs up for another few seconds and she idles fine. Same thing here, it sounds like it's catching when I crank but doesn't. Smells very rich too. Would a clogged cat cause these symptoms? I think it may be my head. It has slightly overheated before and once the motor is really warmed up it won't smoke white but I thought that was just the condensation in the exhaust. It's definitely more white smoke than my 318is while it's cold. Has anyone tried a compression test?

[MBman]

I dont think a CAT plug would cause these symptoms, the O2 sensor is pre cat , so I guess if clogged it may cause some issues, but not a no start. compression is a possibilty but if you have bad compression , you have it all the time , so why would the start be hit and miss. I am sticking with the electrical side of the equation. Trying to find the wire set resistance, and then onto the cps.:) probably be posting back with the same issues , lol.

[THEhomelessONE]

But if your head is slightly warped wouldn't there be a large gap between the compression of the motor hot vs cold? Lol just a hunch idk.

[MBman]

perhaps, but I would think other symptoms as well be present

[THEhomelessONE]

Any updates?

[MBman]

well I have been continuing with the diagnostics, ohmed wires , they are OK, battery is OK, coil and fuel pressures as I mentioned - OK
One thing that is interesting is that I have been waiting 5 seconds in the "on" key position and I have only had one "no start" I dont know if that is coincedence or what. I know that would point to fuel pressure, but all tests ( according to Bently ) are well within specs. I still say it is electrical, so CPS R/R is planned this weekend. I do have a slight miss at idle and I have read that a CPS going bad , may cause a rough idle. Other than that she is running like a beast, strong pull, good performance.
I will keep you posted

[520 Driver]

My 1995 520i is an M50 but this problem exists for me as well. When the temps got up to about 25 F my no-starter started rough and after about 2 minutes of feathering the throttle it idled okay and I drove it to work and home then changed the oil to 5W 30. It has gotten into the teens at night but not to 5 degrees or anything. It starts, runs and idles fine. The plugs are new this summer. Problem solved? I doubt it. I am concerned about more no-starts.

I remarked about my co-worker and the skimmed head and my brothers head leak. My brothers car was losing coolant (but had no leaking) so there is the additional symptom. I have none of that.

Could it be water in a fuel filter? Maybe the fuel keeps it from freezing but in very frigid temps it freezes? I know I'm clutching for straws but I really want to know.

[MBman]

I am running 5-30 , I dont think that is an issue at all , I guess compression is a possibilty but since mine may do a no start - hot , or cold or in between seems to me that would be ruled out. Water in gas, perhaps , but again I dont find mine to be temp dependant , other than it happens more when cold... crazy , I tell ya , crazy

[MBman]

holy crap , it took me an hour to unplug the freeking CPS from the underside of the wire carrier under the manifold, that wire clamp thing about got the best of me, finally go it but had to undo a lot of stuff to semi turn it so I could see it. Replugged and ran the wires but ran out of energy to finish the job, tomorrow is another day
I did ohm out the old one and it was under limits, new one tested higher and within specs. I am optomistic
We will see , providing I did not screw anything else up twisting and turning things

What cant cars be easy:)

[Binjammin]

Don't you just push on the middle of the metal clip so it bends, then pull the connector apart?

http://www.dbcperformance.com/v/vspfiles/photos/1287013003-2.jpg

The metal bar there, the m20 cps connector should be very similar, push the middle of the metal bar towards the plastic and it will bend the locking clips out, then you can just separate the connector halves.

[MBman]

yup it is but I could not get it to work, had to take it off and then it separated , may just have been the angle that I was trying to work at, in any respect it is a 'mouth holding " procedure :)

back together and fired her up, the idle is smoother that is for sure, but the no start - well we will see....
Went ahead and checked the air intake sensor and flap for proper function , all ohmed OK.
I basically have done EVERYTHING I know to do or read to do short of the ecu solder reflow and to check to TPS for close /open. My gut feeling on those items are that they are OK
Have I missed anything???
this car right now is the smoothest running car I have driven ,purrs like a kitten I just want it to start when I want it too.

[MBman]

:mad well the dreaded "no start sometimes " remains . :eyecrazy:eyecrazy, I am in hopes the motronic just needs to learn the new CPS, but not hopeful there
I am at the end of my testing , replacing.
It did not take as long to get her started and the idle is better, so everything i do seems to make a bit of difference. Took her out and really tromped her a few times. Runs like a scalded dog:stickoutt, so it has to be something in the start sequence.

I may try to find a ecu to take apart and resolder

you kind folks have any other thoughts:help:help It may be just the way it is
Some thoughts
I know on my old MB's the relays sometimes get old - is there a relay I could check that is in the start sequence.
Even though my fuel pressure tests were good, could the pump not be strong enough for the start sequence??
TPS - ?? could this possibly be an issue , it seems to pass the tests but during start - could it be sending bad info to ECU?
:help

[Binjammin]

You could have a fuel pump that's starving for fuel on startup, or maybe a clogged filter. Maybe even a very slow leak in your fuel system not leaking fuel out, but air in, taking a few seconds to return the air to the tank.

[MBman]

filter changed last 500 miles, fuel delivery could be an issue I guess, although jumpered the fuel pump and guage shot right up at specifications, running pressure was good as well as residual pressure.
rechecked TPS and it is fine. ECM - perhaps????
I went through all the Bently tests for every elecetrical and sender test - all good.
I have not tested the relays in the E-BOX ( not sure you can)
Do BMW dealers have a way to check ECM ( DME) ?? Referred to in Bently but not clear

[LouieD]

I'm reaching here, but what about a bad fusible link? Just thinking as these cars DO NOT like any variation in amperage/voltage. Could a bad/cracked link cause cold start up issues? Check this out http://bmwe34.net/E34main/Maintenance/Electrical/Fusible_link.htm

I know that the symptoms described on this page are different, but that last "hidden" non-service-able link is described as being related to the DME and it's voltage input.

[MBman]

thanks for that, I will check it out. I got to believe this is electrical , or some input fault, everything else mechanical checks out.
there are a few ecus on ebay pretty cheap right now, may pick one up to take apart and take a look see, you know just for the learning curve.:stickoutt

I am determined to figure this out !:eyecrazy

[E34x2]

When my fusible link cracked the car just stopped. Mine was "inside" the main cable next to the battery but yours may be different. Ended up installing a fuse holder out of a MB 126 chassis as I didn't want to try to replace the cable, worked like a champ. :)

Just went back through the wiring diagram to see if there is something we missed and had an idea...
"What if" your brake booster had a slight leak and when it sets the booster looses all it's vacuum. When you try to start the engine the booster is now a giant hole until vacuum is built-up. After the car sets do you have stored vacuum (press the brake pedal and see if it's rock hard)? If not might pinch off the vacuum hose to the booster with needle-nose vice grips and see if it makes of difference.
Just a thought...

Based on your testing, I'd try to find another DME to plug in.

[shogun]

I would also vote for the fusable link. Easy to check, small money if to be replaced.
http://i31.photobucket.com/albums/c375/demet_/e34/fuseable_link_003b.jpg
comment by Demet:
On mine I wasn't getting spark, so I backtracked to the main relay and it wasn't being triggered on with the ignition ON, then I backtracked to the battery and found the fusable link 2 blown.

Mine's not typical anymore since I cut mine out and replaced it with a fuse, but here's a pic of where it was. That wire powers the main relay & ecu.

demet
89'535 turbo, 91'535, 85'735, 86'528e

[bolloc]

Test based on Bin's hypothesis.... Don't jump the relay. Try leaving fuel pressure gauge on the car overnight, then check before turning on ignition, and then monitor during ignition turn on with your 5 second delay prior to cranking.

[MBman]

OK great ideas, the fusable link is fine.
Just went out and pressed on brake and had resistance, but not rock hard. I will try that in morning, that is an interesting hypothosis, the vacuum line looks fine , clamps all look good, but you never know.
I did not know the booster held vacuum, only that it recieved vacuum from the engine running to boost pedal force??

Interesting on the the fuel pressure, bently says it needs to retain pressures for 20 -30 minutes which it easily did. However if it the pressure does go down overnight which I would suspect, perhaps the pump cant get everthing pressurized with a cold start ??
I need to try that and will when I can carve out some time and am not in a hurry to get to work , just in case:)

[E34x2]

The booster must hold vacuum so if the engine dies while driving you can still stop the car with "power brakes". If the booster has a leak or the check valve is bad the booster will empty while the engine is off resulting in the effect of a large vacuum leak when the engine starts.

Car/trucks that don't have boosters have a "hydro-boost" system that stores pressure in an accumulator for the same purpose. Your car has the vacuum booster.

[MBman]

makes sense , I will check it out tomorrow. I looked in the haynes manual and it gives a pretty detailed test of the system , so will go through that and post back
Good news
I think I found an ECU to buy ( very cheap:)) - suppose to be in good working order - last version for the M20 - Bosch 0261200524, Mine has Bosch 0261200173 but they interchange.

what about the DME relay, the white one, do they ever go bad or intermitant ?
Can they be tested?

[E34x2]

All relays have contacts and can/will burn over time. Back in the mid 80's I spent better part of two day's chasing a bad relay in an RX7 (all under warranty). Passed all the factory tests which were run with key on only. When tested with engine running relay could not carry the current and voltage to the ECU dropped. So sort answer is "yes" they can go bad.
Best way to test for sure is to measure the voltage at the ECU by back probing the connector. Only way to do it without the Bosch breakout box.

[MBman]

thanks for that, just may order and replace, I looked at them and original , just something else to eliminate, they are cheap insurance.
BTW - brake servo is fine, no issues there, and once today she had a hard start and tried the 5 second thing in run for fuel pump spin up and no help, so I think both of those issues are not the issue :(
I have a supposedly good ecu on the way. SO will try that and will give me one to take apart to look at .
Are these just plug and play, anything special other than making sure its the right part :).
And I will replace relays.
As always - more to follow

[bolloc]

I would still test with the fuel pressure gauge though... when the car finally starts its flooded and I am wondering why.

[MBman]

My guess would be non fire plugs,or weak spark, normal injection = flood
but then again its a guess.
I will test the start pressure, in fact today I was trying to figure out a way to actually install a fuel pressure guage so that I could check it on the run and at start from the interior, yet to come up with a solution that was safe :)

[MBman]

OK so I got a tested motronic unit from the last of the M20 line,I got if for $20 . I figure it has the latest ( albeit 1990) fixes and better electronics ( my hope) ect. Put it in and low and behold , it runs , starts great, no issues , knock on wood. I will say the little idle bumps I had are gone as well, she idles a little lower in RPMS and best of all -starts.:D

I am keeping my fingers crossed, in a few days if still no issues I will take apart my original motronic unit and take a look see. I have the motronic relay on order and will put that in as well.

I guess time will tell, I am hopeful this is the final chapter.:redspot:redspot:alright

[Mordan]

wow I hope for you that's the end of it. It is good lesson to everyone. Don't trust the ECU

[MBman]

true, I am in hopes as well! I am also hoping that my disection of the existing ECU will yield some clues. I am going to give this some time to make sure before I rip it apart

[THEhomelessONE]

OMG.. OFF TO PICK N' PULL!!

Glad you got it all sorted out!

I was thinking it was the ecu as well because when I'd unplug the AFM she'd start right up, run very rich after that, and die while driving sometimes.

[E34x2]

Glad you got it fixed. And look how much you learned about how it all works.

[MBman]

yup , me too, actually starting to offer advice - haha - day 5 , perfect starts, perfect running

[MBman]

Ok weeks of perfect starts, all types of weather, THEN @#$%#%#^& yesterday two no starts. It was really cold, dont know if it was a factor, fuel level was below 1/2 tank now I am wondering if fuel delivery - pump - may be a culprit
It is the ONLY thing I have not renewed.
I will tell you that everything I have done has improved the performance, and I am pleased to have done it and know that it is eliminated. I plan to keep this car for some time so the investment in parts has been OK with me. I am not out much
Since the ECU replacement , I have also replaced the relays .
I did take my original ECU apart , but did not see anything there , but then there are so many resistors and solder joints, it really is beyond me.
I think it had issues, not sure what :)
Question of the day - can fuel level in the tank affect the delivery performance of the pump??

[Mordan]

well since you plan on keeping the car, having a spare fuel pump is not a bad move lol.

man, you replaced everything but the pump! that's crazy! :)

Too bad you didn't try to put back the old DME to check for sure if it was indeed the issue.