poor acceleration after manual trans swap in 325e/m20b27

[Leftlane]

Solved!

I wanted to come back to this thread to close it out. I believe the problem is solved.

It indeed was the flywheel being mis-clocked. I disassembled everything after marking the block where the CPS mounting boss on the bellhousing was. Then I put the flywheel back on with the engine at #1 TDC with the metal tab on the perimeter pointing at my mark. Actually, I oriented the flywheel in the closest of the two positions around my mark - it wasn't exactly lined up, just as close as I could align with my particular method (recall you only get 45° adjustment out of the 8 bolt holes).

I then made this video:


Thanks to everybody for your input. I've got to finish getting everything all buttoned up and then it'll be time for a test drive when I'll know for sure I nailed it.

[vicecaptainred]

There should be a line on the crank pulley that matches up to a little line on the block itself to line up cyl#1 to tdc but hell straws and sharpies will get it done too!! Yes my flywheel pin was at like 4oclock and it needed to be at 7oclock (if youre looking at the back of your engine with flywheel on) an i was about 1/4 behind tdc so it was a 90° or two bolt holes late sending the signal. Is it an aftermarket flywheel?

[Leftlane]

Manual cars will start in gear. No neutral safety, or clutch safety.

Thanks! This is great news from my perspective. I was sweating the fact that my car was starting in gear and this was different from as-designed. I guess it's totally factory now ;)

One of the Green/Black wires goes to the DME harness to a Blue/Red wire, and provides the DME with 12v when in Park or Neutral. This 12v signal tells the DME to correct the idle speed for the engine load imposed by the torque converter when the transmission is in gear and the car is stopped. If you disconnect this on an automatic car when in gear, the idle bogs down. Since we don't have a torque converter putting a constant load on the engine, we don't need this function anymore.

Aha! This is good stuff. Great post, thanks for digging it up. Makes sense that the torque converter would tax the engine differently and that would have to be taken into account at low RPMs.

What's slightly confusing to me is that it seemed - looking at the schematic I posted for the manual car - that the manuals in 1985 (or Motronic 1.0 if you like) were getting +voltage on this line at all times (well, while the key was in start or run), so the computer (ICU or DME) would be doing the same thing it was doing for an auto in park and neutral. That confusion stems from the assumption that the ICU is the same for an auto or manual. Thinking about it, they're probably different and the manual ICU just ignores whatever is coming in on this line.** Just spitballin' here, no semblance to fact expressed or implied.

I'm willing to bet your diagrams are from a motronic 1.1 or 1.3 just because I see the "2.5i" in the picture. They were clearly doing something _slightly_ different, but the effect may yet still be the same.

**EDIT: wait, if this guess is true, then keeping my automatic ICU will be problematic because the computer would in theory be trying to apply this load-of-the-torque-converter function against a car which now no longer has such an animal. Idle would then be probably not where it needs to be. Which I think is why you're suggesting to disconnect this at what is the "injection electronis" in your diagram. It does seem that I may have two things incorrect - the flywheel and the fact that I'm sending voltage to the ICU based on how I did the wire splice at the shifter to avoid the neutral safety function of the automatic. Okay this makes my head hurt. I'll stop thinking too much about this and just focus on one thing at a time - flywheel.

Thanks again, all. I'll keep the thread posted as details emerge on the flywheel re-clocking.

[Todd Black 88]

I only think I've found the problem. I won't know for sure until I pull the trans, reclock, and all that jazz. I'd be very interested in your experiences with your manual swap as it relates to the DME and Idle Control Unit (for a motronic v1.0 car, naturally).

I am bugged by the fact that maybe the DME but certainly the Idle Speed Control unit (which is separate in a motronic 1.0 unit such as this) needs to know when we're in park or neutral in an automatic application.

Check out these wiring schematics...
Automatic:


Same but for manual (notice how it sends nothing to the ICU regarding neutral):


...page 122, the idle speed control unit showing either autos or manuals:


So it seems to me that the ICU gets a hot/not-hot signal from an automatic equipped car when the key is in run/start depending on if we're in park and neutral or not. It further seems to me that a manual trans equipped car gets an always-hot signal as long as the key position is in run or start on this same ICU pin (pin 7 just south of C131). This suggests a push by the engineers for manual cars that says "meh, we don't really care if a manual is in gear or not, just give this line some juice because the ____________ function isn't important for manuals." Now fill in the blank :)

Can someone with a factory manual transmission E30/325e from 1984 or 1985 tell me if they've got a neutral safety function? I cannot find that such a thing existed in these earlier cars. The pedal box I got did not have a provision for a switch on the clutch pedal, only the switch for the brake pedal for the brake lights. My car starts in gear the way I wired it and I'm starting to believe the factory manual cars did too (which seems weird, but the mid-80s were a different era, I suppose).

Meanwhile, I'm currently corralling my friends for assistance with the upcoming trans pull.

Manual cars will start in gear. No neutral safety, or clutch safety.

[Andy.B]

IIRC, e30's DO do that. There is a wire connected to the shift selector that informs the DME when it is in neutral/park vs D/R. It was a while ago that I did a manual swap, but I should have some info here somewhere...

Edit: It looks like OP found their issue anyway. That's what I get for not reading to the bottom of a thread first.

I worked on figuring out the wiring differences between the auto and manual transmission, while also sorting out how to make sure cruise control still works. It was actually easier than I thought, but pretty much every auto -> 5speed swap guide I read was incorrect (at least for my car), or made it harder than it needs to be .

First off, although I never have any intention of converting this car back to an automatic transmission, I avoided cutting or hacking anything in a way that would prevent restoring it to its original condition. For the wiring, that meant cutting the harness off of the shift selector side of the connector instead of the car harness side.

Then I modified the shift selector side of the harness to make it a plug in adapter. To do this, cut the harness about 4 inches or so past the range selector connector to create a pigtail harness.

The harness now has 5 wires:
Brown (ground for indicator light)
Green/Black (Park/Neutral signal)
2x Green/Yellow (Hot on run or start)
Blue/White (Reverse light)

To modify:
Connect 1x Green/Yellow to Green Black
Connect 1x Green/Yellow to either one of the two reverse light harness wires
Connect Blue/White to the other reverse light harness wire
Cap off the Brown wire.


I used heat shrink tubing to make it nice and tidy.
This modified harness provides the signal for the manual transmission reverse light switch, and the Park/Neutral signal to the ignition relay to allow the car to start.

For bonus points, cut a car harness side C301 from a parts car to wire into the modified selector range pigtail to connect to the manual transmission reverse light harness, making it a truly plug and play job.

That's it at the shifter, but there are two wires that need to be tracked down and disconnected. The Black/Green wire that provides 12v when in Park/Neutral goes to the DME and Cruise Control module as well.

One of the Green/Black wires goes to the DME harness to a Blue/Red wire, and provides the DME with 12v when in Park or Neutral. This 12v signal tells the DME to correct the idle speed for the engine load imposed by the torque converter when the transmission is in gear and the car is stopped. If you disconnect this on an automatic car when in gear, the idle bogs down. Since we don't have a torque converter putting a constant load on the engine, we don't need this function anymore.

Simply unplug the Green/Black wire from the Blue/Red wire coming out of the DME harness, and tape them up out of the way.


The next wire to disconnect is the 'cruise enable' signal wire (connector C143) from the cruise control harness. The ETK wiring manual states 'For manual transmission C143 is taped to cruise control harness'. C143 is located under the ABS computer on the driver side. The Cruise Control harness side is Blue/Brown and on the car harness side, you will find another Green/Black wire that come from the auto transmission gear selector harness). All you have to do here is disconnect wires and tape them up.
My phone died before I could get pictures of this connector.

What you are left with is a single Green/Black wire in the car harness telling the starter relay that the automatic transmission range switch is in Park/Neutral so it can engage the starter, even though the the range selector is now gone. The DME does not need this signal, since it isn't an automatic car anymore, and the Cruise Control won't work when getting this signal.

The effect of the modified wiring pigtail is picture below... it's really much simpler than it's made out to be:


* I haven't verified the cruise control yet due to a faulty cruise control system, but considering the ETK tells you it should be taped out of the way on a manual car, that one is pretty obvious.

*According the the diagram, the Brown/Red wire coming out of the DME harness next to the Park/Neutral signal is unused. This is not true (at least on my car). It provides the speed signal to the instrument cluster through a Blue/Yellow wire that runs through the dashboard, and plugs into the green connector right next to the Cruise Control's speed signal wire at the cluster.

There you have it. Without cutting anything on the cars harness, you now have the car working as if it was equipped with a manual transmission from the factory. If anybody, for whatever crazy reason, decides they want to swap back at some point, then all they have to do is reconnect those other two Green/Black wires.

[Leftlane]

IIRC, e30's DO do that. There is a wire connected to the shift selector that informs the DME when it is in neutral/park vs D/R. It was a while ago that I did a manual swap, but I should have some info here somewhere...

Edit: It looks like OP found their issue anyway. That's what I get for not reading to the bottom of a thread first.

I only think I've found the problem. I won't know for sure until I pull the trans, reclock, and all that jazz. I'd be very interested in your experiences with your manual swap as it relates to the DME and Idle Control Unit (for a motronic v1.0 car, naturally).

I am bugged by the fact that maybe the DME but certainly the Idle Speed Control unit (which is separate in a motronic 1.0 unit such as this) needs to know when we're in park or neutral in an automatic application.

Check out these wiring schematics...
Automatic:


Same but for manual (notice how it sends nothing to the ICU regarding neutral):


...page 122, the idle speed control unit showing either autos or manuals:


So it seems to me that the ICU gets a hot/not-hot signal from an automatic equipped car when the key is in run/start depending on if we're in park and neutral or not. It further seems to me that a manual trans equipped car gets an always-hot signal as long as the key position is in run or start on this same ICU pin (pin 7 just south of C131). This suggests a push by the engineers for manual cars that says "meh, we don't really care if a manual is in gear or not, just give this line some juice because the ____________ function isn't important for manuals." Now fill in the blank :)

Can someone with a factory manual transmission E30/325e from 1984 or 1985 tell me if they've got a neutral safety function? I cannot find that such a thing existed in these earlier cars. The pedal box I got did not have a provision for a switch on the clutch pedal, only the switch for the brake pedal for the brake lights. My car starts in gear the way I wired it and I'm starting to believe the factory manual cars did too (which seems weird, but the mid-80s were a different era, I suppose).

Meanwhile, I'm currently corralling my friends for assistance with the upcoming trans pull.

[enigmaticdream]

Nice working figuring out what was going on.

[Andy.B]

Boy, if it was a more modern car, I'd say that it thinks it's in neutral, and is mapping the ECU accordingly- but E30 don't do dat shit...

IIRC, e30's DO do that. There is a wire connected to the shift selector that informs the DME when it is in neutral/park vs D/R. It was a while ago that I did a manual swap, but I should have some info here somewhere...

Edit: It looks like OP found their issue anyway. That's what I get for not reading to the bottom of a thread first.

[LateFan]

Very thorough. Good sleuthing.

How did it even run?

[Leftlane]

TDC and other news

Before I tell you what I found, allow me a few words on finding TDC. So it seems TDC is something of a challenge on these cars... compared to my 350cid Chevy-powered CJ5, anyway. Being an army of one, I found it impossible to turn the motor and hold my finger over the spark plug hole to see if I was on the compression stroke.

I read that you can pull the valve cover and ensure that both rockers on cyl #1 were not being actuated while having the piston at its highest travel. I found this to be easier said than done. I also used the straw trick whereby you use a long straw down the #1 cylinder to determine the piston depth. Various turns of the crank while marking the straw depth against a known point yielded a rough high water mark.

I found in the Bentley manual that there's a TDC locator sensor on the front crank pulley, too. It's not used by the DME. It's used by a special BMW Service-test unit. I don't have one of those, but I do have a silver Sharpie. There's a single hole in the perimeter circumference of the lower pulley where this sensor is mounted up against. I felt like this had to be the trigger for the sensor, so I marked it with a line.

Sure enough, when I used the calibrated eyeball on the rockers and the calibrated straw on the piston depth to get TDC on #1 compression stroke, I looked down at the pulley and saw this:


Woo hoo... theory proved by reality. What could be better?

Now it was time to check the location of the flywheel CPS trigger, if I even had one (i.e., if I had the correct flywheel which I wasn't 100% sure of).

It's funny. The Bentley manual doesn't exact say which of the two sensors mounted near each other is the reference sensor. There's a couple diagrams without a legend telling which end is up (or front, or anywhere). Witness:


With the #1 at TDC I took a shot in the dark and pulled the upper of the two sensors and saw this:

Hmmnnn... flywheel teeth. Must be the engine speed sensor. That's when it clicked that the lower one is actually set back further towards the rear of the car. These two sensors should be referred to as the front and rear sensors because it's this distinction that dictates what they're actually measuring. When the flywheel is mounted to the engine, the magnetic pickup for the CPS is actually more rearward than the teeth of the flywheel. So I pulled the lower...er... rear sensor and saw this:

Hmmnn... hard to see what's going on, but feeling around with a screwdriver didn't reveal any bumps of any kind. So I used vicecaptainred's suggestion and balanced the screwdriver in the port while slowly manually rotating the engine by the front pulley (yes, it's a 22mm nut).

Sure enough while turning the crank something bumped the screwdriver and it fell out of the hole. It took some back and forth with the pulley and the newly created site-hole (it was just me doing the work), but eventually I was able to find this:

That is indeed the rectangular reference pin. Well, at least I've got the correct flywheel for an early motronic. That's the good news.

Here's the bad news. The crank was now in this position:

That's about 180° from where it needs to be.

So thanks to all and especially vicecaptainred for their input. I now know what I need to do. I'm going to go into radio silence for a bit while I slowly pull the trans out and figure out a way to reclock the flywheel so that the pickup is in the proper spot. I'm thinking I'll mark the engine block with the location of the CPS sensor so I'll have something to reference once the bellhousing is out and the crank dowel sleeve is pulled. Since there are 8 bolts, each bolt represents 45° of rotation, so hopefully this method will be exact enough. Is that how you did it, vicecaptainred?

[vicecaptainred]

Yes pulling the sensor, if you look thru the hole on the bell housing you should see the pin in the hole with the engine lined up at TDC. if you only have yourself, slide a screwdriver in the sensor hole an slowly rotate the crank(22mm i believe) until you feel it hit the pin. Check location now vs tdc positioning. I did not put a new dowel sleeve back in as it wasnt really needed. Yeah i know, literally no one mentions this but i guess looking back it now....its one of those things you just end up like 'ohhhhhhhhh now i understand' was for me anyways. Pull that sensor and check that pin!

[Leftlane]

I had the same symptoms (starts but runs poor with no acceleration tops at 45mph) in my '84 (5-speed) after a motor swap from an '86(auto) for the life of me couldnt figure it out. Turned out when i put my single mass flywheel on the '86 auto, the dowel sleeve would only let it go on one way but the reference pin on the outside of the flywheel was 90 degrees behind when the motor was at TDC so i repulled the motor an trans, took out the dowel sleeve, reclocked the flywheel(pin should be in the second CPS hole when at TDC

Thanks and oh dear gawd! This is my worst nightmare and up until I read your reply was only theory because I hadn't read about this even being possible.

Okay, vicecaptainred, bare with me. I've got questions.

Is it true that I should be able to see the sensor/reference pin (a rectangular chunk of metal on the outer flywheel if I understand correctly) when I pull the crank position sensor at #1 TDC and thereby diagnose a good or not-good flywheel setup without pulling the trans first?

Just so it permeates my thick skull: On your second attempt, you setup the engine at #1 TDC and pulled the crank sleeve (on the crank itself) which aligns the flywheel in only one mounting position, repositioned (pounding into place?) it into a new hole on the crank output end which would allow the reference pin to point at the CPS (or where it would be when the bellhousing is in place), and bolted everything back together?

Lastly, you did nothing to alter your "early motronic single mass" flywheel as part of your fix? Just the flywheel-locating sleeve on the crank getting put into the proper position?


I never knew that sleeve could be pulled out but it makes sense that it could. Of course it was in the correct hole for the automatic trans flexplate. I just assumed that would be the same spot for manual trans flywheel, but it seems not. None of the auto to manual trans swaps I've read so far have mentioned this little nuance - but then they'd have to have been performed against an early motronic (v1.0) car to run into this.

[vicecaptainred]

I had the same symptoms (starts but runs poor with no acceleration tops at 45mph) in my '84 (5-speed) after a motor swap from an '86(auto) for the life of me couldnt figure it out. Turned out when i put my single mass flywheel on the '86 auto, the dowel sleeve would only let it go on one way but the reference pin on the outside of the flywheel was 90 degrees behind when the motor was at TDC so i repulled the motor an trans, took out the dowel sleeve, reclocked the flywheel(pin should be in the second CPS hole when at TDC

[Leftlane]

Boy, if it was a more modern car, I'd say that it thinks it's in neutral, and is mapping the ECU accordingly- but E30 don't do dat shit...

Thoughts-
Mechanically check ignition timing, like with a light?
If it was timing you to TDC, you'd run like this.

Check running fuel pressure?

Release parking brake?

t

Appreciate the suggestions. A local reputable independent german auto shop checked the timing as they thought the belt might have busted a single tooth. Told me it was fine. I need to do it myself just to double check as I'm retracing all steps not performed by me.

I sort of left that part of the story out. Local shop had the car 5-6 weeks and couldn't find the problem. They make their money on more recent German hardware these days, not the old E30's, so they were working on it as they could rotate it into the shop between real paying jobs and - much appreciated by me - not gouging me on hourly rates.

They did find a cracked intake boot and replaced it, but sadly that didn't prove to be the problem. Ultimately, they tried as many of the "easy things" as they could think of with this cheaper pricing model before wanting to tear into the trans (bad/wrong flywheel?) or other spendy operations.

So now it's home and up to me to figure it out.

Regarding fuel pressure, I've check the fuel _volume_ but not the fuel pressure. The Bentley manual has you disconnect just after the fuel pressure regulator and run the fuel pump for 30 seconds. It states you should fill 30oz of fuel in a container in those 30 seconds. I filled exactly 30oz of fuel for exactly 30secods of running the pump. That was impressive, actually. So I did not move on to checking pressure, but I suppose I should?

I wish it would be the parking brake, but it behaves this way in neutral (clutch in or gear in neutral position) too. Can't be the parking brake.


I've got a couple days off work coming up and I plan to do these:
- new plugs (got 'em waiting in a box)
- new plug wires
- compression/leak-down cylinder tests
- valve lash adjustment
- while #1 on TDC, pull the bellhousing reference sensor and make sure I can see the magnetic pickup on the flywheel (assuming this is where it would be when #1 is at TDC - feel free to chime in on this)
- more AFM futzery?


Again, thanks for all input. I'll keep this thread updated. Nothing worse than a thread that goes into detail for days/weeks/months and then leaves you hanging at the end.

If all else fails, I'm planning on taking it to a BMW dealer for a thorough cleaning out of my bank account. I really want this car to run properly. I called them and they said they've still got the equipment to work on these cars. My local independent guy did not. He'd "gotten rid of" all his E30 stuff long ago. My sense was that amounted to a spare Motronic computer they had lying about for diagnostic R&R purposes.

[Leftlane]

^^^Right… cable length? Pedal mounting? Fittings slipped? Throttle arm or TPS got knocked? AFM is ok you said?

Thanks. For the record, I did not change out the accelerator pedal or cable. However, as noted above, I did remove the auto trans kickdown cable.

I _think_ the AFM is okay based on the fact I saw the ohms dial up as I slowly moved the door by actuating it as much as I could with my finger.

Since then I've seen various videos, like this one and this one that make me think I didn't test the AFM unit as thoroughly as I should have. :/