SOLVED - Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Specific conversations and info for the BMW E28 M5 and M535i.
Post Reply
jhh925
Posts: 913
Joined: Jul 11, 2015 12:23 AM
Location: Reno

SOLVED - Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by jhh925 »

I'm banging my head against the wall on this one. I'm restoring an E28 M5. This was a full "from the ground up tear it down to the studs" kind of rebuild, including a full engine rebuild.

The engine had a Dinan stroker kit installed in the mid/late 90s. If there were different flavors of that kit at the time (e.g., stage 1, stage 2, etc.), then unfortunately I don't have specific info on which version was installed. I do know (in case it's helpful) that the crank is aftermarket and it has hotter cams. The ITBs appear to be the original standard units.

Everything is basically new: all new hoses, plugs (correct NGK for the S38-B35), ignition wires (Bremi), cap & rotor (Bosch), TPS (Facet), coolant temp sensor, etc.

There are no vacuum leaks - confirmed with smoke tests. ITBs are synched. Butterfly throttles are adjusted per factory instructions.

Here's what happens: At part throttle, the engine starts misfiring badly at almost exactly 3000 rpm. On the road, once it starts misfiring, you can floor the throttle, and the issue goes away and the engine winds out as you'd expect.

Things I've tried:
- I swapped the ignition wires (no change)
- I swapped the coil (no change)
- I swapped the ECU box, though I had to keep the same Dinan chip (no change)
- I hooked up a pressure gauge to the fuel feed line, and it's rock solid at 40+ psi
- I've double-checked all the grounding points that the engine wiring harness relies on (at the firewall, at the right-front-bottom of the block, on the head under ITB #6, at the left-front fender wall). All were tight.

BUT ... if I short out pins 2 & 3 on the TPS (essentially fooling the ECU into thinking it's at WOT), the issue goes away.

So, does anyone have insights into the engine management mods that Dinan did with this set up? Was it just the fuel & ignition maps on the chip? Or was there something more? Is it possible at all that they did something to just skip over the part-throttle maps? (Seems unlikely to me.)

Does anyone have any insights on what might be going on? Or suggestions on what I should check next?
Last edited by jhh925 on Jun 23, 2023 2:14 PM, edited 1 time in total.
austin8753
Posts: 1442
Joined: May 16, 2010 1:37 AM
Location: Portland, Oregon, USA

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by austin8753 »

not exactly a common item to have around for an S38, but i'd have a look at the air flow meter.
jhh925
Posts: 913
Joined: Jul 11, 2015 12:23 AM
Location: Reno

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by jhh925 »

austin8753 wrote: May 12, 2023 7:58 PM not exactly a common item to have around for an S38, but i'd have a look at the air flow meter.
Ah, good call. I do have a spare for an S38, so I'll try that. Thx.
jhh925
Posts: 913
Joined: Jul 11, 2015 12:23 AM
Location: Reno

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by jhh925 »

More info:
- I swapped AFMs. Both are the correct AFM for the S38 (Bosch part number ends 025). No change, the misfire persists.
- The misfire does NOT happen when the engine is cold. That is, on initial start-up after sitting all night, the engine runs fine and there's no misfire at 3000 rpm, the engine revs up just fine.
- I hooked up an old inductive timing light to see what's going on with the ignition when the misfires start. Under 3000 rpm, the strobe is consistent and fires exactly as you'd expect. As the engine starts to misfire, the strobe goes wonky ... it's not consistent anymore, and it mostly isn't firing anymore, just occasional flashes. And I'm getting the same wonky strobe light pattern off of all six ignition wires.
jhh925
Posts: 913
Joined: Jul 11, 2015 12:23 AM
Location: Reno

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by jhh925 »

And some more info:
- Pulled the CTS and bench-tested it in hot & cold water. Resistance values for the CTS were consistent with the numbers in Bentley & with the numbers I got a couple of years ago when I bench-tested three CTSs.
- There is coolant flowing behind the CTS. I think this is confirmed because (i) the temp gauge on the dash gives the expected response as the engine warms up, and (ii) I confirmed the coolant tube was wet with coolant when I pulled the CTS (I had previously drained about a gallon of coolant before I pulled the CTS).
- I pulled the ECU and unplugged the CTS. The CTS ground wire has continuity to ground, and the other CTS lead that goes to pin 13 of the ECU has continuity and didn't short out to any of the other terminals in the ECU.
- I put the CTS back in, filled the coolant back in, etc., started the motor, and ran it a bit to get the motor warm. Confirmed that the misfire was still there at 3000 rpm. Disconnected the CTS. Engine continued to run, but at a much higher idle (about 2000 rpm). I then increased the throttle, and the misfire came back as usual, though at a bit under 3000 rpm.
jhh925
Posts: 913
Joined: Jul 11, 2015 12:23 AM
Location: Reno

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by jhh925 »

More info:
- Unplugged the O2 sensor, which as I understand it will force the ECU into open-loop mode (as is the case when the WOT switch closes on the TPS). Same result. Engine misfires at 3000 rpm.
Mike W.
Posts: 26846
Joined: Feb 12, 2006 12:00 PM
Location: California Whine Country

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by Mike W. »

jhh925 wrote: May 17, 2023 1:09 PM More info:
- I swapped AFMs. Both are the correct AFM for the S38 (Bosch part number ends 025). No change, the misfire persists.
- The misfire does NOT happen when the engine is cold. That is, on initial start-up after sitting all night, the engine runs fine and there's no misfire at 3000 rpm, the engine revs up just fine.
- I hooked up an old inductive timing light to see what's going on with the ignition when the misfires start. Under 3000 rpm, the strobe is consistent and fires exactly as you'd expect. As the engine starts to misfire, the strobe goes wonky ... it's not consistent anymore, and it mostly isn't firing anymore, just occasional flashes. And I'm getting the same wonky strobe light pattern off of all six ignition wires.
Since you say the strobe goes wonky, I would look to what is before it in the ignition system. Looking towards the flywheel sensors or the dist cap and rotor. The pattern seems odd, but if spark doesn't seem like it's consistent, look in that direction. Maybe even just spark plug cables, but doubtful. I'd look towards the flywheel speed sensor first.
Preyupy
Posts: 326
Joined: Jun 21, 2012 12:48 AM
Location: Issaquah WA

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by Preyupy »

Do you have an aftermarket chip in the ECU (Dinan, WAR etc)? It sounds like the ECU thinks you are at RPM limit and is activating a rev limiter. It could be possible that the limit has been set at 3000.
jhh925
Posts: 913
Joined: Jul 11, 2015 12:23 AM
Location: Reno

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by jhh925 »

Mike W. wrote: May 19, 2023 2:07 AM Since you say the strobe goes wonky, I would look to what is before it in the ignition system. Looking towards the flywheel sensors or the dist cap and rotor. The pattern seems odd, but if spark doesn't seem like it's consistent, look in that direction. Maybe even just spark plug cables, but doubtful. I'd look towards the flywheel speed sensor first.
Preyupy wrote: May 19, 2023 1:14 PM Do you have an aftermarket chip in the ECU (Dinan, WAR etc)? It sounds like the ECU thinks you are at RPM limit and is activating a rev limiter. It could be possible that the limit has been set at 3000.
Interesting ... especially the thought about hitting the rev limiter. That seems like it could be what the ECU is trying to do. And interestingly, I called Korman to see if they had any thoughts. Korman did a stroker kit with a custom chip back in the day as well, and they seem to still support their old stuff. The guy at Korman also suggested looking at the flywheel sensors (engine speed & crank position). Both are new, and I believe both are aftermarket. I'm going to swap the old one back in to see what happens. Conceivably, the engine speed sensor is giving bad info and for whatever reason, the ECU hits the rev limited in the part throttle condition? Hmmm ...

(I'm traveling all next week, so I'll be hitting this project again at in a week or so. Thanks for the suggestions.)
Preyupy
Posts: 326
Joined: Jun 21, 2012 12:48 AM
Location: Issaquah WA

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by Preyupy »

If the ECU is getting a signal from an unshielded sensor it can easily think that the engine is turning a lot faster than it is. Remember the rev limiter is just looking for a frequency of an ignition pulse that corresponds to a 6 cylinder engine turning 6400 rpm. if it is getting a false signal the ECU can't tell the real impulse from the fake one. In the old days of ignition points if you had POINTS BOUNCE the rev limiters would kick in sometimes at less than 50% of the normal limit RPM.
Fanclutchnut
Posts: 113
Joined: Jan 05, 2021 7:32 PM
Location: Westchester Co NY

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by Fanclutchnut »

Trying to think of simple things. Could the coil be crapping out at higher demand RPM?
jhh925
Posts: 913
Joined: Jul 11, 2015 12:23 AM
Location: Reno

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by jhh925 »

Preyupy wrote: May 19, 2023 8:34 PM If the ECU is getting a signal from an unshielded sensor it can easily think that the engine is turning a lot faster than it is. Remember the rev limiter is just looking for a frequency of an ignition pulse that corresponds to a 6 cylinder engine turning 6400 rpm. if it is getting a false signal the ECU can't tell the real impulse from the fake one. In the old days of ignition points if you had POINTS BOUNCE the rev limiters would kick in sometimes at less than 50% of the normal limit RPM.
Sounds reasonable and I had a similar thought, but why would the rev limited kick in when the TPS is in part throttle, but NOT when the TPS is showing wide-open throttle?
Fanclutchnut wrote: May 20, 2023 8:06 AM Trying to think of simple things. Could the coil be crapping out at higher demand RPM?
I swapped coils early one to another one in the parts bin. Behaved the same way. And it also leaves the same question, why would the coil crap out with the TPS in part throttle, but not in WOT?
Preyupy
Posts: 326
Joined: Jun 21, 2012 12:48 AM
Location: Issaquah WA

Re: Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by Preyupy »

It could also be a problem with the air flow meter. There might be a dead spot in the variable resistance carbon contact. At WOT the ECU looks to a separate fuel and ignition cells in the Map. It does not rely on the AFM for air flow.
jhh925
Posts: 913
Joined: Jul 11, 2015 12:23 AM
Location: Reno

Re: SOLVED - Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by jhh925 »

So, we finally figured out what was wrong. Biig tip of the hat to derrith and also Chris Lingle at GAS for the hours of phone help.

Anyhow, if you've read this far, there were a bunch of clues that something was wrong with the ECU or the chip in the ECU. I sent the ECU off to ecudoctors.com in FLA. They found issues with the unit & repaired it, but upon reinstall, the misfire issue persisted. Derrith sent me two known-good chips suitable for use in an S38B35 running the '079 ECU, but the problem persisted. OK, so maybe it's not the ECU.

The other big clue that I'd latched onto is that the misfire wasn't there when the engine was cold, so I hooked up a potentiometer in place of the coolant temp sensor. After the engine warmed up, I turned up the resistance on the pot to try to fool the ECU into thinking the engine was cold. The idle RPM went up (as you'd expect), but the misfire problem was still there. So whatever the temperature dependence was, it wasn't the CTS signal that was triggering the problem.

On a lark (not sure what made me think to do this), I put the inductive timing light back on to the #1 cylinder ignition wire and tested to see where exactly the spark was firing. It was showing that the plug was firing about 30°degrees before TDC, but online sources & D'Sylva indicated that there shouldn't be that much advance (if any) at idle. Hmmm, so something was causing a giant problem with the ignition timing even when the engine seemed to be running fine.

Derrith & I then had a similar thought at about the same time. Maybe there's something wrong with the flywheel? The engine speed sensor and the crank pos sensor both read off of the flywheel. Derrith then mentioned that there's an inspection port on the transmission bell housing that allows you to check the flywheel position vs the TDC indicator on the harmonic balancer (when the TDC mark is lined up in front, there's a matching TDC mark on the flywheel that should show up centered in that inspection port).

Well, sure enough, the flywheel was off (retarded) by quite a bit. Fuck. Yes, I know what you're thinking, there's an alignment dowel that should prevent the flywheel from getting installed incorrectly. Well, loyal E28'er, read on!

I spent a day dropping the transmission, pulling the clutch, etc. Sure enough, the flywheel was two bolt holes off (8 bolts in total, so clocked in the retarded direction by 90°). There was no dowel pin at all. I ordered the missing part and then had a fair amount of trouble pressing it back into the crankshaft.

So here's my theory: This car has a stroker kit (so aftermarket crankshaft) and about 215,000 miles, This clutch has been replaced before, and likely at least once or twice from the original installation of the stroker kit. My guess is that whoever put the stroker kit in couldn't get the pin back in (or damaged it in trying to do so), so just skipped it. After all, the purpose of the dowel pin is just to make sure the flywheel is clocked right, and if you know what to look for you can just line up the two larger holes. I know I fucked up at that point - I should have noticed that the dowel wasn't there and should have addressed the issue at the time - but I didn't.

Anyhow, I just got everything back together, fired her up, let the motor warm up, and lo & behold the misfire is gone - and the motor runs a bit smoother as well (no surprise).

Thanks all, and don't be me when you do your next clutch job.
Preyupy
Posts: 326
Joined: Jun 21, 2012 12:48 AM
Location: Issaquah WA

Re: SOLVED - Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by Preyupy »

It is amazing how it usually comes down to the basics. Happy to hear you have found the problem and you can now go and have fun with it instead of working on it.
Etip
Posts: 153
Joined: Sep 05, 2021 2:06 AM
Location: Seattle

Re: SOLVED - Help! Misfire at 3000 rpm with Rebuilt S38B-35 with Dinan Stroker Kit

Post by Etip »

Wow!!! Glad you got it sorted out.
Post Reply