Drivability Issues

[Bizboy]

Hello, fellow smol (or big) brain E30 enthusiasts! I've been struggling with my E30 for quite some time and am nearly at my wits end.

Here are the specs of my E30 and highlights of my issue.

1989 325is (M20B25)
MSPNP2 ECU
AFM Delete (GM IAT in its place)
AEM UEGO X Series wideband in OEM O2 sensor port and wired directly to ECU 26 pin optional connector. Signal positive to pin 21 and signal ground to pin 22.
Siemens Deka 60lb/hr (630cc/min) injectors
Walbro 350lph fuel pump

As for my drivability issue, I've finally got the car running pretty decently. However, whenever the engine is at or above about 4800 RPM it starts breaking up pretty good. Almost sounds like an anti lag/2 step lol. If under load, the car will surge or buck because it momentarily loses power every second/half second or so.

Here's what I know about the issue.
-Fixed previous drivability and emissions issues by replacing, with genuine BMW parts, the timing belt, tensioner, distributor cap, distributor rotor, VR trigger sensor, and spark plugs. Spark plug wires were aftermarket and the Siemens deka 60's were also installed at the same time.
-REQ fuel has been updated with the new injectors.
-Datalog shows no loss in RPM or Sync loss
-PW stays even
-AFR has spikes within a couple hundredths of a second. For example, it will be say, 12.6 AFR at 4922 RPM and 0.06 seconds later will read 13.6 AFR then 0.06 seconds later back to 12.6 AFR.
-Tried Richening VE table in affected zones as well as retarding ignition timing. Base map was about 38 degrees advance. tried 32.
-Base timing verified with digital timing light. first tooth angle BTDC is set to 86 degrees.
-High speed tooth logger shows one bar approximately triple the height of the 57 short bars. (Did notice in the affected RPM zone, the short bar immediately after the long bar was lower than the other short bars. Also, the short bars appear to change in height in the shape of a sine wave once it hits the problem RPM range).
-Tried adjusting variable resistor for trigger Sensor in both directions.
-Noise filtering on or off made no change.
-Tried changing injector dead time. currently tuned at 0.9. Went up to 1.2 and down to 0.7. (Richened or leaned out fuel table to correct AFR for each change.)
-EGO control authority set to 0% (Off)
-Accel enrichment set to MAP dot of 1000kpa (Off)

Aside from what is listed above the only two things left I can think of to try, is to verify fuel pressure, and check the ignition coil and its wiring. Which I'm quite certain the fuel pressure will be just fine.

Thank you, generous wizards and or gurus of the mighty E30 and Megasquirt community!

MSQ: https://drive.google.com/file/d/1_QQ...usp=share_link
Datalog: https://drive.google.com/file/d/1JZd...ew?usp=sharing

[Hoffbrohaus]

Well I'm a newbie to megasquirt but thought I'd chime in with my experiences so far.

You mentioned you are running an AEM UEGO wideband o2. I don't know anything about this wideband, but I am running an Innovate wideband with Bosch sensor. The Bosch sensor goes to the innovate gauge and then there is an analog out from the gauge which I ran a signal wire from to pin 21 in the MSPNP. No ground from the gauge to the PNP like you said you did on pin 22. I'm using pin 22 for my wasted spark setup.

Other than that everything you wrote sounds good to a novice like me. I assume you calibrated all your sensors too. Maybe run the spark test, I just can't imagine that's it with new plugs, cap & rotor.

[Bizboy]

Woah, that's really interesting. On the MSPNP2 manual, pin 22 is listed as a signal ground. I haven't looked into wasted spark yet, (other than reading your post about the trial of the gm coil pack from DIY auto tune) so I have no idea how the wiring is supposed to be. But I would've guessed grounding straight to chassis or maybe pin 5 which is listed as chassis ground would be the way to go.

The AEM controller (gauge) sounds like it's the same as your Innovate. I believe AEM also uses a Bosch sensor wired to the controller. And yes, currently I have the analog signal positive wire from the controller output, going to pin 21 and the analog signal ground wire going to pin 22 since it was labeled as signal ground. When I first installed the wideband, I had the analog positive to the signal wire of the factory O2 wiring and the analog ground to the ground wire in the factory O2 wiring. It worked fine the first way, but I found the AFR seemed to read about 1.0 lower than what the controller was reading. The reason I changed it is because of that difference and the AEM controller manual stated to connect the positive and negative analog signal output wires directly to the pins on a standalone ECU labeled as analog signal positive and negative. So I assumed that since the MS has no analog signal negative that the signal ground (pin 22) would be the correct choice. And it did seem to correct that slight difference in the AFR reading.

Since making this post, I've been continuing to do Google searches and reading threads on this forum. I found someone with a somewhat similar issue on their RB20 running MS. Their issue ended up being a fouled spark plug from the initial tuning process. So, when I get a chance ill pull all the new plugs and check them. I would think however, that a fouled spark plug would not run correctly at lower RPM's and my E30 is running great below 4800 right now. The other potential cause I've found on this forum is it seems that MS is very sensitive to the VR trigger sensors distance from the reluctor wheel. Also, there's a whole thread talking about having a specific resistor in one of two configurations (based on MS type) for the trigger sensor. I can't imagine that I would need to add a resistor to the system since the instructions for installation of the MSPNP into the E30 chassis doesn't even mention it. Regardless, I'll check the gap between the VR sensor and the wheel and maybe check resistance to see if it already has what was specified in that thread. Other than that, I haven't had a chance yet to test my coil either.

Thanks for chiming in! I really appreciate any help I can get.

[varg]

37° advance at 100kpa, 4,800rpm? That's pretty high, I'm at 28° there. If it were a CPS signal issue you should see sync losses or at least strange results on the tooth logger, you'd also see weirdness in a data log which shows engine rpm (not smooth traces). AFR spikes are consistent with a misfire and most likely a symptom not the cause. Can it rev past the breaking up? Are the plugs gapped correctly? There's no TPS signal but I'm assuming this was just a free rev since the MAP and pulse width were so low. It was super lean earlier in the log. We have the same injectors but your fuel table is way different than mine, though I have incorporate AFR target turned on and that skews things, I have my table skewed a little rich and use closed loop fueling to pull it back to lean cruise AFR. I also use 1 squirt/simultaneous and you're using 2 squirts/alternating.

[Bizboy]

37° advance at 100kpa, 4,800rpm? That's pretty high, I'm at 28° there. If it were a CPS signal issue you should see sync losses or at least strange results on the tooth logger, you'd also see weirdness in a data log which shows engine rpm (not smooth traces). AFR spikes are consistent with a misfire and most likely a symptom not the cause. Can it rev past the breaking up? Are the plugs gapped correctly? There's no TPS signal but I'm assuming this was just a free rev since the MAP and pulse width were so low. It was super lean earlier in the log. We have the same injectors but your fuel table is way different than mine, though I have incorporate AFR target turned on and that skews things, I have my table skewed a little rich and use closed loop fueling to pull it back to lean cruise AFR. I also use 1 squirt/simultaneous and you're using 2 squirts/alternating.

Sorry about the late reply. Life has thrown several curveballs at me lately.

I hesitate to change my ignition table timing to 28 degrees because that's 9 degrees retarded from where I'm at. Before I had my current issue, I had incorrectly thought my base timing (tooth first angle DBTDC) was correct at the 97 degrees that was set from the base E30 startup tune. Now I know my engine timing marks are lined up correctly at 86 DBTDC. Meaning, (if I understand this correctly) that at 97 DBTDC my ignition table values were essentially 11 degrees retarded. Being 11 degrees retarded, I had an issue where the engine would not rev past 4200-4800 and would fall on its face there, as well as backfire.

Yes, now with my base timing set at 86 degrees it can rev past the breaking up starting point of 4800 rpm. It will continue building revs to my current rev limiter set at 6000 rpm, however will still break up through that 4800-6000 range.

When I installed the plugs I checked the gap on each one and they were all correct according to the Bently manual. (Engine is currently NA.)

Correct. I was just free reving. the breaking up issue happens both free reving and under load. I do not have a variable TPS installed currently. Just the factory 2 position TPS which megasquirt "shouldn't" be using I think.

Yeah, my warm startup settings are not good at all yet. If its dead cold its about 10-11 AFR at startup and increases to 12-14 as it heats up. If its warm at all for example coolant temp at say 130 degrees it runs very lean and surges/hunts between 13-18 AFR.It will stabilize at 12-14 again after 5 to 10 minutes as it heats up. I also have incorporate AFR target turned off and overrun fuel cut turned off since I'm still tuning it in.

Have you changed your fuel setup besides just injectors? 2 squirts/alternating was what came in the E30 base tune. And since I'm new and kind of don't know what I'm doing, I left that setting as is. I'll try changing it to 1 squirt/simultaneous and see what happens.

Thanks for your input. it is very much appreciated.

[varg]

1 squirt simultaneous vs 2 squirts alternating isn't any different from a req_fuel standpoint, but if the dead time is off the error compounds the more squirts there are. It's my experience that these 60lb injector go non-linear below 1.5ms and are basically unusable below 1.3ms pulse width. For reference, my dead time is also different, 0.4ms with a 0.17ms/V Voltage correction, I used this based on someone else's work on the MS forum back in the day, though I've never had the injectors flowed and characterized so I'm unsure if this is correct and I suspect it's a little bit off but can't send the injectors off while I rely on the car. I would save your tune and trial these settings + pull timing back to around what I'm using since you know this works for me and see how it runs. I have an upgraded fuel pump but nothing that would change low load fueling vs your setup. I'm not seeing any reason for your car to be missing at high rpm other than the fueling possibly being off combined with too much spark advance.

[Bizboy]

I changed to 1 squirt simultaneous, 0.4 dead time, 0.17 voltage correction, and copied your fuel and ignition tables (since the dead time was going to throw off all my values anyway, I figured that would be a better starting point). Obviously, I had to make some adjustments to the fuel table but unfortunately, It still has the breaking up issue above 4800. Those changes did however improve some other areas of the drive ability, so I'll keep moving forward using those settings. I'm definitely beginning to suspect that there is a part or parts that are not functioning properly at all. The reason being is that the tune does not stay consistent. I thought the super lean and surging idle was my cold/warm start settings but now I think it is something in the hardware side. For example I ran to the store the other day, and it ran fine (didn't go above 4800) staying at about 13-15 AFR (quick tune, not refined in the slightest). when I got to the store I was gone for maybe 10 minutes before coming back to the E30. When it started, it suddenly had a surging idle and would fluctuate between 16-20 AFR. But the real interesting thing is that for the rpm range it was surging in, all the VE values were the same. Except for when it would drop down in rpm, almost stall, and increase kpa to a point where it would dump more fuel in. Then it would drop to 16 ish AFR and shoot up in rpm again.

I know my IAC valve is in need of replacing so I figure I should start troubleshooting there. I'm thinking since mine is the discontinued L shape valve, and I'd have to rework the intake to fit the OEM T shape valve anyway, I might as well go with an E36/E46 valve since its all configurable through MS. Any suggestions?

[varg]

That sucks to hear, I was hoping it was just down to your tune being too lean at low load, high rpm or the timing advance being too high. Surging idle is tough to avoid in some setups. If you're running closed loop idle disable it until you can get a steady idle on open loop, because the ICV and control loop are too slow to stop a surge but they'll certainly provoke it. Disable closed loop fueling at idle too until it's dialed in. I have an M50/M42 two pin valve on mine, it's one of the only mods I made to the harness when I swapped, doing that mod would require the MS to be reconfigured to drive a two pin ICV.

I'm used to seeing irregularities or lost sync with trigger problems, but I guess it's possible that's still an issue. Have you checked the clearance between the CPS and reluctor wheel? How about the timing itself with reference to the O|T mark? Try to crank in flood clear mode with the cranking timing set to 0° and check that with a timing light. Back to super basics, check that you don't have a fuel delivery issue such as dropping fuel pressure due to a clogged filter or even the soft lines switched where they go to the firewall. Make sure the round connector under the manifold is clean and has good continuity too in case it's an issue with electrical connection to one of the banks of injectors.

[Digitalwave]

Keep an eye on your IAT and MAP values when these large changes are happening, too. I know when Varg was first tuning my E30, I had a faulty MAP sensor that was making us chase our tail. It wasn't until one night when I keyed on and saw it was reading 70 kPa (when it should have been ~103) that I realized the sensor was faulty.

[ADEN]

If you suspect the ICV just disables it from the setting, plug the ICV side air inlet, disconnect ICV connector and crack open the throttle body using the screw and adjust it to 800Rpm (after engine reached operatioperat temp).

I once have an ICV wiring issue (naked wires touching) that was sending 12v back to the board missing the whole controller performance causing huge ignition cuts registering super lean when revving wot and exhaust popping it took me long time to figure it out since symptoms where pointing out to an ignition problem and all my time was wasted replacing ignition parts with no results.