Ok so i've been trying to solve a misfire problem for a while. At first i thought it was just lean misfires. Went richer and eventually progressively swung WAY rich where there was definitley some rich misfiring going on all over the map, not just where the original misfire was. Switched to E85 because of a number of reasons. I just increased the required fuel amount by 35% (req_fuel*1.35) and it started right up and ran decent. Did a bunch of pulls with MLV VE analyzer and through the progression of the "analyzed" tunes, you could clearly see that it was making the map richer from the original misfire point outward (cuz it shows lean when it misfires, so MLV made it richer, which caused more misfire, which spread out after each VE analysis). So i had the idea, instead of approaching from rich, approach from kinda lean (13.5 afr) just in case the original misfire was caused by an overly rich environment initially. I set the VE values leanish from 4k to 5.5k rpms and did more logging. I finally reallized that it's for sure not a fuel problem after doing this because the logs show that part throttle (2 to 3 psi) misfires at 4500ish rpm and so does full boost (only 6-8psi) no matter how rich or lean the mixture is after several VE analysis'. I even tried varying the ignition advance from 0 to 20 degrees in the bins ranging from 120-160kpa, 3900-4900rpm in case the advance was causing the misfire, but with no avail. Problem has shown up in datalogs to definitely be some kind of misfire around 4300-4700 rpms. It shows up pretty much anywhere above 100kpa (basically in boost), and appears to be more of a glitch or electrical noise, and not simply poor fuel or ignition tuning. Please help if you can..
Setup is 1998 BMW M3 S52US with TRM Wiseco 8.5:1 Pistons, OEM headgasket, OEM rods, MS1 V3.0, hall effect, wasted spark on OE coil packs with 6 BIP373's, 65 lb/hr Bosch injectors, 255lph Walbro intank pump, T3/T4 .53 trim turbo (kinda small), large front mount intercooler, mesh wire filter (intake doesn't fit, yet, still fabbing new motor mount), eBay bottom mount tubular SS manifold, E85 fuel. 4340 lightweight flywheel and 500 tq rated F1 clutch, VANOS disconnected. Thats all I can think of off the top
Those duty cycles spike up because the RPM/100: show 150ish which is 15,000 rpm or the RPM numbers are clearly spiking which are causing misfires because the megasquirt is pulling timing and fuel in an effort to curb revving due to the rev limiter being utilized at 7k rpms.
Anyway, the problem is that the car misfires around 4500 rpm under boost no matter what the A/F or ignition timing is. In first and second gear its barely noticable (just a loud backfire, but could easily be the rev cut since its spinning the tires). 3rd - 5th gears are more obviously a misfire from 4300-4700 rpm because it is more difficult to move through that rpm in higher gears, unless you start above 4700, in which it takes off like a rocket. So there is a definite misfire around 4500 rpm.
thanks for the help!!
sam
Silver 1998 M3 Coupe | M50 | Eibach Pro | Home Welded Straight Pipes
Are you using low or high impedance injectors with this setup?
I'll check back here but if you're using low impedance you may want to PM me for a faster response.
I'm a little confused, you stated and your data shows the horrible rpm spike which of course would cause a misfire. Is that the point in the data you are refferring to? If so, obviously something is interfering with your crank signal. If you are using the trigger wheel on the crank pulley, check to make sure your crank pulley bolt and your crank sensor bolt are tight. Also see if you have something triggering at 4500rpm that is causing the rpm spike (vanos?).
Your ignition timing map reminds me of this
Last edited by charger758; 12-08-2011 at 10:37 AM.
You might want to check the megasquirt forums as well (in case you haven't already done that).
-Kenny
'89 535i Twin Turbo Project
Gentlemen, apologies for the late response to your posts, however I was doing field work all day and out of the office.
Kevin325i - Yes they are low impedance! Hopefully that means you've seen a similar problem before and may have a solution!? 65 lb/hr bosch, using 1.1ms opening time, 0.20 batt voltage correction, 27% PWM current limit and 1.1ms PWM time threshold (I'm assuming these parameters are fairly close to correctly modelling the injector since the switch to E85 was an easy req_fuel increase that worked well across the map and even idled well). Anyway I will PM you as well, as you requested. Thanks
Charger758 - I definitely could have been more clear in my post, but yes you are correct that is the point i was referring to. I do believe that the RPM spike is causing the misfire, as you stated, however I'm not sure what is causing the RPM spike. RPM noise is definitely not an unusual problem for a VR based megasquirt install, and sometimes can even occur in compeletly custom rigged, ie. non shielded, hall effect based setups. However, the trigger wheel (which is on the crank inside of the block), the hall sensor, the bracket (which theres not really a bracket, rather a hole on the side of the block that the sensor bolts in to), and the shielded wiring running from the sensor through the harness to the DME clip are all OEM and untouched. If the electrical noise is minimal enough in this setup for the DME, its minimal enough for MS. Now I have tapped into these wires in the harness right at the clip and run it to the MS, but even that is done with a little extension of shielded wire just for shits and giggles to cover my noise bases.
I believe you're on to something with the idea of something being triggered at 4500 rpm, I also thought of this and haven't ruled it out, but can't think of anything. In fact that's part of the reason why I thought bfc would be a better place to post than just the MS forums, since something randomly triggering would be a vehicle platform specific issue that the general ms forums would not pick up on. The vanos i think does retard the cam again somewhere in that 4500 rpm range (maybe it stays engaged longer and retards higher, i'm not sure), but regardless, the vanos is unplugged and more importantly, the OEM DME is not even in the car. I don't think any other sub systems or modules in the car would trigger much of anything at a rpm specific point that the DME doesn't initiate (a/c, etc).
Hahaha you are ABSOLUTELY right about that map, its ridiculous. I've honestly paid very little attention to the timing map, except for tweaking the timing up and down to see if it's the possible cause of the misfire, and so it has stayed very, very conservative and largely untouched since i was using it on the stock motor and it was running into detonation at numbers as low as 10 degrees at 6psi. I'm suspecting that was partly a fueling issue. I did verify base timing with timing locked out on the MS, but really don't completely trust yet that the ms is really sparking when it's supposed to at higher rpms, so i'm fine running FULL RETARD until i get the fueling gremlins taken care of.
eseeling - excellent advice and yes I have checked through quite a few postings. Even found some very similar issues that either never got resolved or solutions were never posted. Problems like the one I have, have forced people to re-do SOO much wiring and buy SOO many new parts, and yet the problem may never get fixed, or even worse, it only "kinda makes it better". I want to hopefully diagnose the real problem and fix it instead of shotgunning at all possible fixes, if that's possible. I'm getting desperate tho..
Captain Bondo - OBDII M50 based motors (m52,s52,etc) use 3-wire hall effect sensors. While the standard 60-2 trigger wheel is still on the front of the crank pulley, the spot for the VR is vacant. The back of the crankshaft has a new, smaller 60-2 trigger wheel machined onto it. I believe BMW did this to avoid false misfires due to crank torsion if I recall correctly.
I do not have a tooth log available. The spike does not occur in neutral or low load conditions through the 4500 rpm range, only around the 100kpa range and up. A tooth log could be useful, however if it was a problem that a tooth log would explain, i would think it would happen at all load conditions and probably not ALWAYS at 4500rpm. Either way, I can try to get one to post up, as it could be useful.
BTW, i have noticed that every once in a while, it will do a clean pull with no evident misfires. This has only happened a handful of times, however, and the misfire always comes back without making any other changes.
Another issue i have noticed is that sometimes when it gets stuck misfiring and backfiring at the 4500 rpm range or after bouncing off the rev limiter, the engine sounds and feels like it runs on 5 cylinders for several seconds. Typically either revving the engine a couple of times or turning the car off and on real quick will bring the 6th cylinder back into commission. This is obviously either a fuel injector sticking or otherwise not opening, OR a plug that stops firing. maybe an overheated coil, or an overly wet plug from excess fuel from the rev cut/misfire (i have it set to ignition cut only) or possibly an overheating BIP373. I don't know. I can't even figure out which cylinder it is because when it happens, by the time i pull over and pop the hood, the cylinder has come back to life and the problem no longer persists. This only happens on occasion, NOT every time it misfires or rev-cuts. I don't know if this is related to the misfire or not. Also, for reference, I'm running colder plugs, NGK BKR8E i believe..
Thank you all for the input!
-Sam
Silver 1998 M3 Coupe | M50 | Eibach Pro | Home Welded Straight Pipes
i was tuning a Aem car that was cutting out like that. Ended up being cause of the low impedance injectors. Aem could not run them and the customer ended up getting a resistor box. and how the car pulls to redline.
RK tunes
Gotcha. That gives you a fighting chance. I fought with a similar setup with a vr and just could not get it to work consistently. I finally bought a vems and had zero issues. But anyways i did learn a few things that helped. Are your bips mounted in the case? This caused me all manner of grief. If they are, hopefully you have a dedicated ground wire and are not using the circuit board in any way. This also gave me grief. Speaking of which double check all of your grounds if you haven't.
-Kenny
'89 535i Twin Turbo Project
Roadkill1991 - That is really interesting.. I may as well just try some resistors and change settings in MS as if they were high-z. I'm sure theres plenty of information on the web about running resistors with low-z injectors on a high-z ecu, however do you know what rating resistors you used? I should just do some research..
How did you end up determining that was the problem? By calling AEM and they just told you it was probably the injectors? thanks!
Captain Bondo - BIP's are mounted inside the case, to the case using the proper mica insulators, rubber boots, etc. Wires tee off from the LED's through 6 resistors to the 6 BIP's as per jon's EXCELLENT m50/m52 megasquirt guide. Each BIP has a ground wire running from it, out of the case, to a grounding bolt on the chassis with each BIP having it's own wire and lug all bolted to the same spot. The gate wires go from the BIP to the harness where they tap in to their respective coil wires in the OEM harness. I used to run the DME in parallel with MS so it would run things like the tach (cam phase sensor was adequate for tach, even tho the crank trigger never reached the DME), IACV (i miss not having to give it throttle until it warms up enough to idle), air conditioning, cruise control, and most awesomely to turn on the main relay, fuel pump, etc.. which i'm currently having to do with a manual switch under the hood.. super awesome for daily driving, not. It turns out a proprietary diode that me and some EEE buddies are trying to spec out, burned up in my OEM DME along with a melted trace on the board. Looks to be a voltage comparator circuit for making sure the power supply to the dme is clean, but, needless to say I still haven't gotten the DME running again. What this means is that there are some sketch temporary jumpers from the DME harness clip to various grounds and a switch so that the car will run and MS will get power, but I really don't think they are the cause of the problem.
MS taps into the same power source that the stock DME got power from, but the DME has to actively switch the main relay on to give itself power once you turn the key.. so basically when the DME fried, MS stopped getting power along with many other systems on the car. I've just manually jumpered and added a switch to those circuits so that i emulate with a switch what the dme did everytime it powered up.
This.. has been an adventure.
Silver 1998 M3 Coupe | M50 | Eibach Pro | Home Welded Straight Pipes
Sounds like numorous issues in your setup. Take the BIP's out the case! Very important. Nothing will cause more noise than this. I made a box and used a regular 1/8th thick x 3/4 wide x 3" long strip of aluminum as a heat strip. Then mounted the BIP's as standard on opposite sides by 1 screw (for V8). Make sure the heat strip/BIP's do not contact the box (my box is metal). Box is very compact and everything was wired through a 9 pin connector to plug into the MS unit. Make sure all the BIP's have dedicated grounds and all grounds through entire system are grounded to 1 lug. I think i would also switch to the oem VR sensor and circuit but its not neccessary. Also you should try high Z injectors.
I had a similar problem with a MS project using low z injectors. In fact your problem sounds exactly like mine. I thought I had tried everything until I wired up some resistors to switch the injectors to high z and it worked perfectly. The flyback circuit causes a ton of noise in the ground. Way more so then the BIP373 coil drivers ever did for me. Your problem may be different, but my problem was the exact same. There was a phantom misfire that showed up around 4500rpm. Or as the turbo spooled and IDC started to rise.
Also, check your logs when it pulls clean. Is it after the car has just started? I found that my problem was more susceptible to the noise issue with low z injectors after the car had run for a few minutes. I never really noticed this because I normally let the car run for a few minutes even if it was already up to temp.
Feel free to PM me if you've anymore questions.
I'm more of a reader/lurker than poster but, if you decide to keep the low z injectors you may want to check this out http://www.jbperf.com/p&h_board/index.html. Personally high z are alot easier to work with from the few cars I've installed them on.
Last edited by Specs; 12-09-2011 at 10:47 AM.
-Kenny
'89 535i Twin Turbo Project
Ok guys, looks like I have something to work with now. I will try throwing some resistors in the mix to avoid any potential PWM issues and will build a little external dedicated coil driver box. I know some high end ecu's like motec avoid putting drivers within the ecu case because of claimed heat sink and noise issues, so I may as well follow suit. It definitely can't hurt anything and will cost very little to implement. Again thank you for the input lughed, kevin325i, specs and Captain Bondo.
specs - as you can see from my post count and join date.. i'm definitely a reader/lurker type as well. The external injector driver board you linked to is very interesting, thanks for sharing. For now, I think I will try the crude, yet elegant, method of using resistors
Silver 1998 M3 Coupe | M50 | Eibach Pro | Home Welded Straight Pipes
do you have access to a different set of coil packs? i'd swap them out and make sure you don't have a coil pack that is breaking down and causing the misfire.
my .02
coming soon, S54's for everything....
I've always built the low ohm flyback circuits into MS units but have had zero good results ever. Low-Z injectors need to be controlled with a P&H box. Put high ohms in it. Start there. Then look into shielding.
re the coil drivers
i just hacked up a stock dme and used it as an interface
used stock gnition drivers in stock location
Long time E34 owner
Now E39 530D touring Sport Auto
Oh man you guys are awesome. The M is fixed.
After you all brought up the problem with the low-z flyback circuit, I started looking into that more. I don't know how I missed that apparently huge problem when throwing my build together (i.e. sourcing injectors). Apparently there is a well documented fix for the MS board that involves running a clean +12V from before the injectors directly to the center legs of the two TIP125 injector drivers and either cutting the 12V trace that those legs went to on the board or simply desoldering and pulling them out of the PCB with needle nose.
I went ahead and pulled the legs up and ran the dedicated +12v for the flyback and took it for a spin. Interestingly, if anything, the noise and problems got worse. I then verified all of my MS grounds and Ignition coil drivers, cleaned up the grounds a bit. Test drive resulted in no change. Then I re-wired the LC-1, which had been tapping directly into the MS harness for power and grounds. I thought maybe the heater circuit or something was causing noise. After re-wiring the LC-1, it still had no effect on the misfire and RPM spike problem.
Went to Fry's, picked up 6x 8.2 Ohm 25 Watt resistors. Stacked them together and wrapped it kinda like a battery pack. Threw them inline with the injector ground wires going to the MS (it doesn't matter which side +/- cuz it resists current either way). Opened up Megatune, increased the opening time a bit, changed to 100%PWM and 25.4ms (the settings for high-Z injectors). Went for a test drive and .... Beautiful.
My datalog looks NOTHING like it ever has before. Everything is SO much cleaner. The most noticeable being NO rpm spikes and the WBO2 log is sooo stable. It's unbelievable. I can't believe I have been wasting my time for so long trying to tune the MS when it had all of that noise with the low-Z injector settings and misfires. I finally got those resistors installed at like 11pm last night and spent like 1.5 hours tuning the fuel (which is so easy with a stable O2 compared to what i was working with before, OMG), and started bumping the timing up. I've upgraded to just retarded and no longer hold the title for FULL Retard.
I'm blown away how this car feels with 6psi, 16*, 8.5:1, E85 and a 1.5 hour super conservative tune. Second gear feels like a sled in the snow (but maybe with a rocket strapped to it) with no traction. This is a small turbo, too. You guys with big turbos and high boost.. oh man I can't wait to upgrade. That said, I cannot tell you how extremely happy I am to have this thing going finally. I'll start bumping up the boost after I get more polished maps. I expect to run approx 20-22 max psi, if my turbo will flow that.
Now I have another concern as far as timing tuning. I saw another thread posted recently that touched on E85 tuning. With a knock-limited fuel/motor, street tuning is straight forward you slowly ramp up to ping then back off a comfortable amount. E85 doesn't knock tho.. so I really need to find MBT (hopefully before I find the limit of combustion pressures my rods/headgasket/etc can handle! lol). So the question is, "How do I make sure I don't pass MBT without a constant load dyno?"
I think what I will do is measure how long it takes to go from a certain rpm to another rpm in like 4th gear (for example 4000-6000) on the same road, same direction, same day and compare the two timing map setting logs to see if the engine is still gaining torque. Once I can't measurably see any reduction in time, i'll knock it back a couple degrees for safety and run with that. That's the plan until I just get some constant load dyno time in somewhere.
Suggestions?
Thank you so much guys..
-Sam
Silver 1998 M3 Coupe | M50 | Eibach Pro | Home Welded Straight Pipes
Honestly, a dyno is you're best bet by far.
If a dyno is 100% out of the question no-way-aint-gonna-happen then start with a pump gas timing table and add a little advance where it seems laggy. Obviously there will be power left on the table but without a dyno it's pretty much unavoidable. You can watch the rpm slope for a good measure of acceleration in a single gear, but if anything changes like a passenger, slightly uphill, lower fuel, etc it's just going to be a pain.
An engine running e85 CAN be knock limited but it's very unlikely an 8.5:1 engine without very high manifold pressures will be able to see this happen.
Last edited by Kevin325i; 12-23-2011 at 03:57 AM.
I am suffering with same problem, and quite frankly it's driving me mad! The problems described by Sam are literally identical to the troubles i am suffering with, however i am running an M20B20 with a V3.0 self built megasquirt and the standard VR crank sensor etc. I'm running Sequential CNP and injection, with a VR type Cam sensor too. I reach 4000RPM, it misses in an incosistent way preventing any further RPM increase, and if you stick with it (full throttle) after a few seconds it nearly always (sometimes does continue to miss less often though) clears and goes like a rocket right up to the soft limiter at 6,200RPM. I don't believe it to be related to Fuel table or Ignition as it will present the same 4000RPM misfire through to just over 5000RPM whilst under no load being revved in neutral / stationary. I can't find any hint of the chronic 60-2 tooth RPM loss, even so i have fitted a 20k resistor in the Crank sensor signal, and looking at the Log, i can't see any spiking in the RPM signal - only the actual misfire causing small peaks etc in the RPM signal. This all seems identical to the problems described above, however i am running High impedance injectors. So i am at a loss! any help would be much appreciated. Either that or i give up and just jump back in the M5 - although very unreliable in itself, atleast it doesnt have a bloody untraceable misfire!
here is a link to the MSL files and a screenshot of a data log i took of the misfire. it was full throttle pulls in second, third and fourth gear.
https://drive.google.com/drive/folde...7D?usp=sharing
thanks,
Sandford
Matt Cramer
1997 BMW 328i convertible, 1972 Chevy C10 pickup, 1966 Dodge Dart slant six
BMW - where "Why doesn't everybody build cars the way they do?" meets "Why can't they build a car the same way everyone else does it?"
Bookmarks