BMW E30 (M30 Swap) - No start after swap to MS3

[Lurker27]

Key info:
V3.57 main board
Was running MS2 3.0.3r previous, no issues
Running 3 coils, wasted spark, wired from the "LED" spark drivers to BIP373s (coils B and C wired direct from pin 8 and 9 on MS2 daughterboard)

I'm sure I've just got something simple wrong in the firing order, or maybe it was Wasted COP versus Wasted Spark (I only tried the latter), but if someone has a working msq for MS3 for a similar engine, that would be awesome.

It seems to find the cranking rpm OK, I know I've got fuel and spark IF all my settings were right since MS2 ran like a champ. I hear firing with MS3, it's just like it's only got maybe 3 cylinders? Just can't quite get it over. You can see some successful rpm spikes associated with it firing in the logs.

If anyone is running a similar wasted spark setup with MS3 in a BMW (60-2 wheel).

Thanks everybody.

[grom_e30]

did you re wire your spark A out put as in the ms2 file its set for js10 in the ms3 its set for the led's (d14)

[Lurker27]

nope, didn't realize that it would need all 3 LED, though that makes sense.-

should it be "wasted Spark" or "wasted COP"? Is my firing order OK? (does it matter)

So I'll rewire to pins 7/8/9 on the underside of the board. (right now, I was still running spark A through the Db37, and just added B and C in ms2 when i converted to wasted spark.

Thanks, this makes a lots of sense, i figured it was something simple... can't believe i missed it this long, really....


Alternately, does "JS10" in MS3 STILL FIRE the LED drivers?

[grom_e30]

Alternately, does "JS10" in MS3 STILL FIRE the LED drivers?

im not actually sure, you could try it and see if you get a config error and see if it runs.

you will be using 'wasted spark' if you are running 3 ignition outputs A,b,C. wasted cop would require 6 outputs a-f wired sequentially but fired in pairs A+D, B+E, C+F.

with any megasquirt it always fires its outputs a,b,c...... and so on, so your wiring needs to match so spark A= cyl 1&6, spark B = cyl 5+2, C = cyl 3+4 (153624 firing order)

[Lurker27]

ms3 attempt 2.msq

ms3 attempt 2.msq

Triumph and defeat!

I swapped my ignition A pin from JS10 to pin 7 and ran LEDs spark, wasted spark - turned right over and maintained idle, though extremely rough - maybe from sitting a couple months, but I figured once it's running I can smooth anything out.

My AFR didnt seem to be working at all, (unless MS3 now supports numbers under 10.5 and I'm REALLY THAT RICH) curiously, so maybe my fueling wasn't right - I messed a little with the multiply map setting but don't recall changing anything else. And I toggled that setting both ways, obviously.

So, no idea what I changed, but no I'm getting NOTHING. Just cranking. Ordered a spark tester last night to see if I am still getting spark but it sure doesn't feel like it. I went ahead and reloaded my MS2 based tune into MS3 environment thinking that gives me a pretty decent "restore point" that i understand, and still got nothing, so I'm pretty convinced it's not in software. Sync loss looks fine, have fuel pressure, smells like fuel's getting injected. Feels like maybe I blew ignition drivers somehow (short?) but all 3 at once? Not even getting pops. I did crank it a little with the spark outputs in the wrong place, but again, it ran after and intermittently sounded pretty smooth. I just can't fathom what's changed so much between a perfectly cromulent megasquirt 2 setup and the nightmare I'm having with ms3.

The engine also didn't seem to be pulling as much vacuum as usual but I'm wondering if I need to rescale my sensors or something?

Current tunes attached, if someone could find the time, but I'm pretty sure I'm back to chasing fuel, spark, compression. My logs were too big to attach here but don't seem to have any information that's relevant (duty cycles seem about in line with what I'm used to seeing from these injectors on startup. And it was JUST running, so I gotta believe I blew some hardware somewhere.

[grom_e30]

do you have a log of a failed start attempt?

[Lurker27]

Welp, We got it to idle, horribly. It can drive a little bit, but not much. Something is terribly wrong. Video of it "running"

https://drive.google.com/file/d/0B94sDC ... sp=sharing

Thinking maybe all this MS3 centric testing fouled the plugs, we swapped to 6 fresh new ones, no joy.

A log and my current MSQ is attached. As far as I can tell the fueling level based on duty cycle is identical to what it was when I was running MS2 - 5-7% when warming up at 60kpa and 20

MSQ: https://drive.google.com/file/d/0B94sDC ... sp=sharing

Datalog: https://drive.google.com/file/d/0B94sDC ... sp=sharing

I feel like this MUST be something with my sync, or still something with spark (placement or timing). I don't think I'm getting much sync loss, though it's very possible I don't really understand sync loss.

Looking at the guide to upgrade form MS2:
"Key differences Ignition mode selection is direct, chose your tach input source. Check your spark output! JS10 is standard on MS2/BG, this will be different on MS3. The regular wheel decoder requires a single "Tooth #1 angle" instead of skip teeth etc. It automatically
calculates the triggers itself and enables the required number of spark outputs depending on how many
cylinders you have and your selection of single coil, wasted spark, coil-on-plug or dual-dizzy."

I think I've got all that set, but I'm wondering if my Tach settings are somehow cross talking - disabling the Tach setting for now, and will try again later. Relevant/inspirational link below:

http://www.msextra.com/forums/viewtopic ... 01&t=60139

[grom_e30]

try cutting the number of squirts per cycle from 6 down to 2 that way you will get a bigger pw at idle and hopefully more even fueling whats happening with the afr in the log its all over the place.

you also have include aft target set to on but appear to have the lambda disabled not sure if thats ideal either

[Lurker27]

I will try that, although I doubt it's any problem with fueling at this point - I cross referenced some old logs when it was running on MS2 and the Duty Cycle #s are similar. (~5% @ 1500rpm, 60kpa, which is how the cars tends to warm up w/ gas pedal slightly depressed)

My AFR are all over the place because my calibrations didn't come over with the msq - my wideband (NAW) is a 0-5V for 10-20AFR linear model, and I added that in last night. Sorry about that.

We also tried, just to make sure it's not anything explicitly in hardware, starting the car with each of the 3 coil packs disconnected. It's more or less equally terrible for each of them, so I know it's not just sparking on X cylinders (where X<6).

It's like the timing is just wrong, but I loaded my OLD WORKING tune to make this one, so how can that possibly be the case unless MS3 handles timing differently? And even so I don't know that I even have noticeable sync loss. (I get 1 count at the very start, but then it stays unless the engine stumbles to near 0 rpm. Reason code 2)

[Lurker27]

Starting to question some really dumb stuff - I toggled multiply MAP off and on a few times but it should definitely be on for my VE table.

Also I'm pretty sure my trigger wheel setting is correct. http://www.msextra.com/forums/viewtopic ... 01&t=39084

I'm coming down on maybe it's in MS3 noise filtering? Would that not show up in sync loss?

[grom_e30]

did you try changing the squirts per cycle? have you got any new logs? i think the tooth#1 should be about correct depending on what way you wired up the sensor will give about 3deg difference.

when you say noise filtering do you mean on the crank sensor input?

[Matt Cramer]

Google drive links are not working; can you attach them to your post instead?

[Lurker27]

Sorry about that - Some previous ones were too big. right click save as seems to help but I'll also attach here.

Some advice I got form an E30 guru is to not import an MS2 file and edit it, but to start 100% from scratch. I'll definitely try that + the squirts setting this week.

I did mean input filtering - but from my log I'm not sure if sync is a problem.

[Matt Cramer]

Not seeing a clear loss of sync.

Is the O2 sensor hooked up correctly?

Can you test for spark on all six cylinders in output test mode?

[Lurker27]

I confirmed all 3 coils are sparking.

O2 is hooked up correctly, just was not calibrated.

[Matt Cramer]

OK, a log with a correct O2 reading will give a bit more information on what's going on.

[Lurker27]

Starting with a fresh file seems to have resolved the issue. Weird.

[Matt Cramer]

Starting with a fresh file seems to have resolved the issue. Weird.

I've seen that happen fairly often - get a junk setting buried in some really odd place, and it can be easier to make a new tune than track it down.