MS3v3 System for CNP and EFI/Turbo "Beetle"

[kangaboy]

I do. I've honestly only used the JimStim upon initially soldering and building the MS.
I'll go to the garage right now and remove the MS from the car and do some reading on how to test things, unless you have some specific testing in mind.

[Schweg]

I do. I've honestly only used the JimStim upon initially soldering and building the MS.
I'll go to the garage right now and remove the MS from the car and do some reading on how to test things, unless you have some specific testing in mind.

Nothing serious I just feel like you could bench test more things then in the car as you have tested all the external items to the ECU and they work


Sent from my iPhone using Tapatalk

[Schweg]

I do. I've honestly only used the JimStim upon initially soldering and building the MS.
I'll go to the garage right now and remove the MS from the car and do some reading on how to test things, unless you have some specific testing in mind.

Nothing serious I just feel like you could bench test more things then in the car as you have tested all the external items to the ECU and they work.

Also I was thinking, and I’m not 100% sure how a jimstim works but you could try and recreate your issue to isolate the problem.


Sent from my iPhone using Tapatalk

[Piledriver]

Can you add a short current hardware setup in first post?

If you have ls2 or ls2 truck coils your dwell settings may be an issue.
I use...
4.2 dwell (3ms makes for a lame or no spark, depends on the coil)
2.0 spark duration
Looks like you are using the proper GM battery voltage dwell compensation table...

LS2 coils are only used in fully sequential setups from the factory and have gotchas on a waste spark setup.
So do big injectors on a little motor.
Might want too go sequential sooner than later.

Still going through it.

[kangaboy]

I have the cam sync sitting on my work bench. I wanted to get this issue figured out before I added any more variables to the equation, but with the way you have explained it, not going sequential could be an issue in itself.
At this point it has to be coils or injectors. I'm one step away from removing them both and swapping out the coils with another set, and sending the injectors out to be flow tested.
I'll put the cam sync in this afternoon and see if I can't get it all wired up and programmed correctly.

[panel]

Looks like you are using the proper GM battery voltage dwell compensation table...

Sorry to but in Pile and kangaboy , I didn't know this was a thing. Where can I get a copy of this table ?

[Piledriver]

Kangaboy--I'd change the dwell before doing anything else, it should fix things instantly if it is the issue.
I have one or two ls2 coils that don't really fire at 3ms dwell.
Not trying to discourage going sequential, just "one change at a time" for troubleshooting.

Looks like you are using the proper GM battery voltage dwell compensation table...

Sorry to but in Pile and kangaboy , I didn't know this was a thing. Where can I get a copy of this table ?

gm-ls2-dwell-batterycomp.png

[kangaboy]

Changed dwell to 4.5ms and the spark duration to 0.7ms.
Acted the same as it had been when taking it out of the driveway and down the lane last week, so I figured it didn't have an impact.
After that I messed with the Fuel VE map and smoothed a few areas out on the 3D map that seemed like they needed some help. I immediately noticed changer here...enough that I was able to drive it up to the gas station and back. I logged the trip and I'll attach it below. There were two events that spiked my AFR and caused spark cut, resulting in a massive backfire when the criteria was met for the motor to run again. I was confident in driving it until that happened.
So with that said, I guess smoothing the fuel map helped, and maybe adjusting dwell as well, but I still feel like there is an underlying issue that caused the couple AFR spikes that I still need to solve.

[kangaboy]

Here is the log I was speaking of.
Right at about 211.00 seconds you can see the start of the event, and the Pulse width goes to zero until the motor comes back down to 2000rpm.
Nothing jumps out at me besides the AFR is swinging wildly and goes up to 20+. Fuel pressure looks good the whole time as well.
So NOW I think that it may be a bad AFR module or sensor...because the car seems to drive fine, other than the wild swings, and the spark/fuel cut that they are causing.
FYI, I found a fella using LS2 Truck coils in wasted spark as I am in the Miata forums and based my current dwell timing off him. Leaving the dwell battery connection as I previously had it, according to the log, it was running about 5.3ms of dwell for the majority of the drive.

[Piledriver]

at just prior to the lean spot in question, the operating point dives deep into fuel cut, and lingers in the 19:1 target area on the 30kpa line...
the result is the following lean spot. This may be due to wall wetting, due to any ful on the walls being stripped off during that hard decel/lean period.

I hadnt used MLV in awhile, bit it appears it now has been upgraded to use the same 9 point delay table as VEAL, rather than the single delay time it traditionally had. Thins means you can now effectively fix your tune from logs, rather than in real time,
(I assume this is only in the registered version)

It also means you can play with the delay table and see if the results follow the delay correctly at your leisure, and with easy undo.

If your fuel flow numbers are sane, the MSextra firmware is really good at calculating tings... you can tune a solid ve table with all non-boosted targets at (for example) 12.5:1, then once you have a solid ve table, change your afr target table, and the resulting afr will be correct.

[buguy]

Is it bad that I have FI but no clue what the hell you just said?!

[kangaboy]

Thins means you can now effectively fix your tune from logs, rather than in real time,
(I assume this is only in the registered version)

It also means you can play with the delay table and see if the results follow the delay correctly at your leisure, and with easy undo.

I actually just noticed that this was possible when i sent the most recent log.
When you open a log, you are able to load your tune as well.
I will have to experiment with this.
How do you feel about the dwell being set at 4.5ms and running at 5.3ms with battery compensation, opposed to the 3ms it was initially set at?
I am confident now with the new fuel pressure regulator and paired fuel pressure sensor that my fuel delivery is adequate. Need to look into the wall wetting that you are speaking of, get some runs in, and get some more logs.
Thanks a ton for the help btw...I feel like adjusting the dwell and VE table, and in-turn giving me a drive-able tune, makes me want to start working on the bug again.

[Piledriver]

Buguy: sorry about that
It makes more sense if you have the log and tune loaded in megalogviewer and play it back...
Wall wetting is literally the fuel that sticks to the port walls, its significant at low to medium loads at least, and the firmware can account for it.(probably does by default)

Get set up for sequential---its easy, and will help tuning at low loads with your big injectors.
Question: you are set up with your fpr manifold referenced? (boost/vacuum)
I expect yes.

I suggest the dwell and spark duration I posted and use, as that is what GM supposedly uses/matches the coil characteristics.
Shortening dwell etc so you can run 7K waste spark is not needed in your case, as you have the needed parts.
If you want to play with batch fuel and waste spark you can just set it back to that in software.

Once you go sequential, you will likely find injection timing matters, from what I have seen, start injection as soon as the intake closes, with our hot heads and valves I suspect it largely vaporizes before the valve opens.(needs less fuel for a target afr)

One more thing: get some nice fast sdcards, you can log most things that matter at ~400hz...

[kangaboy]

I suggest the dwell and spark duration I posted and use, as that is what GM supposedly uses/matches the coil characteristics.

Whoops, missed that in your original post. I'll update dwell to 4.2 as suggested.

Yes, fuel pressure is boost referenced at 1:1 increase.

Full sequential is my goal for this week, and I have all the parts required. I'll have to do some quick reading and figure out how to appropriately orientate my dist drive gear and put it back in, and then get the cam sync wired up.

[buguy]

Yeah get that figured out so you can tell me how to do it in a few weeks!