MEGAsquirt A place to collectively sort out this megasquirt gizmo

'92 ms3 - extreme accel enrichment needed

Thread Tools
 
Search this Thread
 
Old 11-25-2022, 06:17 AM
  #1  
Newb
Thread Starter
 
Alexj84nl's Avatar
 
Join Date: Nov 2021
Posts: 19
Total Cats: 1
Default '92 ms3 - extreme accel enrichment needed

Started a topic yesterday, and a new one today.
I'm running into weird issues with my megasquirt project.
Full specs at bottom, but its a 1992 mostly stock N/A project.
Just getting it to idle correctly, or not stall when pressing the throttle (very slowly), seems way to hard.
After hours of testing, reading and adjusting, I really need some help.

So, the base map shipped with my MS3 from rev didn't work for me.
Hard to start (in cold), running wrong AFR, cannot hit the throttle or it goes lean and stalls.
Weird, right.

- starting for now is OK, AFR around 13.2 and warms up OK.
- Did the idle valve test / configuration (29-96 %)
- idle is around 850rpm, and when I adjust VE map to 14.9, it stays there. But when I change RPM from idle and return to idle, it is always off (just now, 14.2). I can then adjust VE map until 14.9 again, but changing RPM's repeats the behavior. I disabled EGO correction (authority to 0 and only active above 1500rpm)
- From idle, pressing the throttle very gently (even slower than a noticeable TPSdot change), makes the AFR spike up to max AFR (lean) and stalls the car. Tried to tune acceleration enrichment by switching time based and pump based, applied extreme enrichment values, and then the car will not stall but still it's not very smooth. See screenshots/logs.

I think it is really weird that this happens and something is not a bit wrong, but totally wrong.
What could it be?
Didnt have any issues before installing MS, ran 100% perfrect. Maybe the gasket from the throttle body is leaking? Could that cause these problems? (I had to remove TB to install VTPS).




'92 USA California spec, totally stock.
Removed stock ECU, installed MS3mini. Got Vaccuum from fuelpressureregulator. Deleted AFM, installed GM IAT sensor in airfilterbox. (of 1.8)
Also installed a VTPS and Innovate MTX-L. (and calibrated both)
Attached Files
File Type: csv
2022-11-25_11.58.12.csv (1.92 MB, 28 views)
Alexj84nl is offline  
Old 11-25-2022, 12:11 PM
  #2  
Junior Member
 
Panici's Avatar
 
Join Date: Aug 2019
Location: Canada
Posts: 145
Total Cats: 10
Default

Still have your original ECU? Swap that in to rule out any issues with the throttle body.

Perhaps your areas of the VE table that you hit when you're blipping the throttle need attention?
AE just builds on top of these areas, it can't fix a bad VE table.
Panici is offline  
Old 11-25-2022, 02:49 PM
  #3  
Newb
Thread Starter
 
Alexj84nl's Avatar
 
Join Date: Nov 2021
Posts: 19
Total Cats: 1
Default

Thanks.

About the VE map. It is not spot on, but okayish. The more I think about it, Lightly tapping the throttle shouldnt be an issue.

swapping back to original ecu would require me to mount the oem tps again, and other stuff. Not looking forward to do that. But I could try.

What avout trying the base tune from diypnp, would that one work with the mslabs ms3mini?

Alexj84nl is offline  
Old 11-28-2022, 11:10 AM
  #4  
Newb
Thread Starter
 
Alexj84nl's Avatar
 
Join Date: Nov 2021
Posts: 19
Total Cats: 1
Default

I did check the TB with some (flammable) brake cleaner, and I did not notice any RPM change during idle.
I guess there is no vaccuum leak.

Before I switch back to OEM ecu, let me know what you guys think of this log.

First green hill/spike is me pushing down the throttle.
What you see is that simultaneously the RPM declines till 470 rpm (AFR went from 14.7 to 15.7).
Then, the TPS reaches 17% and during this (short) time, you see a massive AFR spike of 21, followed by a AFR of low 11. (while the TPS is still at 17%).
So a lot of variables change in a very short time interval.

My feeling is that acceleration enrichment will not do anything in that very first moment of me hitting the throttle.
I keep reading that my VE map should be good to fix this problem.
But my feeling is that the RPM drop and AFR spike (rich) has to do with some kind of other problem.

What is the white PW line in the graph at the bottom? That reaches 32 (ms).
Injector dead time? Should that go up to 32?


Attached Files
File Type: mlg
2022-11-28_16.34.06.mlg (291.5 KB, 13 views)
Alexj84nl is offline  
Old 11-28-2022, 01:02 PM
  #5  
Junior Member
 
Houleo's Avatar
 
Join Date: Jul 2019
Location: Albany NY
Posts: 53
Total Cats: 12
Default

if you uploaded your tune you may get some good advice. Since this is a CA car I am assuming you have sequential injection? Is your MS and tune set to sequential or batch injection?
Houleo is offline  
Old 11-28-2022, 02:13 PM
  #6  
Newb
Thread Starter
 
Alexj84nl's Avatar
 
Join Date: Nov 2021
Posts: 19
Total Cats: 1
Default

Thanks for the help!
I've got a '92 miata. Afaik only 1993 had got the sequential injection.

I asked rev if it is ok to post his (altered) base map over here. If so, I will.
Alexj84nl is offline  
Old 11-28-2022, 10:26 PM
  #7  
Newb
 
Nethawk's Avatar
 
Join Date: May 2016
Posts: 15
Total Cats: 0
Default

One thing to keep in mind, a misfire from being too rich will show a lean AFR. If the cylinder misses, the wide-band will read all of the un-burnt O2 and show lean. PW is huge during AE in your screenshot. I would reset your tune to a baseline for acceleration enrichment.

Your tune, and a log will help troubleshoot. A log with a slow pedal apply showing the lean stall would be most useful.
Nethawk is offline  
Old 11-29-2022, 07:29 AM
  #8  
Retired Mech Design Engr
iTrader: (3)
 
DNMakinson's Avatar
 
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,009
Total Cats: 857
Default

Originally Posted by Houleo
if you uploaded your tune you may get some good advice. Since this is a CA car I am assuming you have sequential injection? Is your MS and tune set to sequential or batch injection?
^^^^ You need to post your tune. Best if all posted logs correspond with said tune.

Originally Posted by Alexj84nl
Thanks for the help!
I've got a '92 miata. Afaik only 1993 had got the sequential injection.

I asked rev if it is ok to post his (altered) base map over here. If so, I will.
Are you saying that Rev has helped you on the side with an updated base tune?

DNM
DNMakinson is offline  
Old 11-29-2022, 08:28 AM
  #9  
Newb
Thread Starter
 
Alexj84nl's Avatar
 
Join Date: Nov 2021
Posts: 19
Total Cats: 1
Default

Thanks for the help, really appreciated.

Quote: "Your tune, and a log will help troubleshoot. A log with a slow pedal apply showing the lean stall would be most useful."
The logfile and screenshots provided 28-11 are from me applying slowly the throttle pedal. If I do this slower, the car won't even rev up, only down and stall.

Quote: "Are you saying that Rev has helped you on the side with an updated base tune?"
Nope, I received a base file for my car (told Rev specs of mine), and I made adjustments to VE, IGN, and several startup and idle maps. O, and also to the accel enrichment settings.

He was OK with me sharing the tune, so here it is, Hope it helps.
Attached Files
File Type: msq
CurrentTune.msq (274.6 KB, 12 views)
Alexj84nl is offline  
Old 11-30-2022, 03:10 PM
  #10  
Newb
Thread Starter
 
Alexj84nl's Avatar
 
Join Date: Nov 2021
Posts: 19
Total Cats: 1
Default

Went over to diyautotune and took the basemap of a mspnppro (for 90-93). I compared this basefile to the one I had, and made lots of changes to mine.
The resulting tune is attached.

I think it is a step in the right direction.
Car is not firing up good, will have to tweak that again.
When it's warmed up, I can slowly press throttle without stalling. Upon releasing (gently) throttle, it stalls half of the times.
At least I was able to drive around and make a logfile (large one). The smaller logfile is from idle and adjust throttle, releasing it, and a stall.

And again, feedback is really appreciated.
Attached Files
File Type: msq
CurrentTune.msq (274.6 KB, 16 views)
File Type: mlg
2022-11-30_20.50.49-2.mlg (1.48 MB, 16 views)
File Type: mlg
2022-11-30_20.53.25.mlg (156.3 KB, 13 views)
Alexj84nl is offline  
Old 12-01-2022, 09:21 AM
  #11  
Newb
Thread Starter
 
Alexj84nl's Avatar
 
Join Date: Nov 2021
Posts: 19
Total Cats: 1
Default

I analyzed the logs some further, but could use a bit of directions where to look at.
See this screenshot for example.
I lift throttle. AFR goes down (rich), and when i apply a bit of throttle AFR spikes too much lean as possible.
Earlier someone mentioned this could be due to going too rich. Is there any way to know what is happening? (what should I look for in the logs).

I see a lot of things happening with PW and PW2 and VE1/2. Only thing I know is that PW is probably injector pulse width,
It goes from 5 to 6 to 9 to 12. Seems like a lot. Would that cause the lean reading, while it is actually rich?

Thanks.



Another weird situation below.
From Idle, I apply throttle (2.8%) and RPM goes up from 800 to 1100, and then steep down to 400. At the same time, AFR goes to max lean.

Alexj84nl is offline  
Old 12-01-2022, 10:54 AM
  #12  
Retired Mech Design Engr
iTrader: (3)
 
DNMakinson's Avatar
 
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,009
Total Cats: 857
Default

Have you read the manual (available on msextra). Look for TunerStudio Lite. That will let you know a lot. EDIT: this can be more easily found in TS / Help / TS Help.
Forget "PW". Use "PW2". That is your injector pulse width
VE2 has no meaning.
Tune calls for Open Loop (warmup) curve for PWM Duty Cycle (idle control valve), which shows 2% at 80C. Log shows 29%? Do you know how that is? I'm confused.
Use Idle Advance Correction
I'm not sure I believe your AFR, where is your sensor located? Do you have a full exhaust?
In MLViewer, I like to put Field Min Max (under Calculated Fields) for the displayed variable, as it makes for better perspective when reading the traces.

DNM
DNMakinson is offline  
Old 12-01-2022, 01:05 PM
  #13  
Junior Member
 
redursidae's Avatar
 
Join Date: Aug 2019
Posts: 167
Total Cats: 49
Default

I would suggest you turn off acceleration enrichment and overrun fuel cut for now and focus on your VE table and idle timing until the car idles and returns to idle properly. In your tune I see the idle valve PW value of 2 at full warm, but in your log it is 29. Did you make changes in between or enable closed-loop? If you're using closed-loop, I would switch to Basic mode and play with the gain. Use little gain to start off.

To stabilize the idle, start by having your idle RPM as a cell to reduce interpolation, I would also start off by making all the cells from 30kPa down, and up to 1100RPM 50VE. For timing, enable idle timing corrections, and make the values in your timing table the same around the idle area.

I've attached some images of the settings above and msqparts you can use to pull in these changes in TunerStudio.

I hope this helps.
Attached Files
File Type: zip
msparts.zip (4.9 KB, 2 views)
File Type: zip
ChangePictures.zip (608.3 KB, 2 views)
redursidae is offline  
Old 12-01-2022, 01:20 PM
  #14  
Elite Member
iTrader: (10)
 
Reverant's Avatar
 
Join Date: Jun 2006
Location: Athens, Greece
Posts: 5,979
Total Cats: 356
Default

Sequential is disabled in your tune. Go ahead and set it to semi-sequential and see if you need to revisit any accel related settings. I see you have also changed AE to accel pump instead of time-based, I have no experience with accel pump unfortunately. Once I switched your tune back to time-based AE, I noticed that your added % numbers are huge, like at 100% tpsDot, 50% adder. I have 6% adder at 125 tpsdot, for reference.
Reverant is offline  
Old 12-01-2022, 02:54 PM
  #15  
Newb
Thread Starter
 
Alexj84nl's Avatar
 
Join Date: Nov 2021
Posts: 19
Total Cats: 1
Default

I really appreciate all the feedback, great!
Will dive into all of it, and report back. Thanks alot!
Alexj84nl is offline  
Old 12-02-2022, 07:55 AM
  #16  
Newb
Thread Starter
 
Alexj84nl's Avatar
 
Join Date: Nov 2021
Posts: 19
Total Cats: 1
Default

Originally Posted by DNMakinson
Tune calls for Open Loop (warmup) curve for PWM Duty Cycle (idle control valve), which shows 2% at 80C. Log shows 29%? Do you know how that is? I'm confused.
Yeah, I made a mistake. I entered my idle valve min/max values in closed loop idle settings, but when merging my tune with the one from diyautotune, I forgot to fix that table. (valve is min/max at 29/96 (after testing), going below 25 will invert, and open up again)

Originally Posted by DNMakinson
I'm not sure I believe your AFR, where is your sensor located? Do you have a full exhaust?
My MTX-L is 10 years old, but I installed it with a new (original / expensive) sensor.
It is placed in the original OEM place in the exhaust, and wired through original harness.
The readings are just a bit off (gauge vs tunerstudio), not enough to worry about for now.
drove around for 60 minutes with the wideband gauge and oem ECU, the values seemed trustworthy, so I decided to go for the MS install.

Currently changing the tune based on your feedback, and that from redursidae and Reverant.
Makes a whole lot of sense, thanks again.

Alexj84nl is offline  
Old 12-04-2022, 07:40 AM
  #17  
Newb
Thread Starter
 
Alexj84nl's Avatar
 
Join Date: Nov 2021
Posts: 19
Total Cats: 1
Default

So I followed the provided instructions (thanks again) but it could not fix my problem.
Only thing left to try is semi-sequential ignition, but I don't think that is the cure for the problem I'm having.

What I did:
- idle warmup duty steps table adjusted to match my valve
- openloop on (AFR/EGO control - no correction & idle control - open loop (warmup) )
- overrun fuelcut off
- idle advance on & idle RPM timing correction on
- Idle cells for ign and ve are all equal (to provide smooth idle)

The startup/warmup and idle is OK for now, but when pushing down the throttle peddal from idle, makes the car stall or almost stall.
I also have this behavior during driving. (so tps at 2%, and going to 4%, makes the AFR goes lean for seconds before returning to normal/stoich).

Again, a screenshot of this behavior.
Here you see me pressing throttle from idle.
TPS goes up, all other variables stay equal, except PW2. What is causing PW2 to increase? That is not Accel Enrichment, is it?
My feeling is, that little bump needs to increase further.
What you can see is RPM goes down (until 330), AFR goes up.




I experimented with the Accel Enrichment by using time based / pump based and TPS vs MAP.
Also, I started with the reference values from various basemaps and advice from others, and also tried to lower / higher those values to see what happened.
I could feel/see a slight difference, but in the end it just doesn't seem to make a huge difference. And what I need feels like a huge difference/change is needed.

The feeling I got was that the best result I got was pump-based, 100% MAP. (which is contrary to my expectations, I would expect the VTPS would be more accurate).

Here a screenshot from an actual stall from idle


Last logfile from driving around and applying throttle (from 2.4% to 8.3%).
AFR peaks at max lean for almost 3 seconds.

Attached Files
File Type: msq
CurrentTune.msq (261.8 KB, 14 views)
File Type: mlg
2022-12-03_15.32.59.mlg (1.32 MB, 19 views)
File Type: mlg
2022-12-03_15.02.24- mapdot.mlg (488.6 KB, 16 views)
Alexj84nl is offline  
Old 12-04-2022, 10:48 AM
  #18  
Elite Member
iTrader: (10)
 
Reverant's Avatar
 
Join Date: Jun 2006
Location: Athens, Greece
Posts: 5,979
Total Cats: 356
Default

Try adding 10 or 15 units to your entire VE table and see if it makes it better. A small TPS increase should not make the car stall even with 0 accel enrich.
Reverant is offline  
Old 12-06-2022, 01:22 AM
  #19  
Junior Member
 
redursidae's Avatar
 
Join Date: Aug 2019
Posts: 167
Total Cats: 49
Default

That's definitely and odd behavior, and it's almost like you have a fuel pressure problem, or something else wrong with fuel. You say your vacuum is from the fuel pressure regulator? What if you grab your vacuum from some other place and make sure the fuel pressure regulator has its own line?
redursidae is offline  
Old 12-06-2022, 09:52 AM
  #20  
Junior Member
 
Panici's Avatar
 
Join Date: Aug 2019
Location: Canada
Posts: 145
Total Cats: 10
Default

Originally Posted by Reverant
Try adding 10 or 15 units to your entire VE table and see if it makes it better. A small TPS increase should not make the car stall even with 0 accel enrich.
Agreed, it looks like you need to try more fuel overall. Better to fatten things up, you can lean it back out once you've got drivability sorted.

FWIW when starting on a fresh street tune, I actually turn AE off completely. I am still able to drive around as long as I'm smooth with the throttle inputs.


What I did notice from that log: As your RPM drops below idle, the MAP reading goes way up (obviously). I would start by adding fuel in the upper left side of your VE table.
That is to say, RPM columns at and below idle with MAP readings from ~25kpa to 100kpa.
Panici is offline  


Quick Reply: '92 ms3 - extreme accel enrichment needed



All times are GMT -4. The time now is 10:52 AM.