MEGAsquirt A place to collectively sort out this megasquirt gizmo

Car dies immediately on cold start, but then starts up totally fine

Thread Tools
 
Search this Thread
 
Old 03-23-2019, 11:52 AM
  #1  
Newb
Thread Starter
 
Dwalk51's Avatar
 
Join Date: Aug 2016
Location: Atlanta, Georgia
Posts: 43
Total Cats: 5
Default Car dies immediately on cold start, but then starts up totally fine

Hi all,

I'm dealing with an issue I can't quite figure out. I'll lay out what's happening and then I'll describe what I've tried to fix it so far.

I've got a 1994 Miata with an MSPNP 3 from the MKTurbo Kit (shout out to Lars!). From a cold start, (between 35F and 65ishF), the car dies immediately upon start up. It catches, revs to about 2000rpm, and then promptly falls and dies. Using the throttle doesn't seem to help or affect it, so I can't 'catch' the falling revs and keep it running. As soon as it's dead though I can crank it back up and it revs to 2000rpm and goes through a normal start up no problem (WUE, ASE, all of that). I don't get any sync errors or check engine lights on either start up. I mention the temperature at which this happens because it doesn't happen at warmer 'cold-start' temperatures. We've had a few days in the high 60s, low 70s here in Atlanta and it doesn't exhibit the same symptoms.

Because it doesn't exhibit the same symptoms at warmer temperatures, I'm assuming I've goofed some table in my settings related to temperature and fueling/air. I've tried adjusting fuel up and down, initial idle valve % up and down, priming pulse, and cranking pulse, but I just can't seem to nail this thing down. I've even got hot restarts pretty well sorted (long ASE taper between 130F and 180F and an early EGO), but I can't figure out cold initial start.

If anyone has any suggestions, I'd be immensely grateful. I've included a log and my tune as attachments to this post.

Thanks!
Dylan
Attached Files
File Type: msq
94 Turbo Miata dwalk51 Tune.msq (242.1 KB, 150 views)
File Type: msl
ColdStart032219.msl (173.6 KB, 85 views)
Dwalk51 is offline  
Old 03-23-2019, 12:18 PM
  #2  
Junior Member
 
Mudflap's Avatar
 
Join Date: May 2017
Location: Louisville, Co
Posts: 466
Total Cats: 86
Default

Move to Florida?
Mudflap is offline  
Old 03-23-2019, 01:24 PM
  #3  
Junior Member
 
Mudflap's Avatar
 
Join Date: May 2017
Location: Louisville, Co
Posts: 466
Total Cats: 86
Default

I think I figured it out. Go plot Spark Table1 and watch the behavior as you drop off in RPM. What is happening is that you are too rapidly losing power. Basically losing momentum and dying. It goes to 35deg timing, then drops rapidly to that MAP 30, RPM 1500 spark cell at 18deg. Too dramatic.

Then if you look at the graph over where you are actually running, it starts up at 33deg timing. But then only drops a little to 28deg and maintains.

So I believe you can fix this by bringing up your timing cells. I'm more at 23-25deg at 1500 RPM. Then at 20deg at idle.
Mudflap is offline  
Old 03-23-2019, 01:54 PM
  #4  
Senior Member
 
SpartanSV's Avatar
 
Join Date: Apr 2017
Location: Greeley, CO
Posts: 1,220
Total Cats: 162
Default

Your ASE settings are nonsensical. Set reasonable curves first. Once that is fixed I'd start adding to WUE.

I primarily run e85, but my car likes ~11 AFR after a ~30 degree cold start. That ratio allows me to start the car and immediately start driving. Anything leaner gives driveability issues during warmup.
SpartanSV is offline  
Old 03-23-2019, 03:17 PM
  #5  
Retired Mech Design Engr
iTrader: (3)
 
DNMakinson's Avatar
 
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,009
Total Cats: 856
Default

Originally Posted by Mudflap
I think I figured it out. Go plot Spark Table1 and watch the behavior as you drop off in RPM. What is happening is that you are too rapidly losing power. Basically losing momentum and dying. It goes to 35deg timing, then drops rapidly to that MAP 30, RPM 1500 spark cell at 18deg. Too dramatic.

Then if you look at the graph over where you are actually running, it starts up at 33deg timing. But then only drops a little to 28deg and maintains.

So I believe you can fix this by bringing up your timing cells. I'm more at 23-25deg at 1500 RPM. Then at 20deg at idle.
Wow, I thought his Advance was pretty radical below 2000 RPM.

Re: SpartanSV comment:
Yes, AFR may be a bit lean right after start. It is amazing how much that 2nd start can deceive you into thinking you are closer than you are. I would also try adding a bit more ASE.
DNMakinson is offline  
Old 03-24-2019, 08:17 PM
  #6  
Newb
 
SA58Junkie's Avatar
 
Join Date: Oct 2017
Posts: 37
Total Cats: 1
Default

So 67% duty cycle is all the way open and 26% is all the way closed on your valve? At least that is what your settings say. How does it run without PID control? BTW, your spark table literally has almost no resolution in boost. I'd probably fix that why you're at it.

Edit: I just noticed you are in ATL. What part? I live in Powder Springs and I have a MKturbo VVT NB2 miata.
SA58Junkie is offline  
Old 03-29-2019, 08:55 AM
  #7  
Newb
Thread Starter
 
Dwalk51's Avatar
 
Join Date: Aug 2016
Location: Atlanta, Georgia
Posts: 43
Total Cats: 5
Talking Working again!

Sorry for the slow response guys, I really appreciate all the help. Work was hellish this week and I was slammed, but I've been reading the comments.

Originally Posted by SpartanSV
Your ASE settings are nonsensical. Set reasonable curves first. Once that is fixed I'd start adding to WUE.

I primarily run e85, but my car likes ~11 AFR after a ~30 degree cold start. That ratio allows me to start the car and immediately start driving. Anything leaner gives driveability issues during warmup.
Nonsensical how? Too much or too little in a specific area?

Originally Posted by DNMakinson
Wow, I thought his Advance was pretty radical below 2000 RPM.

Re: SpartanSV comment:
Yes, AFR may be a bit lean right after start. It is amazing how much that 2nd start can deceive you into thinking you are closer than you are. I would also try adding a bit more ASE.
Ok yes! I did try adding more ASE, with good results. I made some adjustments to the tune, I'll post the results below. It's much better now!

Originally Posted by SA58Junkie
So 67% duty cycle is all the way open and 26% is all the way closed on your valve? At least that is what your settings say. How does it run without PID control? BTW, your spark table literally has almost no resolution in boost. I'd probably fix that why you're at it.

Edit: I just noticed you are in ATL. What part? I live in Powder Springs and I have a MKturbo VVT NB2 miata.
Yep, that's about what seemed right. I've checked it a few times via "Output Test Mode - Idle Valve" when it's warmed up. When I mess with the % duties there, I stop getting a response basically around 27%, and on the high end I stop getting a response above 67%. I thought that was how you found and set those values.

I'm in Midtown, and no way, that's great. I've been loving the MKTurbo kit. I just don't get enough time to drive it. I'd be down to find a weekend to go driving together if you're into it.

Originally Posted by Mudflap
I think I figured it out. Go plot Spark Table1 and watch the behavior as you drop off in RPM. What is happening is that you are too rapidly losing power. Basically losing momentum and dying. It goes to 35deg timing, then drops rapidly to that MAP 30, RPM 1500 spark cell at 18deg. Too dramatic.

Then if you look at the graph over where you are actually running, it starts up at 33deg timing. But then only drops a little to 28deg and maintains.

So I believe you can fix this by bringing up your timing cells. I'm more at 23-25deg at 1500 RPM. Then at 20deg at idle.
Ok yes! Watching the spark map when playing back the log in MegaLog Viewer, it really *snapped* back to 18deg. I upped the timing in those cells and around the idle cells, added ASE as was suggested, and have tried cold-starting it again a few times. Thankfully, it's much better now and hasn't died on first start up since I made the changes. I've included an updated log of a cold start up and my current tune as attachments. Cold start up still isn't quite as predictable and 'clean' as stock, but I'll continue to tweak it. I'm just super happy to have it not die the first time I start it. Thanks everyone for the feedback! Anyone have tips on how to clean up start up, or where to go to find more? I.e. on stock it feels like it catches after turning over maybe 0.5s and immediately goes from cranking to running. On my MS it cranks for 1.3s (according to log) and then it 'gurgles' from cranking into running over a second, but the overall impression is much sloppier than stock. It's totally functional, just ain't pretty.

Thanks again for all the help and suggestions!
Attached Files
File Type: msl
ColdStart032419.msl (296.6 KB, 66 views)
Dwalk51 is offline  
Old 04-01-2019, 07:13 PM
  #8  
Senior Member
 
SpartanSV's Avatar
 
Join Date: Apr 2017
Location: Greeley, CO
Posts: 1,220
Total Cats: 162
Default

Originally Posted by Dwalk51
Nonsensical how? Too much or too little in a specific area?
If I need to explain what makes these settings nonsensical then you clearly aren't trying hard enough.

Hint: I would reference a known good base map if someone told me one of my settings is incredibly stupid.



SpartanSV is offline  
Old 04-02-2019, 12:01 AM
  #9  
Cpt. Slow
iTrader: (25)
 
curly's Avatar
 
Join Date: Oct 2005
Location: Oregon City, OR
Posts: 14,181
Total Cats: 1,132
Default

That right there is someone desperate to get rid of hot restart issues.
curly is offline  
Old 04-02-2019, 09:48 AM
  #10  
Junior Member
 
thebigtuna's Avatar
 
Join Date: Nov 2016
Location: Columbia, MO
Posts: 160
Total Cats: 12
Default

Megasquirt Support Forum (MSEXTRA) ? ASE and ASE taper. (View topic)
Megasquirt Support Forum (MSEXTRA) ? Need help understanding ASE taper (View topic)

I would highly recommend checking these two threads out. I was pretty confused on when to adjust ase vs ase taper myself. The second one specifically helped with making it all click. Take your time and try to understand what each setting does. Your ASE and ASE Taper are bonkers.
thebigtuna is offline  
Old 04-04-2019, 02:07 PM
  #11  
Newb
Thread Starter
 
Dwalk51's Avatar
 
Join Date: Aug 2016
Location: Atlanta, Georgia
Posts: 43
Total Cats: 5
Default

Thanks for the feedback guys. I've actually read through that first forum thread before, and the second one is new to me. That second thread cleared up a question I had about how many revolutions 1 cycle really is. Conceptually, I get what ASE and ASE taper are doing, and I realize my settings probably look crazy. Curly hit the nail on the head, I've had terrible hot restart issues. And for whatever it's worth, my crazy ASE settings have helped tremendously. I'm not opposed to changing my ASE settings, but if it's stupid and it works, then it ain't stupid, right?

My logic behind my ASE settings is this:
  1. my engine is reading a falsely hot air temperature (IAT) on hot restart when the engine has been off for at least 10 min
  2. Because the engine is reading a false-hot IAT, it's pulling fuel (even though I've flattened the air density compensation curve)
  3. To compensate for my engine pulling fuel on hot restart, I want to dump more fuel until some air has flowed past the IAT sensor and it's reading correctly again
  4. To dump more fuel for just a few second (ok, maybe up to 10 seconds), I'm using a crazy high ASE Taper value that only applies for a range of warm coolant temperatures, so that my high ASE taper is only in effect for hot restarts
  5. I set EGO to kick in shortly after start up (7 seconds, I believe), and a relatively low CLT minimum, so that EGO can work on hot restarts
    1. I've got EGO set with a relatively high authority, but I'm crippling how quickly it can act so that it doesn't make wild swings
Someone described similar settings somewhere else on this site, but I can't remember where I read about them. For whatever it's worth, my hot-restarts are much better now. By the time ASE is tapering off, EGO is kicking in and helps stabilize my idle. It's not quite as good as stock, but it's much much better than before where every time I restarted the car I had to keep the car running with the throttle while I backed out of the parking lot, navigating other cars and pedestrians.

I'll post a log of some hot restarts. You'll see. But also tell me if I'm crazy, because if there's a better way to handle lean hot restarts, I'm dying to know. I've searched the seven seas of Miata forum content without much luck until the ASE/EGO trick I've described here.
Dwalk51 is offline  
Old 04-04-2019, 03:27 PM
  #12  
Retired Mech Design Engr
iTrader: (3)
 
DNMakinson's Avatar
 
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,009
Total Cats: 856
Default

Yes post. Interested to see if your MAT is reading falsely hot or not. Where is your MAT sensor located?
DNMakinson is offline  
Old 04-04-2019, 06:23 PM
  #13  
Junior Member
 
Mudflap's Avatar
 
Join Date: May 2017
Location: Louisville, Co
Posts: 466
Total Cats: 86
Default

so you are saying you have an IAT (or MAT) sensor that is heat soaked due to location? MOVE IT. Your problems with a restart are just the tip of the iceberg if you are letting your intake manifold soak the IAT. You should have thermal isolation by using a non-metallic intake hose to your throttle body. It should live just at the output of your intercooler.
Mudflap is offline  
Related Topics
Thread
Thread Starter
Forum
Replies
Last Post
dwink
MEGAsquirt
16
12-31-2015 04:54 PM
Onyxyth
MEGAsquirt
3
03-31-2014 12:25 PM
Ryan_G
MEGAsquirt
6
11-09-2012 03:45 AM
Tony the Tiger
MSPNP
17
11-03-2012 10:01 PM
Ben
MEGAsquirt
40
12-03-2007 03:35 PM



Quick Reply: Car dies immediately on cold start, but then starts up totally fine



All times are GMT -4. The time now is 06:02 AM.