MS3X Hot Starts Idle like poo poo?
#1
MS3X Hot Starts Idle like poo poo?
So I was reading through several threads on here and on other forums about shitty idle during warm/hot starts. With my current tune, the car runs pretty well on cold starts, but if I shut it off, and start it again later before too long, the car surges and hunts during idle, and goes from lean to rich, back to lean. (16:1 -> 13:1 -> 16:1) If I start driving while its doing this, it works fine. Accel enrichment is working too as it goes much richer under WOT even when warm.
I am thinking that this has to be able to be tuned out with ASE, WUE or something like that but I am not sure. The car idles perfectly fine when cold so I have no clue here guys.
I can post a my current tune so you all can see if there are any glaring issues? I can take a datalog with this tune if need-be tomorrow.
Currently running 420cc RX8 Injectors if that helps out.
I am thinking that this has to be able to be tuned out with ASE, WUE or something like that but I am not sure. The car idles perfectly fine when cold so I have no clue here guys.
I can post a my current tune so you all can see if there are any glaring issues? I can take a datalog with this tune if need-be tomorrow.
Currently running 420cc RX8 Injectors if that helps out.
#2
SadFab CEO
iTrader: (3)
Join Date: Sep 2012
Location: your mom's house phoenix, AZ
Posts: 4,560
Total Cats: 1,142
Cant look at the msq at the moment, but fwiw my hot starts were horrible with RX8 injectors regardless of my tune. Just switching to EV14s improved things immensely. I think the EV6(?)s have a real shitty dead time vs temp curve. Your lean-rich fluctuations tell me you have closed loop idle enabled? and its having a hard time dealing with.... something.
But for starters, where is your IAT sensor located?
But for starters, where is your IAT sensor located?
#3
This post helped me get acceptable hot restarts on FF injectors.
https://www.miataturbo.net/megasquir...estarts-79819/
https://www.miataturbo.net/megasquir...estarts-79819/
#4
Retired Mech Design Engr
iTrader: (3)
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,011
Total Cats: 859
And I can give an update:
Still on yellow tops. Went to a return style, manifold referenced fuel system. That did not help. I am now running a nominal to slightly rich idle VE table and 15% EGO authority. Results are quite acceptable.
For me, the worst re-starts were after a 20 minute shutdown.
Still on yellow tops. Went to a return style, manifold referenced fuel system. That did not help. I am now running a nominal to slightly rich idle VE table and 15% EGO authority. Results are quite acceptable.
For me, the worst re-starts were after a 20 minute shutdown.
#5
Cant look at the msq at the moment, but fwiw my hot starts were horrible with RX8 injectors regardless of my tune. Just switching to EV14s improved things immensely. I think the EV6(?)s have a real shitty dead time vs temp curve. Your lean-rich fluctuations tell me you have closed loop idle enabled? and its having a hard time dealing with.... something.
But for starters, where is your IAT sensor located?
But for starters, where is your IAT sensor located?
#6
And I can give an update:
Still on yellow tops. Went to a return style, manifold referenced fuel system. That did not help. I am now running a nominal to slightly rich idle VE table and 15% EGO authority. Results are quite acceptable.
For me, the worst re-starts were after a 20 minute shutdown.
Still on yellow tops. Went to a return style, manifold referenced fuel system. That did not help. I am now running a nominal to slightly rich idle VE table and 15% EGO authority. Results are quite acceptable.
For me, the worst re-starts were after a 20 minute shutdown.
I can take a log for you all of the issue and post that with my most current tune some time today.
#8
Braineack is correct (he usually is). You're using the default air density table which is too aggressive.
I bet if you posted a log you'd see fuel correction in the ~95% range. That means the air density table is pulling ~5% of your fuel out.
If you flatten the 'hot' section of that table (anything under 100% fuel) it will help immensely. I think I pull 1-2% fuel at 86 degrees - less than half what the default table pulls.
I bet if you posted a log you'd see fuel correction in the ~95% range. That means the air density table is pulling ~5% of your fuel out.
If you flatten the 'hot' section of that table (anything under 100% fuel) it will help immensely. I think I pull 1-2% fuel at 86 degrees - less than half what the default table pulls.
#10
Here is where I have my MAT Air Density Table set to now (see attached)
What should my CLT correction table look like if I need to change it?
#11
Boost Czar
iTrader: (62)
Join Date: May 2005
Location: Chantilly, VA
Posts: 79,688
Total Cats: 4,113
make it 100% ACROSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS SSSSSSSSSSSSSSSSSSSSSSSS
#12
Retired Mech Design Engr
iTrader: (3)
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,011
Total Cats: 859
If you set it to 100% across the board, then whatever temperature (MAT) you tune at, and you should be as consistent as possible, becomes your basis and you keep that temp at 100%, and then correct the rest of the curve later. You don't know what temp should be 100% yet, until you know at what MAT you tune the VE table at. Mine is at 70* because that is a common temp in my area; in the afternoons in winter, or mornings in summer.
After VE is tuned, then on cool mornings or hot afternoons, adjust MAT curve at some stable operating point such as 50kPa and 3000 RPM.
After VE is tuned, then on cool mornings or hot afternoons, adjust MAT curve at some stable operating point such as 50kPa and 3000 RPM.
#13
If you set it to 100% across the board, then whatever temperature (MAT) you tune at, and you should be as consistent as possible, becomes your basis and you keep that temp at 100%, and then correct the rest of the curve later. You don't know what temp should be 100% yet, until you know at what MAT you tune the VE table at. Mine is at 70* because that is a common temp in my area; in the afternoons in winter, or mornings in summer.
After VE is tuned, then on cool mornings or hot afternoons, adjust MAT curve at some stable operating point such as 50kPa and 3000 RPM.
After VE is tuned, then on cool mornings or hot afternoons, adjust MAT curve at some stable operating point such as 50kPa and 3000 RPM.
#15
Retired Mech Design Engr
iTrader: (3)
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,011
Total Cats: 859
I don't know of any practical use of the CLT/MAT table. IIRC, it was an attempt by the FW people to be a means to address hot restart issues that did not work as conceived. At least I have never seen anyone use it for anything. Mine is all zeroes.
#18
Retired Mech Design Engr
iTrader: (3)
Join Date: Jan 2013
Location: Seneca, SC
Posts: 5,011
Total Cats: 859
Yes. It is an attempt to have actual MAT better approximate IAT readings.
However, I have found the same as this guy: MSExtra Discussion "if any of the users that posted previous threads about tuning this curve could provide updates or helpful advice on direction or how to approach tuning this curve it would be great."
ive read all the posts from previous and some of you were in the middle of it, but lots of unfinished posts, so just wondering if some of you had more/better progress to pass on.
There were some earlier discussions on the subject where users complained that they had opposite effects than what jsmcortina had desired.
This is how I think about it. If I have my MAT correction: 100%@70*, and 98%@100* (theoretical). Then I set up a blend of 30%CLT / 70%MAT, and the CLT is 180* and MAT (measured) is 70*. CLT/MAT is going to calculate a resultant "MAT" of .3*180 + .7*70 = 103*. So I will use the 98% when tuning, and will therefore fatten up idle VE. Then, when CLT=70*, and MAT=70*, "MAT" will be 70, and I will use the 100% value and end up richer.
Thus, this correction has the effect of pulling fuel when CLT is high vs when CLT is low.
It is true that at higher intake flow, the CLT/MAT curve would tend to Zero, and the measured MAT will be used, and go back to 100%.
NOTE: code may actually use Rankine, not Fahrenheit, but the effect is similar.
Thus I would tend to agree with most MSExtra posts that CLT/MAT is counterproductive.
Hence I asked if you had seen cases where it was meaningful. Maybe with some sort of reversed curve.
One final note. It would seem that if MAT correction is set to 100% across the full range, then CLT/MAT would have no effect whatsoever.
Thoughts?
However, I have found the same as this guy: MSExtra Discussion "if any of the users that posted previous threads about tuning this curve could provide updates or helpful advice on direction or how to approach tuning this curve it would be great."
ive read all the posts from previous and some of you were in the middle of it, but lots of unfinished posts, so just wondering if some of you had more/better progress to pass on.
There were some earlier discussions on the subject where users complained that they had opposite effects than what jsmcortina had desired.
This is how I think about it. If I have my MAT correction: 100%@70*, and 98%@100* (theoretical). Then I set up a blend of 30%CLT / 70%MAT, and the CLT is 180* and MAT (measured) is 70*. CLT/MAT is going to calculate a resultant "MAT" of .3*180 + .7*70 = 103*. So I will use the 98% when tuning, and will therefore fatten up idle VE. Then, when CLT=70*, and MAT=70*, "MAT" will be 70, and I will use the 100% value and end up richer.
Thus, this correction has the effect of pulling fuel when CLT is high vs when CLT is low.
It is true that at higher intake flow, the CLT/MAT curve would tend to Zero, and the measured MAT will be used, and go back to 100%.
NOTE: code may actually use Rankine, not Fahrenheit, but the effect is similar.
Thus I would tend to agree with most MSExtra posts that CLT/MAT is counterproductive.
Hence I asked if you had seen cases where it was meaningful. Maybe with some sort of reversed curve.
One final note. It would seem that if MAT correction is set to 100% across the full range, then CLT/MAT would have no effect whatsoever.
Thoughts?
Last edited by DNMakinson; 04-17-2017 at 08:56 AM. Reason: Added quote of germane part of MSExtra thread