Lean spike at low RPM with partial throttle
03-07-2016, 10:30 PM
#1
Newb
Thread Starter
Join Date: Feb 2016
Posts: 8
Total Cats: 1
Lean spike at low RPM with partial throttle
Having a couple issues with my megasquirt tune and I'm a bit out of my depth. For reference this is in a MSM with FM full intake, RX-8 injectors that were reconditioned and tested, and an MS3 Basic from Rev. The megasquirt was installed and tuned by a local dyno shop. Haven't driven the car much since getting it back due to the following issues, but it runs great otherwise. So much better than on the stock ECU. This is my first experience with tuning so I'm sure there are dumb questions in here.
First issue I'm having is a lean spike when accelerating slowly with constant, partial throttle. The tuner was convinced it was an acceleration enrichment issue, but it's happening even if the tpsdot is 0. I'll be at 10% or so throttle and as the RPMS climb to 1900-2300 the AFR will got to 18, and then back down to normal. It's lean enough that I actually lose power and have to give it more throttle to get past it. I drove around and ran VEAL in this area for a while and now my VE tables have a big ridge along the 1800rpm line up to about 120kpa. After adding all that fuel and manually smoothing out the area around the ridge so it doesn't stutter, the lean spike is narrower and more like 16.5-17 (still more lean than the AFR target) and I'm also starting to run a bit rich around the spike. I'm not sure if I should worry about fouling plugs if I keep adding fuel in the area, and I have no idea why it goes lean all of a sudden right there. The issue is totally consistent and I can reproduce it every time.
The lean spot was super frustrating when it was icy out as that's where most of my daily driving was (so I didn't actually drive it.) Now that it's warm out and traction isn't an issue, I'm generally at higher RPMs and it's not as noticeable.
I've attached my initial tune from the dyno shop, my current tune, a recent log (sorry it's so big), screenshots of my VE, Spark, and AFR target tables, and screenshots from megalogviewer of a couple examples.
My second issue: it seems like maybe the coolant temp sensor isn't calibrated right. When it's cold out, the coolant temp reads about 20-30 degrees F warmer than my intake temp sensor and the ambient temp, even if it's been sitting for a couple days. Biggest difference is when it's cold—near or below freezing. It's made it so I need to bring a laptop with me to the car and adjust cranking pulse widths if the overnight temp has changed too much. Doesn't seem to affect anything once the car is running though. Also hasn’t been much of an issue now that the temps are above 50 or so.
Is there a good way to check how the calibration is set? DIYautotune has values for the stock sensor, but I'm not sure where to see the current calibration settings. I can open the calibration window but everything is blank. And if I do put in new calibration settings, what all is it going to **** with? I assume mostly startup related stuff since it seems like the top end of the temp curve is accurate.
Original tune: https://www.dropbox.com/s/8r4r95kch8...e%201.msq?dl=0
Current tune after tweaking: https://www.dropbox.com/s/raci46mxpw...33.53.msq?dl=0
Recent log, go to record 3532 for an example: https://www.dropbox.com/s/7so7t3skjg...40.42.msl?dl=0
First issue I'm having is a lean spike when accelerating slowly with constant, partial throttle. The tuner was convinced it was an acceleration enrichment issue, but it's happening even if the tpsdot is 0. I'll be at 10% or so throttle and as the RPMS climb to 1900-2300 the AFR will got to 18, and then back down to normal. It's lean enough that I actually lose power and have to give it more throttle to get past it. I drove around and ran VEAL in this area for a while and now my VE tables have a big ridge along the 1800rpm line up to about 120kpa. After adding all that fuel and manually smoothing out the area around the ridge so it doesn't stutter, the lean spike is narrower and more like 16.5-17 (still more lean than the AFR target) and I'm also starting to run a bit rich around the spike. I'm not sure if I should worry about fouling plugs if I keep adding fuel in the area, and I have no idea why it goes lean all of a sudden right there. The issue is totally consistent and I can reproduce it every time.
The lean spot was super frustrating when it was icy out as that's where most of my daily driving was (so I didn't actually drive it.) Now that it's warm out and traction isn't an issue, I'm generally at higher RPMs and it's not as noticeable.
I've attached my initial tune from the dyno shop, my current tune, a recent log (sorry it's so big), screenshots of my VE, Spark, and AFR target tables, and screenshots from megalogviewer of a couple examples.
My second issue: it seems like maybe the coolant temp sensor isn't calibrated right. When it's cold out, the coolant temp reads about 20-30 degrees F warmer than my intake temp sensor and the ambient temp, even if it's been sitting for a couple days. Biggest difference is when it's cold—near or below freezing. It's made it so I need to bring a laptop with me to the car and adjust cranking pulse widths if the overnight temp has changed too much. Doesn't seem to affect anything once the car is running though. Also hasn’t been much of an issue now that the temps are above 50 or so.
Is there a good way to check how the calibration is set? DIYautotune has values for the stock sensor, but I'm not sure where to see the current calibration settings. I can open the calibration window but everything is blank. And if I do put in new calibration settings, what all is it going to **** with? I assume mostly startup related stuff since it seems like the top end of the temp curve is accurate.
Original tune: https://www.dropbox.com/s/8r4r95kch8...e%201.msq?dl=0
Current tune after tweaking: https://www.dropbox.com/s/raci46mxpw...33.53.msq?dl=0
Recent log, go to record 3532 for an example: https://www.dropbox.com/s/7so7t3skjg...40.42.msl?dl=0
Reply
1
1
03-25-2016, 10:49 AM
#2
Newb
Thread Starter
Join Date: Feb 2016
Posts: 8
Total Cats: 1
I got ahold of my tuner and he's suggested removing and adding timing in the area to see what works. He said that'd be the next thing he did anyway.
From my very limited knowledge and poking around for other ignition tables, mine looks close to other MSM tables and the datalog seems smooth during the spike. To start he suggested taking a few degrees away and then smoothing the area. Does that make sense?
From my very limited knowledge and poking around for other ignition tables, mine looks close to other MSM tables and the datalog seems smooth during the spike. To start he suggested taking a few degrees away and then smoothing the area. Does that make sense?
Reply
0
0
03-25-2016, 10:59 AM
#3
your tuner sounds kinda dumb to suggest AE/EAE when the issue is at steady state
I think his idea to take away timing is also stupid, since your table looks fine, and none of us ever had to dip timing in a specific spot like that.
assuming you don't have a hardware failure or intermittent connection of some sort..........
.........are the injector dead times and voltage comps set up correctly? kinda sorta sounds like an injector issue
I think his idea to take away timing is also stupid, since your table looks fine, and none of us ever had to dip timing in a specific spot like that.
assuming you don't have a hardware failure or intermittent connection of some sort..........
.........are the injector dead times and voltage comps set up correctly? kinda sorta sounds like an injector issue
Reply
0
0
03-25-2016, 11:30 AM
#4
Newb
Thread Starter
Join Date: Feb 2016
Posts: 8
Total Cats: 1
Not sure, here's my dead time table:
I guess it seems weird to me that it'd be some kind of hardware failure since it's so consistent and the car doesn't seem to have any unexpected issues elsewhere (only other issue besides already mention is I think the WUE needs tweaking, but that seems like a normal tuning thing and not a hardware issue.) Buuut, I'm a total newb so that's all just guessing.
I guess it seems weird to me that it'd be some kind of hardware failure since it's so consistent and the car doesn't seem to have any unexpected issues elsewhere (only other issue besides already mention is I think the WUE needs tweaking, but that seems like a normal tuning thing and not a hardware issue.) Buuut, I'm a total newb so that's all just guessing.
Reply
0
0
03-26-2016, 11:40 AM
#5
I'm having a similar problem with my MS after I turned on EAE. It'll buck pretty bad right around 2k, and although I'm tuning it out, that's always the problem spot. Even with the stock ecu there was a bit of a dead spot there. It looks like your AE could use some work since the AFR is going lean whenever map goes up and vice versa. I know that AE is for transient conditions, but EAE is always working and I sometimes get some strange needle movement on my EAE correction gauge. (Take advice from a guy with a poorly tuned car wisely)
What AE are you using?
What AE are you using?
Reply
0
0
03-26-2016, 12:53 PM
#6
Newb
Thread Starter
Join Date: Feb 2016
Posts: 8
Total Cats: 1
In my screenshots I was intentionally making the lean thing happen so I could try and figure out what was going on, so it may give a bit of an inaccurate picture of what's happening in transients. I'm sure my AE could use some work, but it actually does seem to be setup decently well. In other parts of the log when I go from, say, 20% throttle to 100%, AE gives it a good squirt of fuel and the AFR's track with the target table.
It's using time based AE right now. From what I understand pump based AE should be better, but I've had trouble tracking down much info about it other than that it mimics the accel pump on a carb. Not knowing a ton about tuning I dunno what makes that better. Haven't tried EAE as it sounds like a bitch to tune even for people who know what they're doing really well. Which I don't, and apparently maybe my tuner doesn't either.
It's using time based AE right now. From what I understand pump based AE should be better, but I've had trouble tracking down much info about it other than that it mimics the accel pump on a carb. Not knowing a ton about tuning I dunno what makes that better. Haven't tried EAE as it sounds like a bitch to tune even for people who know what they're doing really well. Which I don't, and apparently maybe my tuner doesn't either.
Reply
0
0
03-26-2016, 09:48 PM
#7
Newb
Join Date: Dec 2011
Posts: 29
Total Cats: 6
I can't see anything obvious.
A shot in the dark might be to try explicitly disabling the VTCS and see if it changes the behaviour. Your configuration has it configured to be inactive at CLT >= 122F or RPM >= 2300 so it _shouldn't_ be in effect but ... like I said a shot in the dark.
Advanced Engine -> Programmable On/Off Outputs -> PA7 - Inj H and turn off the Enable checkbox. Note: this may have a bit of an effect on cold starts as the tumbling affects AFR. Turn it back on if it doesn't change anything - I keep mine off but since you have tuned already with it on you'll probably want it left where it was.
Other bits:
I'm using the yellow top RX8 injectors too. I find that at PW < about 1.6 I get lean conditions too. However this is not the cause of the lean condition in the area you highlighted. When tuning, I use a little bit of injector small pulsewidths to prevent the full lean conditions on low MAPs (which is where low PW occurs) so EGO doesn't over compensate. Once I have things tuned I turn it back off again because at low MAPs full lean is not an issue since you are decelerating and going full lean might even provide better fuel economy.
One small suggestion: you can set Table Choices-> Fuel Table Switching and Spark Table Switching to off so you only see the one table (VE1 and Ignition Table 1) that is being used. This has nothing to do with your problem, it just cleans up the menus a little.
A shot in the dark might be to try explicitly disabling the VTCS and see if it changes the behaviour. Your configuration has it configured to be inactive at CLT >= 122F or RPM >= 2300 so it _shouldn't_ be in effect but ... like I said a shot in the dark.
Advanced Engine -> Programmable On/Off Outputs -> PA7 - Inj H and turn off the Enable checkbox. Note: this may have a bit of an effect on cold starts as the tumbling affects AFR. Turn it back on if it doesn't change anything - I keep mine off but since you have tuned already with it on you'll probably want it left where it was.
Other bits:
I'm using the yellow top RX8 injectors too. I find that at PW < about 1.6 I get lean conditions too. However this is not the cause of the lean condition in the area you highlighted. When tuning, I use a little bit of injector small pulsewidths to prevent the full lean conditions on low MAPs (which is where low PW occurs) so EGO doesn't over compensate. Once I have things tuned I turn it back off again because at low MAPs full lean is not an issue since you are decelerating and going full lean might even provide better fuel economy.
One small suggestion: you can set Table Choices-> Fuel Table Switching and Spark Table Switching to off so you only see the one table (VE1 and Ignition Table 1) that is being used. This has nothing to do with your problem, it just cleans up the menus a little.
Reply
0
0
03-27-2016, 12:11 AM
#9
Newb
Join Date: Dec 2011
Posts: 29
Total Cats: 6
Just noticed that you have VVT turned on (Boost VVT->VVT Settings : No. of VVTs should be Off) - we MSM folk don't have VVT.
I have no idea if this could be causing the symptoms you are seeing : maybe the VVT output is tied to something else or influences something else? - The VVT table comes into play around 1600RPM. I suspect this is just another shot in the dark.
I'm also assuming you have a EBC? You have boost control turned on too.
edited.
I have no idea if this could be causing the symptoms you are seeing : maybe the VVT output is tied to something else or influences something else? - The VVT table comes into play around 1600RPM. I suspect this is just another shot in the dark.
I'm also assuming you have a EBC? You have boost control turned on too.
edited.
Last edited by VcrMiata; 03-27-2016 at 12:22 AM.
Reply
0
0
03-27-2016, 03:23 PM
#10
Newb
Thread Starter
Join Date: Feb 2016
Posts: 8
Total Cats: 1
Shots in the dark are totally welcome. I feel like that's the most likely way I'm gonna figure it out.
I have a manual boost controller, so I don't know why that'd be turned on either.
Might not have time to try anything till later this week but really appreciate all the suggestions. If nothing else at least I'm cleaning up dumb settings in the tune.
I have a manual boost controller, so I don't know why that'd be turned on either.
Might not have time to try anything till later this week but really appreciate all the suggestions. If nothing else at least I'm cleaning up dumb settings in the tune.
Reply
0
0
03-28-2016, 12:17 PM
#11
Newb
Join Date: Dec 2011
Posts: 29
Total Cats: 6
Missed the bit about your issues driving when icy and I think I know what is causing your issue there- Over-Run Fuel Cut.
Your settings are configured to pull fuel when MAP is less than 50Kpa - that is way too high. On light thottle (i.e. feathering around idle) you will get bucking like crazy with that MAP setting. This would make for very adventurous driving in icy conditions.
On my drive, I use the following:
RPM > than: 1800
MAP < than: 20 (only active when decelerating heavily in gear - needs to be safely under idle Kpa)
TPS < than: 2
CLT > than(C): 80
After Delay: .5 (just noticed this on my tune, should be higher to prevent fuel cut when shifting - just increased mine to 2.5 sec)
Progressive Fuel Cut: On
Progressive Fues Return: ON
Return Transition Time:.5
Lowest RPM to re-engage: 1600
RPMdot for Upper RPM: 1000
RPMdot for Lower RPM: 500
Duration of Fuel adders: 0
Delay EGO after fuel return: 2
Your settings are configured to pull fuel when MAP is less than 50Kpa - that is way too high. On light thottle (i.e. feathering around idle) you will get bucking like crazy with that MAP setting. This would make for very adventurous driving in icy conditions.
On my drive, I use the following:
RPM > than: 1800
MAP < than: 20 (only active when decelerating heavily in gear - needs to be safely under idle Kpa)
TPS < than: 2
CLT > than(C): 80
After Delay: .5 (just noticed this on my tune, should be higher to prevent fuel cut when shifting - just increased mine to 2.5 sec)
Progressive Fuel Cut: On
Progressive Fues Return: ON
Return Transition Time:.5
Lowest RPM to re-engage: 1600
RPMdot for Upper RPM: 1000
RPMdot for Lower RPM: 500
Duration of Fuel adders: 0
Delay EGO after fuel return: 2
Last edited by VcrMiata; 03-29-2016 at 11:23 AM.
Reply
0
0
04-02-2016, 09:38 PM
#12
Newb
Thread Starter
Join Date: Feb 2016
Posts: 8
Total Cats: 1
Allright, tried all the above and it wound up making everything suuuper rich, but still with a lean (er) spike right at the same RPM range. I rolled everything back to original and will try making the changes one by one (like I should have in the first place) but I'm not hopeful since the same general curve was there, just richer.
Small upside, got the right temp sensor calibration stuff from rev, so at least I'll be able to get cold start sorted.
Small upside, got the right temp sensor calibration stuff from rev, so at least I'll be able to get cold start sorted.
Reply
0
0
04-02-2016, 10:49 PM
#13
Newb
Join Date: Dec 2011
Posts: 29
Total Cats: 6
The only thing that might have made things "super rich" would be turning off VTCS but only while the engine is cold - your tune disables VTCS when RPM >= 2300 or when CLT >= 122F.
[That's why I mentioned to re-enable it after testing - otherwise you would have to rework your ASE and WUE curves to prevent going rich when cold.]
Hopefully restoring your base map and taking it one step at a time will help you find the spike problem.
[That's why I mentioned to re-enable it after testing - otherwise you would have to rework your ASE and WUE curves to prevent going rich when cold.]
Hopefully restoring your base map and taking it one step at a time will help you find the spike problem.
Reply
0
0
04-03-2016, 12:26 AM
#14
Your issue sounds scary familiar. I have a 2003 G35 coupe with an HKS twin turbo and a Haltech ECU. When I first installed the ECU/injectors/fuel system I had a lean spike at 2400 RPM that I couldn't tune out no matter what I did with the fuel map.
The OEM fuel system on the G35/350z is returnless, and most people (myself included) add a fuel return line back to the tank so they can run big fuel pumps (340 lph). The most common way to do this was to remove the OEM fuel rail damper from the back of the drivers side fuel rail and install an adapter fitting in its place that connected to a AN line that runs back to the fuel tank. A company called CJ motorsports made such an adapter and was the most popular and widely used (they told me they sold over 500).
After pulling my hair out and trying everything I could think of (tuning, more tuning, swapping pumps, swapping injectors) I showed the datalogs to a colleague at work and he suggested that it looked like resonance. That's when it clicked and I realized that my problems were likely caused by the removal of the fuel damper. So I redesigned the adapter to add a mount for the OE damper and it dramatically reduced the lean spike. In fact I added a second damper to the passeneger side fuel rail and saw further improvements. Nissan even added a second damper themselves on the 2007+ 350Z.
After I proved (very publicly, on the my350z forums) that the adapter and lack of damper had caused all my problems, the owner of CJ motorsports admitted that his own personal car had the same issue and he had never been able to figure it out. As a result CJ motorsports redesigned their whole product line to retain OEM Nissan fuel rail dampers.
If you want to read the whole saga, here is a link to the thread:
Lean Spot At ~2400 RPM...What the? - MY350Z.COM Forums
The OEM fuel system on the G35/350z is returnless, and most people (myself included) add a fuel return line back to the tank so they can run big fuel pumps (340 lph). The most common way to do this was to remove the OEM fuel rail damper from the back of the drivers side fuel rail and install an adapter fitting in its place that connected to a AN line that runs back to the fuel tank. A company called CJ motorsports made such an adapter and was the most popular and widely used (they told me they sold over 500).
After pulling my hair out and trying everything I could think of (tuning, more tuning, swapping pumps, swapping injectors) I showed the datalogs to a colleague at work and he suggested that it looked like resonance. That's when it clicked and I realized that my problems were likely caused by the removal of the fuel damper. So I redesigned the adapter to add a mount for the OE damper and it dramatically reduced the lean spike. In fact I added a second damper to the passeneger side fuel rail and saw further improvements. Nissan even added a second damper themselves on the 2007+ 350Z.
After I proved (very publicly, on the my350z forums) that the adapter and lack of damper had caused all my problems, the owner of CJ motorsports admitted that his own personal car had the same issue and he had never been able to figure it out. As a result CJ motorsports redesigned their whole product line to retain OEM Nissan fuel rail dampers.
If you want to read the whole saga, here is a link to the thread:
Lean Spot At ~2400 RPM...What the? - MY350Z.COM Forums
Reply
1
1
04-03-2016, 12:37 AM
#15
All that being said, my Miata is a 94 with a Walbro 190HP, FIC 750cc injectors, stock fuel rails modified for dual feed, stock rail-mounted FPR, and MS3 and I have never had an issue like I did with my G35. I am not very familiar with the Mazdaspeed - is the fuel system returnless? If so, is there a damper on the rail? If not, are you using the OE rail-mounted FPR? A FPR acts like a damper.
Reply
0
0
04-22-2016, 07:42 PM
#16
Newb
Thread Starter
Join Date: Feb 2016
Posts: 8
Total Cats: 1
Work has been crazy lately so I haven't had any time to do more testing since my last work, and I missed your post mx592.
I don't know that much about the MSM's fuel system in particular. I believe it's actually the same as the stock NB from those years. The only change I made was having the yellowtop injectors installed. I don't know if the issue you're describing is possible/likely in the miata fuel system, but the symptoms definitely sound similar. I wonder if there's a way to install something on the fuel rail wrong that'd cause it maybe? Interesting story regardless.
I don't know that much about the MSM's fuel system in particular. I believe it's actually the same as the stock NB from those years. The only change I made was having the yellowtop injectors installed. I don't know if the issue you're describing is possible/likely in the miata fuel system, but the symptoms definitely sound similar. I wonder if there's a way to install something on the fuel rail wrong that'd cause it maybe? Interesting story regardless.
Reply
0
0
Thread
Thread Starter
Forum
Replies
Last Post
thumpetto007
Engine Performance
95
12-24-2017 10:55 PM
