LSJ Turbo Dynamic issues

[eronq9]

Already had MAF dialed in on a friends LSJT, s256et.
Set timing low (10deg) pre spool WOT to raise egts, then ramped spark with rpms to 20 deg at redline (E85)
Car pulls great in MAF only at WOT, if you WOT around 3300-3400 it will hit max boost (20psi) before 4000rpm.

Anyways got VE decently close in open loop with the MAF disabled, started raining so I had to stop doing any pulls but got cruising dialed in pretty good. Turned dynamic bac on set at 3400 rpm and sometimes, especially lower gears, the car would hesitate likes its hitting the rev limiter. 4th gear you could spool through starting at 3000rpm and it wouldn't do it. Should I try messing with the VE table some more, lower VE switch over, or possibly look for some settings that might have accidentally gotten changed?

MAF only and VE only work great, but with both enabled it doesnt like it.

[AP83]

I have my dynamic set at 2,000 rpm. No problems with spool here.

Car made 418 at 22psi and 16 degrees. E30 mix. No knock. No traction either.

[AP83]

Also, what's the plug gap?

I had issues with the car breaking up under the initial spool, until I gapped down to 0.028.

[Speedytec]

We found a similar weird Dynamic switch over issue while tuning our Harrop SC-ed LSJ...

MAF and VE dialed in seperately with a Wideband O2 and also with fueltrims for the closed loop area.
PE is set for a straight 11.8 AFR. Dynamic VE/MAF blend switch-over point is the stock 3400rpm setting.

- On MAF only (Dynamic @ 200rpm) while in PE mode, the measured AFR line follows the commanded 11.8 AFR closely. (as it should.)
Dynamic set @200rpm (so MAF only):
Rob - VE Table tuned -dynamic switch point @200rpm.JPG

- In VE/MAF blend and PE enabled, Dynamic @3400rpm, we see a too lean AFR before 3400rpm (~13.5AFR) and after that rpm it makes the correct commanded ~11.8 AFR...
- This AFR change point moves with the Dynamic set point; if we set Dynamic f.i. @2000rpm, after that the AFR is correct...
Dynamic set @3400rpm:
Rob - VE Table tuned -dynamic switch point @3500rpm.JPG

- So VE table seems too lean => richened the whole VE map by 20%: gives the exact same too lean AFR behavior (~13.5AFR) before dynamic switch point, correct ~11.8 AFR after that rpm... So richening the VE map makes no difference at all in measured AFR or injector duty cycle compared to the original VE map...
Dynamic set @3400rpm + VE table +20%:
Rob - VE Table values x 1.2 - dynamic switch @3500rpm.JPG


Very confused and it seems that altering the VE table has no effect in normal VE/MAF blend mode? Are there different tables affecting PE AFR which we canor see/alter on the LSJ ecu??
Yes, we can simply lower the dynamic rpm to overcome this issue and go earlier MAF only, but I really want to know what we are missing here??

[AP83]

Its a primitive PCM that we only have access to half of.

I have always only used Dynamic at low rpms to help keep things smooth where the MAF is lower flow, namely idle and off-idle.

In my understanding, the P12 only uses the VE table in dynamic as a check against the MAF.

[eronq9]

Its a primitive PCM that we only have access to half of.

I have always only used Dynamic at low rpms to help keep things smooth where the MAF is lower flow, namely idle and off-idle.

In my understanding, the P12 only uses the VE table in dynamic as a check against the MAF.

It uses VE when theres quick throttle changes, I did see the dynamic lean issues on my friends cobalt when tuning the TVS as well but it went away as I spent more time on it. MAF should still be the main airflow table. Best bet is to log Maf Airflow, Dynamic, and Volumetric Efficiency Airflow. You can see quickly from those graphs how VE works.

Heres a pic from my TVS car, however its a 2004 so OS may be different

Dynamic.JPG

For now im going to just leave it off even if the car doesnt run as smoothly

[eronq9]

Ve is also important in starting the engine

Startup.JPG

As you can see the MAF doesnt measure the the air already in the manifold at atmospheric pressure that is first pulled into the engine on start that the VE can see from the MAP sensor

[Speedytec]

I know that VE is used for quick air changes to correct the slow MAF sensor, that's why I want to keep using the VE-MAF blend, but WHAT is causing it to run leaner than commanded in PE-mode untill this dynamic MAF only switch point?
And the LSJ ecu is not fully unlocked, but richen-up the VE map with 20% more fuel seems to have no effect at all on the lower rpm's too-lean part, but setting the dynamic rpm lower does change the lean AFR switch point... Really weird, or the LSJ's VE table is not the VE table that it uses used during normal (PE) operation??

ps; what parameters did you log there? Will check the LSJ's primary table.

[Speedytec]

PS2; you said; "I did see the dynamic lean issues on my friends cobalt when tuning the TVS as well but it went away as I spent more time on it"...
You mean you spend more time on tuning the VE table? We have that pretty well tuned in with fuel trim logging and also with Wideband O2 sensor...

[Speedytec]

same problem: https://www.hptuners.com/forum/showt...eling-Question

[cobaltssoverbooster]

did i understand that you tuned the maf, then tuned the ve, then re-enabled the dynamic switch over?

if that is how you did that then you need to disable the maf, and turn dynamic off. Completely tune the ve all the way through pe and then reactivate your maf and dynamic blend.
At this point it will run bad for a bit but you would now fix that by tuning your maf in.
The reason it runs bad when you tune maf first is because the base screws with the maf offset. the base is used as a maf compare table. Therefore the base is tuned first so it is correct and then the maf is tuned to match the base. If the maf fails and deviates from the base while driving the ecu will throw a maf signal code to then notify of the issue.

Your skills in setting up calibration devices and setting the stoich value and pe offset to match those devices is going to help smooth this issue out.
as with always, the injector settings in the lsj can make or break your dynamic transition since you dont get any o2 sensor calibration fields or transient condition modifiers to non-pe fueling

The ve fuel trims is your slow responder off the o2 sensor. the maf sensor is the fastest density calculating sensor on the vehicle.
and i quote from a prior post in 2009, (data is congruent to many articles on electrical engineering in the field of aerodynamics and automotive airflow)

"MAF systems are much more flexible in their ability to compensate for engine changes since they actually measure airflow instead of computing it based on preprogrammed assumptions. They are self-compensating for most reasonable upgrades, as well as extremely accurate under low-speed, part-throttle operation. On the other hand, the MAF meter, mounted as it is ahead of the throttle-body, can become an airflow restriction on high-horsepower engines. On nonstock engine retrofits or EFI conversions on engines never produced with fuel injection, it may be hard to package an MAF meter within the confines of the engine bay and available intake manifolding.

Which Is Best?
In a perfect world, virtually all street-performance engines would use Mass Air, due to its superior accuracy and greater tolerance for engine changes. In the past there was a problem on high-horsepower engines because larger-capacity MAF sensors were scarce and prohibitively expensive. Nowadays, oversize MAF sensors are available from Pro-M, Granatelli Racing, and other sources that are compatible with Ford engines and computers. Custom MAF calibration keyed to the specific vehicle, engine, and injector size is also available. With a correctly calibrated oversize meter, reflashing the computer usually isn't required".

[eronq9]

did i understand that you tuned the maf, then tuned the ve, then re-enabled the dynamic switch over?

if that is how you did that then you need to disable the maf, and turn dynamic off. Completely tune the ve all the way through pe and then reactivate your maf and dynamic blend.
At this point it will run bad for a bit but you would now fix that by tuning your maf in.
The reason it runs bad when you tune maf first is because the base screws with the maf offset. the base is used as a maf compare table. Therefore the base is tuned first so it is correct and then the maf is tuned to match the base. If the maf fails and deviates from the base while driving the ecu will throw a maf signal code to then notify of the issue.

Your skills in setting up calibration devices and setting the stoich value and pe offset to match those devices is going to help smooth this issue out.
as with always, the injector settings in the lsj can make or break your dynamic transition since you dont get any o2 sensor calibration fields or transient condition modifiers to non-pe fueling

The ve fuel trims is your slow responder off the o2 sensor. the maf sensor is the fastest density calculating sensor on the vehicle.
and i quote from a prior post in 2009, (data is congruent to many articles on electrical engineering in the field of aerodynamics and automotive airflow)

"MAF systems are much more flexible in their ability to compensate for engine changes since they actually measure airflow instead of computing it based on preprogrammed assumptions. They are self-compensating for most reasonable upgrades, as well as extremely accurate under low-speed, part-throttle operation. On the other hand, the MAF meter, mounted as it is ahead of the throttle-body, can become an airflow restriction on high-horsepower engines. On nonstock engine retrofits or EFI conversions on engines never produced with fuel injection, it may be hard to package an MAF meter within the confines of the engine bay and available intake manifolding.

Which Is Best?
In a perfect world, virtually all street-performance engines would use Mass Air, due to its superior accuracy and greater tolerance for engine changes. In the past there was a problem on high-horsepower engines because larger-capacity MAF sensors were scarce and prohibitively expensive. Nowadays, oversize MAF sensors are available from Pro-M, Granatelli Racing, and other sources that are compatible with Ford engines and computers. Custom MAF calibration keyed to the specific vehicle, engine, and injector size is also available. With a correctly calibrated oversize meter, reflashing the computer usually isn't required".

cobaltssoverbooster,

thanks for the good information, now i see why when I continued to retune both tables on the TVS setups, they fixed the issues. So essentially VE should be tuned first once injector data is set and then MAF afterwards? This makes sense, I may give this a shot once I have a nice day outside where I can really get the VE table within 1, then retune MAF.

Ill then test dynamic open loop at 14.7 and see how it goes. Thanks

[Speedytec]

did i understand that you tuned the maf, then tuned the ve, then re-enabled the dynamic switch over?

if that is how you did that then you need to disable the maf, and turn dynamic off. Completely tune the ve all the way through pe and then reactivate your maf and dynamic blend.
At this point it will run bad for a bit but you would now fix that by tuning your maf in.
The reason it runs bad when you tune maf first is because the base screws with the maf offset. the base is used as a maf compare table. Therefore the base is tuned first so it is correct and then the maf is tuned to match the base. If the maf fails and deviates from the base while driving the ecu will throw a maf signal code to then notify of the issue.

Your skills in setting up calibration devices and setting the stoich value and pe offset to match those devices is going to help smooth this issue out.
as with always, the injector settings in the lsj can make or break your dynamic transition since you dont get any o2 sensor calibration fields or transient condition modifiers to non-pe fueling

The ve fuel trims is your slow responder off the o2 sensor. the maf sensor is the fastest density calculating sensor on the vehicle.
and i quote from a prior post in 2009, (data is congruent to many articles on electrical engineering in the field of aerodynamics and automotive airflow)

"MAF systems are much more flexible in their ability to compensate for engine changes since they actually measure airflow instead of computing it based on preprogrammed assumptions. They are self-compensating for most reasonable upgrades, as well as extremely accurate under low-speed, part-throttle operation. On the other hand, the MAF meter, mounted as it is ahead of the throttle-body, can become an airflow restriction on high-horsepower engines. On nonstock engine retrofits or EFI conversions on engines never produced with fuel injection, it may be hard to package an MAF meter within the confines of the engine bay and available intake manifolding.

Which Is Best?
In a perfect world, virtually all street-performance engines would use Mass Air, due to its superior accuracy and greater tolerance for engine changes. In the past there was a problem on high-horsepower engines because larger-capacity MAF sensors were scarce and prohibitively expensive. Nowadays, oversize MAF sensors are available from Pro-M, Granatelli Racing, and other sources that are compatible with Ford engines and computers. Custom MAF calibration keyed to the specific vehicle, engine, and injector size is also available. With a correctly calibrated oversize meter, reflashing the computer usually isn't required".

Good info, just trying to understand it all correctly...
(What do you mean exactly with "Base"? I assume the VE table? MAF offset is a parameter I saw in other ecu's, but not present in HPT for the LSJ. And "Dynamic turned off" you mean a switch point below idle rpm?)

I'm not sure anymore what we tuned first, but that does not really matter as it's a continuing proces...
What we do:
First set the correct injector parameters. (We run a boost referenced system, so one IFR value and offsetts for the whole pressure range.)
- For MAF tuning we set Dynamic below idle, so 200rpm or so. Dial in the MAF curve with fuel trims in closed loop and Wideband O2 in open loop mode.
- For VE tuning we fail-set the MAF sensor and tune VE. Also with fuel trims in closed loop and with wideband in Open Loop. I don't think we turn Dynamic off here. Is that necessary? MAF is failed and not active.
(In Open Loop tuning we zero out the OL Gas (Gear)/ Gas (P/N)/ IVT Gain tables, as recommended everywhere.)

You say that after VE is fully tuned, we should re-tune the MAF (with dynamic set low) again because it is still active in the background? But when I set dynamic low, the MAF AFR already follows commanded AFR exactly, so not much to improve there.
Or do you mean that once we enable dynamic again, we must re-tune the MAF in the blend area? That is not logical then, as without Dynamic blend the MAF is dialed in perfectly...

And as said; 20% enrichment of the VE table seem to have no effect at all on the measured open loop AFR once you set dynamic back to the 3400rpm blend...

[cobaltssoverbooster]

so in lnf and other ecu's they have maf base tables which is a 3rd axis correction plane. in the lsj the base i refer to is the VE table. Your assumption here was correct.
When dynamic is turned off the enable rpm i set to the maximum allowed rpm value. If you tune the ve you not only set the maf fail high to the minimum value but you also raise the dynamic enable to its max allowed value. this makes the ecu ignore maf input regardless of its activation state just in case something is missed in setting up the ve file.

you shouldn't be tuning the maf before you tune the ve table. Thats what i was afraid you were doing. People who wish to be quick and dirty tuners will enable the maf through 200rpm to max rpm by lowering the dynamic enable rpm which allows it to always remain active. in this method it is possible to tune the maf without touching the ve settings but it is crude and not recommended as it is not proper. If the ve is to be tuned in any maf assisted setup it is always done first before setting the maf in high end practices.

VE is the table that determines how air is being used by the engine. Everyone associates it with fuel because they raise a value in the table and it goes richer. Vice versa for leaning out the mixture. Truth is the VE table is programmed to the airflow the engine is using. Volumetric Efficiency is a term used for giving a value to how well something pumps. (Pumping efficiency) So, since the VE is the main efficiency table, the ecu will recall its values ALWAYS, and FIRST. The MAF is a sensor that is fast reacting, due to its description previously posted. The ecu uses this sensor to add to the accuracy of the VE table since the VE table compared against MAP, BARO, and O2 tend to be slower reacting and don't adjust for variances as fast as the maf allows. Since the MAF is a secondary to the VE, that is why it is always tuned second. Everything is compared to the VE....VE+MAP correction+Baro correction+O2 correction+MAf correction.(and some others).

There is a possibility that the maf can be adjusted for prior to the ve and not effect its afr output but that is only if the ecu recognizes it as a heavily weighted Volumetric Efficiency value. In which case if tuned against a wideband with proper setup can net results that dont change when ve is changed from underneath it. I usually dont see this happening as when i mix them up i personally end up with errors which is why i always follow the proper calibration process.
could be why your blend issue is covering up your ve table errors. worth checking out.

[eronq9]

so in lnf and other ecu's they have maf base tables which is a 3rd axis correction plane. in the lsj the base i refer to is the VE table. Your assumption here was correct.
When dynamic is turned off the enable rpm i set to the maximum allowed rpm value. If you tune the ve you not only set the maf fail high to the minimum value but you also raise the dynamic enable to its max allowed value. this makes the ecu ignore maf input regardless of its activation state just in case something is missed in setting up the ve file.

you shouldn't be tuning the maf before you tune the ve table. Thats what i was afraid you were doing. People who wish to be quick and dirty tuners will enable the maf through 200rpm to max rpm by lowering the dynamic enable rpm which allows it to always remain active. in this method it is possible to tune the maf without touching the ve settings but it is crude and not recommended as it is not proper. If the ve is to be tuned in any maf assisted setup it is always done first before setting the maf in high end practices.

VE is the table that determines how air is being used by the engine. Everyone associates it with fuel because they raise a value in the table and it goes richer. Vice versa for leaning out the mixture. Truth is the VE table is programmed to the airflow the engine is using. Volumetric Efficiency is a term used for giving a value to how well something pumps. (Pumping efficiency) So, since the VE is the main efficiency table, the ecu will recall its values ALWAYS, and FIRST. The MAF is a sensor that is fast reacting, due to its description previously posted. The ecu uses this sensor to add to the accuracy of the VE table since the VE table compared against MAP, BARO, and O2 tend to be slower reacting and don't adjust for variances as fast as the maf allows. Since the MAF is a secondary to the VE, that is why it is always tuned second. Everything is compared to the VE....VE+MAP correction+Baro correction+O2 correction+MAf correction.(and some others).

There is a possibility that the maf can be adjusted for prior to the ve and not effect its afr output but that is only if the ecu recognizes it as a heavily weighted Volumetric Efficiency value. In which case if tuned against a wideband with proper setup can net results that dont change when ve is changed from underneath it. I usually dont see this happening as when i mix them up i personally end up with errors which is why i always follow the proper calibration process.
could be why your blend issue is covering up your ve table errors. worth checking out.

This is definitely some really good info about how the VE table is really used, I always understood it as the pumping efficiency but I always just kinda thought of it as an alternative to MAF airflow, now it seems to make alot more sense in terms of how the computer is analyzing the data from the sensors, and why my TVS setup suffers no issues and hits commanded AFR perfect in dynamic, because I tuned VE and retuned MAF several times meaning the last time I updated my MAF calibration it was over a correct tuned VE table where I has the settings at 7000rpm disable 6900rpm enable. I never understood why It just "got better" over time, but now I really understand whats going on here.

[Speedytec]

so in lnf and other ecu's they have maf base tables which is a 3rd axis correction plane. in the lsj the base i refer to is the VE table. Your assumption here was correct.
When dynamic is turned off the enable rpm i set to the maximum allowed rpm value. If you tune the ve you not only set the maf fail high to the minimum value but you also raise the dynamic enable to its max allowed value. this makes the ecu ignore maf input regardless of its activation state just in case something is missed in setting up the ve file.

you shouldn't be tuning the maf before you tune the ve table. Thats what i was afraid you were doing. People who wish to be quick and dirty tuners will enable the maf through 200rpm to max rpm by lowering the dynamic enable rpm which allows it to always remain active. in this method it is possible to tune the maf without touching the ve settings but it is crude and not recommended as it is not proper. If the ve is to be tuned in any maf assisted setup it is always done first before setting the maf in high end practices.

VE is the table that determines how air is being used by the engine. Everyone associates it with fuel because they raise a value in the table and it goes richer. Vice versa for leaning out the mixture. Truth is the VE table is programmed to the airflow the engine is using. Volumetric Efficiency is a term used for giving a value to how well something pumps. (Pumping efficiency) So, since the VE is the main efficiency table, the ecu will recall its values ALWAYS, and FIRST. The MAF is a sensor that is fast reacting, due to its description previously posted. The ecu uses this sensor to add to the accuracy of the VE table since the VE table compared against MAP, BARO, and O2 tend to be slower reacting and don't adjust for variances as fast as the maf allows. Since the MAF is a secondary to the VE, that is why it is always tuned second. Everything is compared to the VE....VE+MAP correction+Baro correction+O2 correction+MAf correction.(and some others).

There is a possibility that the maf can be adjusted for prior to the ve and not effect its afr output but that is only if the ecu recognizes it as a heavily weighted Volumetric Efficiency value. In which case if tuned against a wideband with proper setup can net results that dont change when ve is changed from underneath it. I usually dont see this happening as when i mix them up i personally end up with errors which is why i always follow the proper calibration process.
could be why your blend issue is covering up your ve table errors. worth checking out.

I fully understand your explanation about VE. (Actually its GMVE, which contains a bit more than a pure VE table. There's a long explanation about this by RedSupra or so somewhere)
And sorry by dragging this along, but let's be clear here that this is not about a on-shot quick tune: "We tuned MAF, We tuned VE => We have an AFR issue... Help?"

- Simply lets assume the MAF was in an "untuned" state while we mapped the VE table. (MAF failed, but we did not raise the Dynamic to max rpm)
- Then after that we tune the MAF sensor (again) with Dynamic set below idle; MAF curve needs almost no corrections to be spot on with the commanded AFR. So MAF is tuned after VE.
- Then we blend in Dynamic again @3400rpm and we find this weird too lean AFR below the Dynamic switch point. (see previous post for the exact behavior and screenshots)

I have to talk to my mate to ask if after the VE table was raised by 20%, he re-checked/tuned the MAF curve again. But I find it really weird that raising the LSJ's VE table by 20% has absolutely no effect on measured AFR in the area where VE should be effective.
(And in the case that the by VE calculated fuel is "adjusted" with the MAF sensor value, why is it not corrected to the correct AFR, as in MAF only mode the AFR is spot-on??)

Lot's of confusion, but we simply have to dive into this more deeply and see if it corrects itself by more VE & MAF tuning. And report back.

[Area47]

with a properly tuned ve table, it will still exhibit the same issues. save yourself the head ache and time and disable the VE referencing. Went through this in 2008.

[AP83]

with a properly tuned ve table, it will still exhibit the same issues. save yourself the head ache and time and disable the VE referencing. Went through this in 2008.

This has always been my understanding and practice. I only use dynamic down low because it does seem to actually make a difference at idle and off-idle. Maybe it's all in my head. Down the road though, the MAF is king.

[cobaltssoverbooster]

Dynamic can not be limited to only low rpm as once its activation rpm is set low(below idle rpm) in the lsj it forces the system to run dynamic entirely. (MAF tune) This is the "disabled VE referencing" i'm assuming Area47 is speaking of. If the maf is fully active the entire time then the switch over point stumble will not be an issue since the tables do not technically switch(crossover point is eliminated in this method.)
Still would be beneficial to complete the ve tune as a method for sensor fault protection.

I do not think you guys are trying to be maf only tuners, you know more than the people who shoot for that method. Other people are reading in here to so i added it in so the other readers can see what the difference is. I am sorry for the confusion.

As for your lean error i have solved that before by monitoring just prior to activation, activation point, and post activation point. I then use the data to blend the crossover. Unfortunately crossover is not a very clean process, which has to be why Area47 claims he always see's this issue regardless of ve and maf tuned quality. You can limit it by sacrificing rich/lean events on each table at the crossover to blend the result into an acceptable afr result. This can cause a consistent offset error though so its not recommended. I ended up leaving the lean event but just pushed the maf activation up to a higher rpm such as 3500-4000 rpm. By pushing the dynamic blend to above the cruise zone the maf will be calibrated strictly for an rpm accelerating state which caused a reduction in error at crossover. I just tested this on my yukon, since the factory crossover is set to dynamic enable of 3500. by pushing it lower than idle i can fully tune the maf no problem but with the activation set to 1800 i noticed an afr swing as the tables swung between cruise, light acceleration, and deceleration and crossed the crossover point. Have you guys tried pushing the dynamic up to a more mid rpm zone to see if the error reduces? I also have an injector theory behind the higher activation point but i dont feel i should include it as it may cause more confusion. If this works for you guys ill throw it in here as something to ponder on.

[Speedytec]

Just digging through the forum for this and the V8 guys seem to have exact similar issues:
https://www.hptuners.com/forum/showt...ransition-Lean.

Running full MAF from low rpms is a simple fix, but will log MAF airflow, VE airflow and Dynamic airflow to see if we can see something that can explain this.

(Although it seems that the VE table in LSJ-HPT is only active once the MAF fails, but not doing much in Dynamic mode...)