LSA LQ4 Tuning Help

[kingtal0n]

I have an old set of fic650s that’ll run 720cc at 58 pounds that I may toss in and see what happens, should be running a higher DC with those if I can’t run Lsa injectors.

You are wasting time fooling around with injectors and pressure. Put the largest injectors you have in, And tune the damn thing.

You could put the lowest pressure possible 38 or 44psi or whatever you want to fuel your wide open throttle max condition ideal is 40-50% duty cycle but you can go 80% or whatever it doesn't make that big of a difference when you are no where near optimization point of the process. I just don't like high fuel pressure because it wears everything out and strains the fuel pump and heats the fuel and kills voltage at the pump and kills the flow rate of the fuel and can lead to catastrophic failure when using typical in-tank fuel pumps but thats just me.

So tune it. Adjust the VE table to achieve target a/f values then check torque or target afr. The target is set either by 'commanded afr' or by your own desired torque metric. Ask for clarification if you need it. Either is fine in gen3 applications because the gen3 computer doesnt give a fuck all about torque model if you the user/tuner understand how to compensate for high or low torque outside of the model e.g. force motor current table. I have an advantage that I build and tune my own transmissions so I was able to figure this out quickly but many here do not understand that and do not build their own engines and transmissions yet.

The point on the log I took a snapshot of is not a transient condition, the steady state values of the VE table are intended to reflect static or holding conditions of the TPS/KPA/IAT/etc... however RPM moving is not technically a transient. I cannot think of any modifiers involving rate of change RPM that has to do with fuel. I've never had to program a transient fuel condition specifically for RPM changing that wasn't embedded to the fuel map. RPM is always changing even at idle. The piston speeds up, then it slows down, then it speeds up again, slows down again. As the RPM is rising with some load this is a static feature for a given vehicle weight and gear ratio. For example on a load dyno you can stop the engine from accelerating and hold the specific load spot on the VE table and tune it like a steady exact spot. However this will NEVER reflect real road conditions because the vehicle will never be heavy enough or geared enough to achieve that same condition, so it would be pointless to tune it that way, i.e. you are not looking for a steady state loaded condition in boost to tune. Furthermore each gear has its own rate of change of RPM so even if you did tune one gear perfectly the other gear ratios and rates will never line up. Therefore to tune VE properly we imagine the philosophy of tuning a VE table- for example, to keep it smooth. To Give it what it wants. To Make a logical shape to the VE table no matter what our wideband or other sensors say. And mainly not to rely on the ECU or SCANNER to tune the engine overall. But that is getting in advanced topics nobody will read it when they are novices. Nobody wants to hear about how I do not use widebands or logging to tune my engine.

[GHuggins]

I assume reading plugs, combustion colors, mpg numbers and sound/feel? I did my own low hp stuff like that many years ago. Even played with sensors cause that's all we had to make adjustments with at that time. I prefer to see data on top of everything else now. Just extra assurance.

[SiriusC1024]

If the offset is incorrect the VE Tables near low non-linear pulses will be nonsense, but the normal driving portions will barely notice because pulses are much higher. Unless the injector is very large for some reason.

I already talked about that in this thread.

More responsiveness because a larger proportion of injector flow is linear. Decreased pressure means increased duty cycle. Even with data to address this the data is often slightly inaccurate for aftermarket injectors.

LSA injectors fire at an angle. Standard straight spray injectors will have issues with port wetting.

When you reduce fuel pressure offset and SPA will be affected. No data available to compensate, so that's another problem. Offset can be roughed in by multiplying by SQRT(P2/P1), 0.866 in this case, but SPA is another matter.

You're too quick with your "corrections" just like with the wideband formula on the other thread.

Look at OP's VE. Now look at log and what STFT's are doing. Injectors are oversized. It's obvious he's having a little trouble getting it right at idle. Return to idle is part of normal driving. Idle surge, hunting, and stalling can be extremely annoying in a street car. Don't Bill Clinton me about "define normal driving". We all know what that means.

What mean normal driving? DEFINE normal driving please.

bill define.png

The idea of turning it down to 3 bar is good, but you still have to take into account how this affects High Slope, Low Slope, and Breakpoint. Offset, SPA, Min Pulse, Min Fuel Milligrams all matter as much as IFR.

[schiavonske1013]

I already talked about that in this thread.



You're too quick with your "corrections" just like with the wideband formula on the other thread.

Look at OP's VE. Now look at log and what STFT's are doing. Injectors are oversized. It's obvious he's having a little trouble getting it right at idle. Return to idle is part of normal driving, too. Idle surge, hunting, and stalling can be extremely annoying in a street car. It will also hold true for VE during decel. Don't Bill Clinton me about "define normal driving". We all know what that means.


bill define.png

The idea of turning it down to 3 bar is good, but you still have to take into account how this affects High Slope, Low Slope, and Breakpoint. Offset, SPA, Min Pulse, Min Fuel Milligrams all matter as much as IFR.

At this point I may just toss in the fuel injector connection 650s and see what happens, easy to change on this setup and can’t hurt. Still tossing around the thought of Lsa injectors to combat transients, there’s claims they can handle good power if the regulator is at 4 bar and boost referenced, but I guess check what I have first. What kingtalon says lines up with people that claim “you can throw id1700s at 650hp and be fine” but that seems too high.

[schiavonske1013]

Ok well have him define "fine".

I disagree about the big injectors thing. Put in the minimum sized injectors that'll get you ~80% DC. This keeps nonlinear behavior to a minimum.

Injectors need good GM data to run right in a GM vehicle.

LSA injectors would be the best for this application, but I figured there was a reason you didn't want them.

Only real reason was the setup came with them, reading into people claim LSA injectors were "useless with more boost", but then looking at a few people randomly who said if you set it up right they are fine. I may just go that route.

[SiriusC1024]

Ok well have him define "fine".

I disagree about the big injectors thing. Put in the minimum sized injectors that'll get you ~80% DC. This keeps nonlinear behavior to a minimum.

Injectors need good GM data to run right in a GM vehicle.

LSA injectors would be the best for this application, but I figured there was a reason you didn't want them.

[SiriusC1024]

Sorry for the repost. Technical difficulties here...

If you decide to go bigger than LSA injectors then at least make sure there is GM data.

[schiavonske1013]

Ok well have him define "fine".

I disagree about the big injectors thing. Put in the minimum sized injectors that'll get you ~80% DC. This keeps nonlinear behavior to a minimum.

Injectors need good GM data to run right in a GM vehicle. There's really no such thing as universal injector data.

LSA injectors would be the best for this application, but I figured there was a reason you didn't want them.

The setup came with the blower, most people claim LSA injectors are "useless" if you up the boost, but then there are a few who would say if setup right they are fine.

[SiriusC1024]

Haha

[schiavonske1013]

Haha

I've had to fix so many little things on this truck since I put it together, I just want it to work lol

[kingtal0n]

I assume reading plugs, combustion colors, mpg numbers and sound/feel? I did my own low hp stuff like that many years ago. Even played with sensors cause that's all we had to make adjustments with at that time. I prefer to see data on top of everything else now. Just extra assurance.

In 1995 that is how I do things because there is no wideband sensor. I bought a https://www.dataq.com/ Serial 9-pin analog input for my narrowband to use it on a carb blower application just to find tan center and back then I didn't know about optimal a/f ratios 11:1 or 12:1 for a blower/turbo/nitrous or whatever we just looked at the plugs for detonation and the dynojet for peak output while being cautious with loading and gear changes on the actual road in terms of timing. That was back when a 12 second car was fast. In 1995 Toyota Supra has 440rwhp capable completely factory at sea level in FLorida and this was considered pretty good, something to 'beat'.

By 1998 I started tuning TPI systems with the EEPROM got into building 700R4 transmissions and realized the deficiency of burning proms due to down time (sadly its alot like stopping the engine to upload a new tune file for the OEM ecu). In 2001 I jumped at an opportunity to beta test a commander 950 for my TPI and wrote software that uses narrowbands to auto-tune the engine and it is essentially the outline for all modern closed loop wideband controllers doing the same exact thing just better with faster computers. That was the birth of the commander 950 pro ECU and I ran my software on a twin turbo 355 small block chevy that I had garbled together on a best buy paycheck as a college drop out.
https://www.supraforums.com/threads/.../post-13987029



So to answer your question, yes I absolutely was born in a pre-computer era that do not trust sensors or widebands or whatever when I can avoid them and physically understand how an engine functions by virtue of its mechanical, sensational, and perturbances. My wideband is the soot forming pseudo particulate at wide open throttle at the exhaust exit with correct smells , then we get the clean appearance of spark plugs a certain way I expect them to look over time. My knock sensor is in the steric hindrance of hydrocarbons during chain reaction of combustion due to energy input, rod angle and the critical moment you decide to break an O2 bond yielding oxygen radicals at some frequency rate. My timing and torque is application dependent and typically I tune daily drivers with 2x 3x 4x factory output using factory internals so minimum cylinder pressure and control the EGT is the name of the game. My EGT sensor is the rate of fuel flowing compared to the work being done and displacement/heat capacity of the engine and its glorious fluids. I can smell when the oil temperatures are getting high, can't you smell it? The sort of heat emanating from the engine and the change in roughness as viscosity dwindles. I've been tuning engines my entire life since before I was old enough to drive and I have developed these skills and alongside them computer programming skills which allow me to understand input/output well enough to break the ECU wide open and extract what I want. If I was tuning a later model Gen4 5 ECU I would use external microcontrollers to input signals that it wants to 'see' to trick it like a super AFC for example, performing 'magic' and setting a new standard, but I have no need for such extremities as the gen3 computer is ideal for performance applications sub 1200rwhp using factory modern 02-07 V8 Chevrolet engines which is all I need in daily driver applications.

Widebands
When I see wideband output I do not see a number on the gauge. I do not care what the number on the gauge is. I Never used to log it either, that is new just so i can share tuning efforts online, I never needed it to tune although I admit it does help to review the log , it isn't necessary to tune an engine. People get hung up on electronics and logging when it really just takes your attention away from the real issues such as setting up mechanical aspects with far more important impactive on the way an engine runs and responds such as pressure testing and PCV components, which is ironic people spend so much time looking in the ECU and scanner for an answer and wind up tuning around some ridiculous behaviors only because they failed to challenge the engine mechanically prepared for adequate service interval.

With experience top tuners learn about the little offsets, we learn to 'see' the true number on the wideband gauge. We convey that information to our spark plug appearance and bounce it from the dynojet and other tests, and after enough engines we no longer need to look at the plugs because we already know how to correlate the behaviors of the engine with the number on the gauge which yield the optimal settings for a given placement even if and when those settings do not make any sense to the ECU.

I could have answered with a simple yes but then what fun would that be

[GHuggins]

Not sure how y'all got it to post responses backwards to one another unless you just deleted post but cool...

LSA injectors will work best in this application and with the the boost ref and no ethanol they'll be fine. A lot of the transient problem you're having is coming from the distance to valve that this setup has created. It often requires OL tables to be modified along with inj timing to be advanced. You have a blower where the inj hole points down instead of at an angle plus adapter plates that are exaggerating the issue. Any injectors could be tuned in but would require some serious changes and experiments. The lsa injectors are already injecting at an angle as they are purposely built for this blower. Make sure you turn them right when installed.

You might also want to keep in mind that boost spikes are caused by misfires. Something about how the airflow is slowed due to the misfiring cylinder. Might be worth checking and or changing plugs while installing the other injectors. Get some nice OE boost plugs. Can clean up a lot on their own.

[SiriusC1024]

Not sure how y'all got it to post responses backwards to one another unless you just deleted post but cool...

I hit refresh then it showed two of the same posts, so I hit delete on one, but they both went away. Closed the tab reopened and redid it. When you hit refresh it'll show your reply posted as well as the same reply sitting there ready to go. Kind of annoying.

[GHuggins]

In 1995 that is how I do things because there is no wideband sensor. I bought a https://www.dataq.com/ Serial 9-pin analog input for my narrowband to use it on a carb blower application just to find tan center and back then I didn't know about optimal a/f ratios 11:1 or 12:1 for a blower/turbo/nitrous or whatever we just looked at the plugs for detonation and the dynojet for peak output while being cautious with loading and gear changes on the actual road in terms of timing. That was back when a 12 second car was fast. In 1995 Toyota Supra has 440rwhp capable completely factory at sea level in FLorida and this was considered pretty good, something to 'beat'.

By 1998 I started tuning TPI systems with the EEPROM got into building 700R4 transmissions and realized the deficiency of burning proms due to down time (sadly its alot like stopping the engine to upload a new tune file for the OEM ecu). In 2001 I jumped at an opportunity to beta test a commander 950 for my TPI and wrote software that uses narrowbands to auto-tune the engine and it is essentially the outline for all modern closed loop wideband controllers doing the same exact thing just better with faster computers. That was the birth of the commander 950 pro ECU and I ran my software on a twin turbo 355 small block chevy that I had garbled together on a best buy paycheck as a college drop out.
https://www.supraforums.com/threads/.../post-13987029



So to answer your question, yes I absolutely was born in a pre-computer era that do not trust sensors or widebands or whatever when I can avoid them and physically understand how an engine functions by virtue of its mechanical, sensational, and perturbances. My wideband is the soot forming pseudo particulate at wide open throttle at the exhaust exit with correct smells , then we get the clean appearance of spark plugs a certain way I expect them to look over time. My knock sensor is in the steric hindrance of hydrocarbons during chain reaction of combustion due to energy input, rod angle and the critical moment you decide to break an O2 bond yielding oxygen radicals at some frequency rate. My timing and torque is application dependent and typically I tune daily drivers with 2x 3x 4x factory output using factory internals so minimum cylinder pressure and control the EGT is the name of the game. My EGT sensor is the rate of fuel flowing compared to the work being done and displacement/heat capacity of the engine and its glorious fluids. I can smell when the oil temperatures are getting high, can't you smell it? The sort of heat emanating from the engine and the change in roughness as viscosity dwindles. I've been tuning engines my entire life since before I was old enough to drive and I have developed these skills and alongside them computer programming skills which allow me to understand input/output well enough to break the ECU wide open and extract what I want. If I was tuning a later model Gen4 5 ECU I would use external microcontrollers to input signals that it wants to 'see' to trick it like a super AFC for example, performing 'magic' and setting a new standard, but I have no need for such extremities as the gen3 computer is ideal for performance applications sub 1200rwhp using factory modern 02-07 V8 Chevrolet engines which is all I need in daily driver applications.

Widebands
When I see wideband output I do not see a number on the gauge. I do not care what the number on the gauge is. I Never used to log it either, that is new just so i can share tuning efforts online, I never needed it to tune although I admit it does help to review the log , it isn't necessary to tune an engine. People get hung up on electronics and logging when it really just takes your attention away from the real issues such as setting up mechanical aspects with far more important impactive on the way an engine runs and responds such as pressure testing and PCV components, which is ironic people spend so much time looking in the ECU and scanner for an answer and wind up tuning around some ridiculous behaviors only because they failed to challenge the engine mechanically prepared for adequate service interval.

With experience top tuners learn about the little offsets, we learn to 'see' the true number on the wideband gauge. We convey that information to our spark plug appearance and bounce it from the dynojet and other tests, and after enough engines we no longer need to look at the plugs because we already know how to correlate the behaviors of the engine with the number on the gauge which yield the optimal settings for a given placement even if and when those settings do not make any sense to the ECU.

I could have answered with a simple yes but then what fun would that be

I hated tuning TPI although I had to do it with OE settings and in the dealership to make them run right. You learned a lot playing with the older engines. I probably tuned as many carbureted flat head V8's when I started in the performance industry as I did big hp LS's due to where the shop was. Lots of wealthy people with very old show cars. Even worked on an old Truman clone car with the suicide doors. Funny thing though, I used widebands on them on top of TB flow matching meters to balance everything out. Personally I'm glad the newer stuff is here even though they are a pain to get right and yes you are often skewing things to work around junk AM parts or part combo's.

[schiavonske1013]

Not sure how y'all got it to post responses backwards to one another unless you just deleted post but cool...

LSA injectors will work best in this application and with the the boost ref and no ethanol they'll be fine. A lot of the transient problem you're having is coming from the distance to valve that this setup has created. It often requires OL tables to be modified along with inj timing to be advanced. You have a blower where the inj hole points down instead of at an angle plus adapter plates that are exaggerating the issue. Any injectors could be tuned in but would require some serious changes and experiments. The lsa injectors are already injecting at an angle as they are purposely built for this blower. Make sure you turn them right when installed.

You might also want to keep in mind that boost spikes are caused by misfires. Something about how the airflow is slowed due to the misfiring cylinder. Might be worth checking and or changing plugs while installing the other injectors. Get some nice OE boost plugs. Can clean up a lot on their own.

I did actually change from cathedral port with adapters to rectangular port to ditch the adapters after reading some of smokeshows and others comments on the situation hoping that would help, but it didn't do nearly as much. But two votes for LSA injectors lol. Also just got in new TR7IX so by coincidence that will get taken care of.

[kingtal0n]

You're too quick with your "corrections" just like with the wideband formula on the other thread.

I told you that you need analog offset term. I show you the term you forgot. Then you made a new thread because you swore I was wrong. But you were told the same exact thing again by somebody else that knows.

I prefer for all of my customers to use the AEM CAN setups for this reason...

Hard to say on this one just for the simple fact that the O2's are already prioritizing to the rich side and not switching uniformly. This can be a wideband that needs to be cleaned or just an offset issue. You should only need to use a math to correct the wideband if an offset exist and even then I would want to gather data at a higher rpm to verify this as the O2 and wideband errors alike start coming more in line with one another at higher rpms so an offset issue would be more evident here.

You got told twice that the offset is a critical aspect of wideband setting up. And you failed to mention the offset when teaching the math formula to the other people so it was a CORRECTIVE MEASURE that I will and always will take to explain to people that you MUST account for these offsets to line up the wideband log with the actual wideband air fuel ratios that you have determined, as an advanced knowledgeable tuner, to be accurate. Which can only happen if you actually know the true wideband number, not the number on the gauge but the actual wideband number based on all of your internal brain offsets not just the analog offset.

Look at OP's VE. Now look at log and what STFT's are doing. Injectors are oversized. It's obvious he's having a little trouble getting it right at idle. Return to idle is part of normal driving. Idle surge, hunting, and stalling can be extremely annoying in a street car. Don't Bill Clinton me about "define normal driving". We all know what that means.

Bah Large injectors do not cause problems. I've tuned 2000cc/min injectors on a 2L engine to idle perfectly on sr20det and 4g63 at 15:1. Its not a injector problem. Its a user problem.
Want a little tips for these,
1. Watch wideband curve behavior, ignore the number just focus on the behavior and treat it like a picture from the past instead of a real time agent.
2. Settle wideband curve behavior no matter what nonsense you need to do to the ECU
3. Tune the airflow adders and throttle cracker tables to gently resume to idle
4. Tune the idle PID to control the idle properly based on cam roughness and so forth so that the idle correction doesn't overcorrect or interfere with return to idle on a lightweight rotating engine (lightweight flywheel and knife edged crank for example need very tenuous PID controller settings).
5. Smooth the VE table in the idle regions and as part of the idle control features, e.g. adding fuel to a VE cell can draw the engine to that cell which can help prevent oscillation by 'directing traffic' when it becomes unstable.



I could do it. That is what frustrates me. I can tune all of this shit perfectly put whatever injectors you want in there I'll fix the signals and make it run perfect in gen3. You have to tune a performance engine without closed loop FIRST so that the baseline running conditions are PERFECT and VE table is PERFECT before you enable the closed loop shit which puts spikes onto the fuel map and leads to oscillation behaviors. It happens when you try to tune a ECU while the computer is also making its own corrections at the same time, its like having two tuners tuning the same ECU at the same time, of course it will become unstable. You can fight the computer and make corrections over long periods of time to follow the ECU but why when the behind the scenes IAT and CTS is just going to scale your a/f whenever the temperature change anyways. The real goal should be to generate adequate VE curvature not ideal VE values. This stabilize operations and then you can let the ECU fight with itself as conditions change.

[SiriusC1024]

I told you that you need analog offset term. I show you the term you forgot. Then you made a new thread because you swore I was wrong. But you were told the same exact thing again by somebody else that knows.

You got told twice that the offset is a critical aspect of wideband setting up. And you failed to mention the offset when teaching the math formula to the other people so it was a CORRECTIVE MEASURE that I will and always will take to explain to people that you MUST account for these offsets to line up the wideband log with the actual wideband air fuel ratios that you have determined, as an advanced knowledgeable tuner, to be accurate. Which can only happen if you actually know the true wideband number, not the number on the gauge but the actual wideband number based on all of your internal brain offsets not just the analog offset.

Go to that CL vs OL debate thread and make the offset math work to correct post #2. I'll wait. It's not an offset problem.

As far as the wideband egr math. No offset was needed since wideband was tracking commanded in that case. I didn't forget anything. I gave him the formula that works. Also, it's ironic that I'm being "corrected" by the guy who couldn't even get the formula right for the particular wideband. Why don't you go to that thread and correct things?

Back to the subject. This will tune much easier with LSA injectors.

[schiavonske1013]

Went from tr6 to tr7ix and it already feels better, but Lsa injectors here I come. I appreciate everyone’s responses, once I get that done I’ll post a log.

[kingtal0n]

Go to that CL vs OL debate thread and make the offset math work to correct post #2. I'll wait. It's not an offset problem.

Its always an offset problem when the input is analog. The END USER has total control over what the wideband logs in the scanner. Therefore, if the scanner is logging wrong values which disagree with other analyzers, spark plugs, or other devices such as narrowbands, then the offset is wrong because the tuner failed to correct it. There is no other adjustment. The wideband doesn't have a little dial on the face for you to adjust the analog output but maybe it should. It only reads whatever you tell it to.

As far as the wideband egr math. No offset was needed since wideband was tracking commanded in that case. I didn't forget anything. I gave him the formula that works.

Eh no. That is very scary. There is always some offset, every car, every time. Even if the offset is small there is still some. If they match then its wrong. If you put the right offset and they match then its still wrong. Current flowing to ground changes offset and current always changes from the ECU perspective. You must ALWAYS include offset term unless its a canbus. The fix is to use a canbus wideband which does not have offset analog. Your entire thread could be thrown away into the trash if you just had a canbus wideband which is digital communication between microcontrollers eliminating the offset of analog voltage. Your thread basically revolves around a tuner who does not yet realize that THEY ARE THE ONES CONTROLLING THE SCANNER VIA OFFSET. If its wrong, its your fault for not using a canbus wideband or for not understanding analog offset.

You MUST teach analog offset terms when giving people formulas no matter what. It is 99.9999% impossible for there to be no offset when current changes for any analog microcontroller reading at multiple levels of grounding and current flow and temperature etc... When the new hptuners ECU releases it will have canbus digital widebands to eliminate analog offset. The new era approaches. Better tell them quick that widebands are useless and they need narrowbands.... oh they will laugh at you

Also, it's ironic that I'm being "corrected" by the guy who couldn't even get the formula right for the particular wideband. Why don't you go to that thread and correct things?

I made 2 mistakes on this forum out of all the posts I've ever made, the first was dyslexia at its finest. The second was assuming you knew what an offset term was and that you had embedded it into the ratio term. Basically my mistake was thinking you had some grasp of the analog offset and feeble attempt to correct for it. But the truth was far from that, you didn't even realize you needed one.

[SiriusC1024]

Show me where. Do it. Make the math work or stfu.