HP Tuner Math Channel Help

[acammer]

So, I agree with pretty much everything everyone said here. The calculated AFR is probably not going to match the actual AFR unless every variable is perfectly accounted for, and the engine perfectly consumes the delivered fuel under all conditions. In other words, it's not gonna happen.

I also feel you on wanting someone to actually test it, and prove it out to you. So, I did. Please see the screenshots below. This is a car I was doing some tuning on that has a good wideband in it. I created your Math channel with Mass Airflow divided by Instantaneous Fuel Flow Estimate, and because this car has ID1300x injectors, the entire tune is scaled in half, so I divided the result by 2 in order to make it come out right. The bottom-most row of graphing compares the the calculated AFR (white) against the wideband (red).

The results are about what I would expect. It trends the same as the actual AFR in most places, but they separate, sometimes significantly, in a few places. This is likely due to several of the factors mentioned in the posts above - the fuel flow is estimated, the airflow is a calibrated amount subject to inaccuracies, etc. I would not consider this approach a very useful tuning tool, narrow-band O2 sensors will give you a ballpark to shoot for in the absence of an actual wideband O2 sensor. If it's anything with boost on board, a wideband should be considered a requirement before any tuning occurs.

I like the idea though. And I love playing with Math channels - you can absolutely do some unique things there to get views into various pieces of engine/vehicle operation.

[gtstorey]

Originally Posted by Dark Sun

And I understand what you are saying and I’ve been studying this ECM along with other tune files so I can get an understanding of it all. That’s why I asked if someone would be willing to test out this math channel and compare it with there wide band reading to see if it’s even accurate. But all I’ve gotten is lip service on how no one’s going to try it, how it’s a guess and the sensors aren’t accurate, and that the ECM just uses a set of variables in open loop WOT to basically guess at how much fuel to use. That ECM knows exactly how much fuel it uses with the airflow to maintain the correct AFR. Yes most cars don’t have a fuel pressure sensor, but I’m not work on one of those and don’t care about that as it’s not something I’m concerned with at this time. If I get a Pontiac 6000 SE, then maybe I’ll look at that configuration.

I’m not sure why you think you are smarter than everyone else, but I was trying to help. But after this second round of I’m smarter than everyone else, I don’t care if you try using the smell of the exhaust to figure out your AFR.

[Little Walter]

Doesn't the ECM get it's AFR info from the O2 Sensors and MAF sensor while in the close loop mode? Also on the Gen5 Zl's do they go into open loop mode at WOT? Thanks in advance for your answer.

[gtstorey]

Originally Posted by Little Walter

Doesn't the ECM get it's AFR info from the O2 Sensors and MAF sensor while in the close loop mode? Also on the Gen5 Zl's do they go into open loop mode at WOT? Thanks in advance for your answer.

Yes. Of course even closed loop on narrow band o2 sensors the ECM is still to some extent guessing. That is why it drives it from slightly rich to slightly lean when in closed loop. It only knows rich or lean, not an exact number or ratio.

Except it doesn’t use MAF for AFR. It uses a comparison of MAF to VE (actually on these I think VVE) to calculate how much fuel it thinks it needs, turns on the injectors for a length of time it thinks it needs, and then compares it to whether the narrowband say it was rich or lean. If it doesn’t get it correct, it adjusts things to “trim” it back to correct.

This is very much a simplified explanation. I could go somewhat deeper, but my knowledge doesn’t cover all of the details of all of the variables it uses in these calculations and table look ups.

[Shodown]

Bro lol. You bought a $35-$40k car, $550 for HPT plus a laptop, but want to skimp on a $200 wideband? For real?

Also the LSA has been around for 15 years. There's nothing new to it. Playing with bypass springs pretending to increase boost and trying to trick the ECM into reading accurate AFRs may be fun and all, but the recipes for reliable power are out there...swap pullies, get bigger injectors/fuel pump and tune appropriately (with a wideband). It's very simple. No need to reinvent the wheel.

[Little Walter]

Originally Posted by gtstorey

Yes. Of course even closed loop on narrow band o2 sensors the ECM is still to some extent guessing. That is why it drives it from slightly rich to slightly lean when in closed loop. It only knows rich or lean, not an exact number or ratio.

Except it doesn’t use MAF for AFR. It uses a comparison of MAF to VE (actually on these I think VVE) to calculate how much fuel it thinks it needs, turns on the injectors for a length of time it thinks it needs, and then compares it to whether the narrowband say it was rich or lean. If it doesn’t get it correct, it adjusts things to “trim” it back to correct.

This is very much a simplified explanation. I could go somewhat deeper, but my knowledge doesn’t cover all of the details of all of the variables it uses in these calculations and table look ups.

Thanks a million.

[1JEWLDSSRS]

Originally Posted by Shodown

Bro lol. You bought a $35-$40k car, $550 for HPT plus a laptop, but want to skimp on a $200 wideband? For real?

Also the LSA has been around for 15 years. There's nothing new to it. Playing with bypass springs pretending to increase boost and trying to trick the ECM into reading accurate AFRs may be fun and all, but the recipes for reliable power are out there...swap pullies, get bigger injectors/fuel pump and tune appropriately (with a wideband). It's very simple. No need to reinvent the wheel.

lol. Didn't you guys try to tell OP this last time?? In another thread?? It's so easy...Wow man..

[Shodown]

Originally Posted by 1JEWLDSSRS

lol. Didn't you guys try to tell OP this last time?? In another thread?? It's so easy...Wow man..

Yes.

[Dark Sun]

Originally Posted by gtstorey

Yes. Of course even closed loop on narrow band o2 sensors the ECM is still to some extent guessing. That is why it drives it from slightly rich to slightly lean when in closed loop. It only knows rich or lean, not an exact number or ratio.

Except it doesn’t use MAF for AFR. It uses a comparison of MAF to VE (actually on these I think VVE) to calculate how much fuel it thinks it needs, turns on the injectors for a length of time it thinks it needs, and then compares it to whether the narrowband say it was rich or lean. If it doesn’t get it correct, it adjusts things to “trim” it back to correct.

This is very much a simplified explanation. I could go somewhat deeper, but my knowledge doesn’t cover all of the details of all of the variables it uses in these calculations and table look ups.

And in WOT and PE in open loop, it’s guessing at what to do for AFR as it doesn’t have a wide band to read, so how does it KNOW if it’s lean or not?

[gtstorey]

I'll try one more time even though I said I would pull out. It doesn't know in PE what the AFR is. It is just supplying the amount of fuel that was programmed in by the designers based upon what they expect it to need with a lot of table lookups, with an adjustment carried over by the LTFT from running in closed loop. This is why it has to have a wideband when running it outside of stock parameters. You have to adjust the model to now tell it how much fuel it needs at these new parameters.

I'm sure the original settings for WOT/PE where done with some combination of computer modeling and by using some version of an o2 analyzer, probably a lab grade 5 gas analyzer.

All of this is really the difference between it running in open loop vs closed. Closed loop means it has something to adjust by and tweaks things based upon AFR. Open loop is just following the programming. Reads pressure, rpm, temps etc, etc, etc and says it needs this much fuel.

[Dark Sun]

Originally Posted by Shodown

Bro lol. You bought a $35-$40k car, $550 for HPT plus a laptop, but want to skimp on a $200 wideband? For real?

Also the LSA has been around for 15 years. There's nothing new to it. Playing with bypass springs pretending to increase boost and trying to trick the ECM into reading accurate AFRs may be fun and all, but the recipes for reliable power are out there...swap pullies, get bigger injectors/fuel pump and tune appropriately (with a wideband). It's very simple. No need to reinvent the wheel.

So how much PSI are you at with all the mods you listed? I know there’s recipes out there and everyone is a cook, but nothing tastes the same when all the cooks follow the same recipe.
So as I stated before that you mentioned, I raised the boost with my first trick from stock to 9.25 psi and everyone is scolding my recipe, I did some research and math, and put a 10% lower pulley on two days ago, now I’m making 13.2 psi hitting the torque management boost disable, injectors are 112% and my cat temps are good. My MAF is just a few percent off at WOT, my calculated AFR for my stock ECM is within 1% of commanded lambda AND the PE table. I am getting a wide band soon, injectors, cam, springs, etc. So take your standard chicken soup recipe like the other 10 thousand have done, I personally don’t follow others with anything and revel with others telling me it can’t be done, you are wrong, or it’s impossible, then doing it. Can you personally design parts in AutoCad, have them made, cut weld and fabricate it all in stainless, build a rear differential, transmission, a twin turbo draw through carburetor system on a LS engine making 20 psi on stock pistons, get that ground up design and build to pass DOT regulations, get it titled, registered, and tagged without ANY help from others because there is no RECIPE for it and run it to 219 mph? Everyone said I couldn’t use hydraulic steering on a motorcycle, did it, everyone said I couldn’t use that transmission and differential setup, did it, everyone said the turbo system wouldn’t work, did it, everyone said you can use a single sided hub centric front wheel on swing arms for a motorcycle, did it. Do you know the exact PSI factory GM hypereutectic pistons will break ring lands no matter the AFR and .060” ring gaps? I do. Thanks for input.

[gtstorey]

Sure I can design my own parts and have them made by someone else. I could do that 30 years ago (when I was a Rocket Scientist, or at least an Engineering co-op with NASA). That has nothing to do with knowing how a GM ECM setup is done. You may know how you think it should be done, but that doesn't mean they did it that way.

Of course your commanded and calculated are following extremely close. It's calculating how long to turn on the injectors from cylinder volume and the commanded AFR. You are then calculating AFR from the same airflow and fuel flow parameters. The only reason it is different is a time lag.

[Dark Sun]

Originally Posted by gtstorey

I'll try one more time even though I said I would pull out. It doesn't know in PE what the AFR is. It is just supplying the amount of fuel that was programmed in by the designers based upon what they expect it to need with a lot of table lookups, with an adjustment carried over by the LTFT from running in closed loop. This is why it has to have a wideband when running it outside of stock parameters. You have to adjust the model to now tell it how much fuel it needs at these new parameters.

I'm sure the original settings for WOT/PE where done with some combination of computer modeling and by using some version of an o2 analyzer, probably a lab grade 5 gas analyzer.

All of this is really the difference between it running in open loop vs closed. Closed loop means it has something to adjust by and tweaks things based upon AFR. Open loop is just following the programming. Reads pressure, rpm, temps etc, etc, etc and says it needs this much fuel.

And I understand this exactly, you stated what I’m looking at in this math channel I’m working on. It’s programmed with a known amount of air interring the engine, it injects a calculated amount of fuel to create a AFR without a wide band sensor. As I said in another forum, in the 1940’s they controlled AFR on 3000-4000 hp aircraft engines with both mechanical and exhaust superchargers on the same engine with pencils, paper, gauges and levers. The fastest aircraft engine ever in the SR-71 used gauges and levers so the pilot could adjust the AFR using airflow, fuel pressure, and EGT. I’m getting the wide band and some other things soon. I just for the life of me don’t understand why everyone knocks and beats on someone asking questions. I simply asked if anyone would compare the channel to a wide band reading. A few have and there cars are already modded showing the math reads lean in areas on there comparison. So I will do some tweaking on it. I know that it’s not going to ever be “perfect” as the math has no way to “see” the unburned fuel trapped between the piston and cylinder walls, but that can also be calculated to a degree. I mean really, drag cars in the 60’-80’s used EGT as a standard measure of AFR and so did every piston engine aircraft from the 40’s till the 90’s with carburetors. Some even had up to 30 psi boost and 20:1 AFRs. It’s not rocket science, it’s simple math. Thanks for the information.

[Dark Sun]

Originally Posted by gtstorey

Sure I can design my own parts and have them made by someone else. I could do that 30 years ago (when I was a Rocket Scientist, or at least an Engineering co-op with NASA). That has nothing to do with knowing how a GM ECM setup is done. You may know how you think it should be done, but that doesn't mean they did it that way.

Of course your commanded and calculated are following extremely close. It's calculating how long to turn on the injectors from cylinder volume and the commanded AFR. You are then calculating AFR from the same airflow and fuel flow parameters. The only reason it is different is a time lag.

See that’s a great response with useful information. I know it’s not perfect and was asking for the comparison so I can understand and tune the math. It was my first attempt and math channel and I know it needs some work. As far as understanding the ECM, I’m studying every aspect of it and how every table cell by cell is related and calculated for AFR, MAF, injector timing, spark, knock, torque management, TCM commands, everything. But I still believe that the AFR can be accurately measured with the math to a very close percentage if I continue to work on it.