Why does Banish recommend tuning MAF before VE?

[futuretech]

Seems contrary to everything I have read here.

[Montecarlodrag]

For every problem you have, there are many solutions.
I have done it both ways and the results were the same (As long as you do it right)

To tune MAF you disable VE... To tune VE you disable MAF, so we can say the order is not important because they are tuned completely independent of each other.


There is no absolute truth, and I'm absolutely certain of it

[02Ls1]

Not to hijack this thread but i have a question.

I have tune VE and MAF open loop with a wideband and left them on open loop.
My question is, After finishing tuning either one can i go back and change it to closed loop?

Example: I tune VE and leave it running with VE only, after finishing tuning it can i go back and be on closed loop and my air fuel will stay the same as it was open loop?

[Z06-Hammer]

Seems contrary to everything I have read here.

I have found a couple of inconsistencies in his video.

For example....

When he programs his Lambda pro into HP Tuners he uses Volts / 4 + 0.55. By using this formula you come up with.

0V = 0.55
1V = 0.8
2V = 1.05
3V = 1.3
4V = 1.55
5V = 1.8

Lambda should have been 1 @ 1.88V but is now 1.02.

Just wondering why did he not use Volts / 4.166 + 0.55?

Also when he tunes the MAF on the Trailblazer he uses a commanded AFR of 14.16. When he sets up his Lambda Error he uses commanded afr / 14.56 which gives you a Lambda difference of .03. So was he tuning with .05 Lambda added?

[Maxwell Power]

I have found a couple of inconsistencies in his video.

For example....

When he programs his Lambda pro into HP Tuners he uses Volts / 4 + 0.55. By using this formula you come up with.

0V = 0.55
1V = 0.8
2V = 1.05
3V = 1.3
4V = 1.55
5V = 1.8

Lambda should have been 1 @ 1.88V but is now 1.02.

Just wondering why did he not use Volts / 4.166 + 0.55?

Also when he tunes the MAF on the Trailblazer he uses a commanded AFR of 14.16. When he sets up his Lambda Error he uses commanded afr / 14.56 which gives you a Lambda difference of .03. So was he tuning with .05 Lambda added?

I just watched both videos this morning. He just mentioned 14.1, he didn't put it in. He left it at 14.56.

I think anyone who watched the video could find a few small errors. The errors could very well have been just in the editing. I don't really think they matter much as the point is made and the lessons are there to be learned.

edit: I misread your post. so the first part of my response doesn't apply.

[Maxwell Power]

one thing that is important to notice is that the ecu targets a lambda of 1.0. So regardless of what you put in that box, lambda is still lambda. The O2 sensors respond the same, the voltage doesn't change, the ecu still has the same lambda target.

Basically what I'm saying is: No. He wasn't tuning with .05 lambda added. The ecu was targeting lambda and his sensor was reading lambda.

The actual AFR value for lambda will change between fillups. This is why he tunes in lambda. Lambda is self correcting. The input value in the hp tuners software does not tell the ecu to command a different lambda. It tells the ecu what AFR is theoretically going to be at lambda = 1.

[Z06-Hammer]

one thing that is important to notice is that the ecu targets a lambda of 1.0. So regardless of what you put in that box, lambda is still lambda. The O2 sensors respond the same, the voltage doesn't change, the ecu still has the same lambda target.

Basically what I'm saying is: No. He wasn't tuning with .05 lambda added. The ecu was targeting lambda and his sensor was reading lambda.

The actual AFR value for lambda will change between fillups. This is why he tunes in lambda. Lambda is self correcting. The input value in the hp tuners software does not tell the ecu to command a different lambda. It tells the ecu what AFR is theoretically going to be at lambda = 1.

Here is my reasoning....

The stoich ratio can be changed in the ECM. He was commanding 14.16 AFR for stoich which will now show as Lambda 1 in the scanner. The same as if he was tuning for E85. He would change the stoich ratio to 9.8

So now [PID.6001] = 14.16 AFR or Lambda 1

He entered in [PID.6001]/14.56 which would make the Lambda error .03 off.

His formula should have been [PID.6001]/14.16

[Maxwell Power]

I understand your reasoning, but I think you're mistaken.

Telling the ECM what stoich is, ISN'T for finding stoich. It's used for calculations on fuel mass required to hit lambda. The car will always seek 1.0 Lambda, it isn't seeking .98. If you change the stoich ratio you just throw off the initial calculation. The ecu will still seek stoich of lambda 1.0.

EDIT:

However, if the stoich input is still for the base calculation for fueling, then I guess I could see that once you calibrated the maf you'd have an error IF you changed the stoich input value afterwards. Otherwise, it's relatively moot, because you programmed the MAF for those values. I would also suppose that since fuel quality and ethanol percentage change from station to station, it would be irrelevant to worry about .3 AFR in the stoich table because it's going to change every time you put gas in it anyway.

[Z06-Hammer]

Found this thread that Greg responded in. Finally puts my mind to rest.

http://www.hptuners.com/forum/showthread.php?t=26662

[eficalibrator]

Guys, can I take a week off and go waterskiing without missing drama? Sheesh!

1) I don't care if you do MAF or VE first. I care that you do both. Independently. They should come up with the same g/cyl numbers when you're done. I personally prefer to do MAF first because it's easier and gets the car up and running quicker. You can also use the MAF to calculate VE or GMVE if you have good data and test conditions.

2) I used my ECM Lambda PRO (ECM website link here) in the DVD. Click that link and see that the 0-5v range is 0.55 to 1.75 Lambda. This means my equation constants should be (V/4.166)+0.55 if you do the math. I'll have to check the custom PID to see if it got changed when I was working on another project with a different client in between. (I often end up using different widebands on different projects) Keep in mind that there were gaps between when I shot the on-dyno work and when we shot the screen setups for the video. This stuff doesn't come together overnight, so I occasionally get sidetracked. Poop occasionally happens, and you need to be diligent when you're setting things up on your own.

3) I don't care about AFR for anything other than setting the stoich point in the PCM once in the beginning. The thread above demonstrates that pretty clearly. I only adjust it to match the ethanol percentage in the fuel that I'm tuning that day. It's not always E0 or E10.

[DSteck]

Guys, can I take a week off and go waterskiing without missing drama? Sheesh!

1) I don't care if you do MAF or VE first. I care that you do both. Independently. They should come up with the same g/cyl numbers when you're done. I personally prefer to do MAF first because it's easier and gets the car up and running quicker. You can also use the MAF to calculate VE or GMVE if you have good data and test conditions.

2) I used my ECM Lambda PRO (ECM website link here) in the DVD. Click that link and see that the 0-5v range is 0.55 to 1.75 Lambda. This means my equation constants of (V/4.166)+0.55 are right if you do the math.

3) I don't care about AFR for anything other than setting the stoich point in the PCM once in the beginning. The thread above demonstrates that pretty clearly. I only adjust it to match the ethanol percentage in the fuel that I'm tuning that day. It's not always E0 or E10.

Bingo. Dialing in a MAF table is a lot quicker and easier than a SD table, and gives you the opportunity to just get the car going and back-calculate an SD table based off of injector flow rate, injector pulsewidth, and air/fuel ratio.

[Cantalope Kid]

back-calculate an SD table based off of injector flow rate, injector pulsewidth, and air/fuel ratio.

Never thought of doing that... How would one do so (link)?

[redhardsupra]

http://redhardsupra.blogspot.com/200...ss-models.html that's the base of how I do it. I have since completely 'mechanized' the process with custom software, but it's still the same principles.

Since both models are built from the same data (fuel consumption), I calculate them both simultaneously. Sometimes, to verify correctness, if I've gathered enough data, I make multiple set of samples, create tables from one of the subsets, see if the results change when using different samples. Another trick is 'projecting' one model's results onto another model's space. If i have a car with a well tuned MAF-only setup, I calculate the VE table from it. Or vice-versa, it's all equivalent.

[DSteck]

Mine does it all automatically using the scanner to calculate what can be considered an AFR error for the SD table. It acts the same as tuning in SD mode and using lambda error. It's amazing how easy it is to develop a base SD table.

[Justinmofo]

Hey DSteck, do you have a configuration for that? I dont understand how you set that up.

[DSteck]

This is one I'm not sharing. I imagine redhardsupra will... But I won't.

[eficalibrator]

You can also use the MAF to calculate VE or GMVE if you have good data and test conditions.

Bingo. Dialing in a MAF table is a lot quicker and easier than a SD table, and gives you the opportunity to just get the car going and back-calculate an SD table based off of injector flow rate, injector pulsewidth, and air/fuel ratio.

Never thought of doing that... How would one do so (link)?

Mine does it all automatically using the scanner to calculate what can be considered an AFR error for the SD table. It acts the same as tuning in SD mode and using lambda error. It's amazing how easy it is to develop a base SD table.

This is one I'm not sharing. I imagine redhardsupra will... But I won't.

Hrmm... If only there was some experienced trainer that would teach people how fuel injection works, write a couple books, make a DVD or two, and also teach classes on exactly this subject...

[DSteck]

Hrmm... If only there was some experienced trainer that would teach people how fuel injection works, write a couple books, make a DVD or two, and also teach classes on exactly this subject...

Does not exist.

Instead, just get hammered drunk and punch some custom PIDs in. Sooner or later, probability will win out and you'll type the right things.

[eficalibrator]

I like the way you think, Dave. First round's on me, but we start with Jack Daniels.

[Justinmofo]

Got the books, watched the DVDs. Just thought this site was for people to help others out. But I understand not sharing info, don't want anything to be easy.
Just thought you could point me in the right direction.