AFR error PID setup guide

[c.u]

For the hard shift,Applie the TCS patch.

[jim2092]

12.5/14.7=.85.... 1/.85=1.176

True for regular gas and 1.136 for E10?

[Kyne]

True for regular gas and 1.136 for E10?

Ugh, this is why you need to learn lambda, equivalence and lambda are the same for all fuels. .88L is the same no matter what. You don't shoot for 12.5 for e10 you technically shoot for 11.99 at the same lambda. So 11.99/14.1=.85 1/.85= 1.176. Which would be the same fueling.

[jim2092]

As it turns out I do tune in lambda without even converting it from AFR. I think you misspoke when you said "lambda is the same for all fuels". Obviously that is not true. But interesting that you see .85 lambda as "the" number for WOT PE and I have been thinking 12.5 AFR is "the" number under those conditions and using a lambda PE number to attain that result for E10 fuel. Not a big difference of course but interesting none the less.

[Kyne]

Lambda is the same for all fuels, 1.00 is stoich for e85, e10, e0, methanol, everything. It's kind of the definition of lambda. I just used .85 for an example, that's a good starting point for NA but I end up usually in the .87-.88 range at the top end. The EQ is just a ratio of enrichment, so you add 17.6% fuel no matter what the stoich to get .85L.

[jim2092]

I think we are getting into the area of semantics, but the lambda value of 1.0 is attained at a different ratio of air to fuel for each type of fuel since lambda is defined as the ratio of air to fuel that results in complete combustion. And of course that is what I meant by saying that lambda is different for different fuels. Hopefully on this we can agree and I will agree that 1.0 = 1.0 = 1.0.

[Kyne]

Lambda is the same, stoichometry is different.

[2004gtols1]

Except when the maf is failed it runs off of low. And I've never had the shifting problem that I've noticed, have you done the traction control fix?

That is what I was thinking it did just like the gen 3 setups but wasn't positive. I already made both tables match...

For the hard shift,Applie the TCS patch.

I seen the TCS patch since a few weeks ago when I tried starting his base tune, at the time it wouldn't even start with the MAF failed.

I followed the how to sticky to SD tune a gen4 and followed the write to the T and the car wouldn't even start lol. That write up is a pile of shit.....had to go back and undo everything it said to do pretty much.

I'll apply the TCS patch and do a full write and see how the shifting feels then.

[2004gtols1]

I agree you should switch to Lambda but the idea of setting up is the same idea. First set up the error PID, I use my wideband in lambda divided by the inverse of thee equivalence ratio which is commanded lambda on the E40. That will give you a ratio, to turn it into a percent you can use in past multiply it by 100 and subtract 100. Here is my function:



You can also do the normal way of ((actual-theoretical)/theoretical)*100

This is to show you where the settings are, click the folder to the right of 12, where it says histogram settings.


This is my MAF low histo, you can see I am plotting my Lambda Error on the table and I have it filtered so that its only lower than 5800Hz.


Finally here is the table setup, just copy the axis from the calibration table and remove the commas.


Same idea as this to set up a VE table. Also you will have to change in the calibration from MAF to VE but I just use the scanner to force open loop.

Was about to set this up but just realized no matter how far I zoom in I can't quite read what you put in for the function on the lambda error PID setup.

I went to the photo bucket link for the pic and it is the same quality. It's just not clear enough in the screen shots. Looks like it says AUX20122 but can't really tell what symbols are around it and that kinda needs to be right for it to work properly haha. I can see the others just not that first screen shot for some reason.

Didn't really think of this till now but if you posted the config file wouldn't I be able to just save and open it up in the scanner and tune away?

As I stated before is that I'm using a AEM WB I just set it up with the pre determined settings that HPT provides for the EIO input. Is that going to work okay for tuning with lambda instead of AFR?

Also do you have all those histos run at the same time? Thought if you have too many things running at once it skews the log sum and makes it slower to react etc.

Thanks again for the help and going through the trouble of making the screen shots and what not.

[Kyne]

As long as you are under 24 things in the table it will run fast. I'm not going to post the config file because you need to learn how to make the config. The AEM can be piped in directly as lambda. Just find a wideband transfer function spreadsheet. Click on the help button just above the function string, it will bring up the pid names so you can choose, [AUX.20122] is my wideband you have to find yours. You'll need your wideband and AFRhires or EQ also. The pids have brackets and I uses parenthesis like I did before to show you what the function is. Mine is ([AUX.20122]/(1/[PID.68]))*100-100 aux is wideband and pid68 is EQ. I need 1/EQ to get lambda you'll just use hires.

[2004gtols1]

Here's the config I made to try out using afr error.