First time tuning an E38 from scratch

[mowton]

Isn't that what I'm already doing in my config? On the lower left of the table display I have commanded lambda and lambda error listed above the LC-1 PID. I plot that lambda error in my histograms and use those error values as my factor for editing my VE table.

IIRC, we have been over this LC-1 debate already and came to the conclusion that changing the output values in the LC-1 only works for Innovate's logworks software. You can't alter what the LC-1 tells HPT. If you are running at lambda, HPT is going to display 14.7. You need to make custom PIDs (which I have) to alter that reading into whatever format you choose.

Its been a while and I didn't go back through the thread...sorry. But the LC-1 manual shows the ability to generate the Analog out in Lambda or AFR. You would need to set the HPTuners User defined Wideband (EIO) as Lambda based on the voltage settings from the LC-1. Based on that and your Commanded Lambda PID (Commanded AFR/Stoich AFR) will give you the 2 variables for the error PID.

Ed M

[bikewiz]

I'm going to play around with it tonight and make sure my LC-1 is set up correctly.

[bikewiz]

I've done a ton of reading on the LC-1 setup in HPTuners. From what I gather, there is nothing wrong with the way I have the wideband set up. The default LC-1 setup is in gasoline scale where 1.0 Lambda = 14.7 AFR. The LC-1 EIO displays AFR by default. If you create a custom PID with the LC-1 EIO/14.7 you have a Lambda reading. That is what I have done. Also, my EIO AFR readings match my LC-1 gauge in the car, so no voltage offset is necessary.

This was a frustrating evening of trying to tune SD. I went back a few tunes where I knew I wasn't going lean. I made sure all my MAF DTC's were set properly. I then logged lambda error on my histogram. I saw i was a little lean above 2400 rpm and about 10% rich below 2400. I made adjustments to the VE table in Bluecat and loaded the new equations into a new tune file. I flashed the ECM and fired it up, and I'm still 8-10% rich in the lower RPM range. WTF?

I must be doing something wrong. Overlooking a setting or something other modifier affecting my settings. Scratching my head here and getting frustrated. HELP!

[bikewiz]

Here's one last tune from the evening with a log file.

[mowton]

I've done a ton of reading on the LC-1 setup in HPTuners. From what I gather, there is nothing wrong with the way I have the wideband set up. The default LC-1 setup is in gasoline scale where 1.0 Lambda = 14.7 AFR. The LC-1 EIO displays AFR by default. If you create a custom PID with the LC-1 EIO/14.7 you have a Lambda reading. That is what I have done. Also, my EIO AFR readings match my LC-1 gauge in the car, so no voltage offset is necessary.

This was a frustrating evening of trying to tune SD. I went back a few tunes where I knew I wasn't going lean. I made sure all my MAF DTC's were set properly. I then logged lambda error on my histogram. I saw i was a little lean above 2400 rpm and about 10% rich below 2400. I made adjustments to the VE table in Bluecat and loaded the new equations into a new tune file. I flashed the ECM and fired it up, and I'm still 8-10% rich in the lower RPM range. WTF?

I must be doing something wrong. Overlooking a setting or something other modifier affecting my settings. Scratching my head here and getting frustrated. HELP!

OK Matt, back off the ledge :-), know it can get frustrating at times and information sometimes gets jumbled in the postings.

4 questions/observations....(I think 4 best will solve the problem though)

1) what config are you using know? I seem to be able to see the LC-1 WB (readings based on 14.7 AFR) and a commanded Lambda (based on Cmd AFR Hi res/9.76) but can't seem to find the Error PID.

2) Make sure in all you AFR to Lambda PIDS and Error PID that they are all moving in the right direction....what I mean is if the WB is reading lean than the error s/b 1.XX. If rich than the error s/b 0.XX.

3) You are not entering PE mode.....Some where's between here and your MAF tuning files the PE MAP enable went from 15 kPa @ 30% pedal to 75 kPa @ 30% pedal. Recommend you either change back to the 15 KpA or lower pedal to like 10% (my preference). As a note, in all logs never saw more than like 31 or 32% pedal. As its a transplant, you have verified the pedal to TB plate will go full scale. Not sure if you haven't pushed it yet or I just missed it.

4) Most important, if you are lean than you need to add to the VE (multiply by 1.XX), not pull (multiply by 0.XX)

Ed M

[bikewiz]

OK Matt, back off the ledge :-), know it can get frustrating at times and information sometimes gets jumbled in the postings.

4 questions/observations....(I think 4 best will solve the problem though)

1) what config are you using know? I seem to be able to see the LC-1 WB (readings based on 14.7 AFR) and a commanded Lambda (based on Cmd AFR Hi res/9.76) but can't seem to find the Error PID.

2) Make sure in all you AFR to Lambda PIDS and Error PID that they are all moving in the right direction....what I mean is if the WB is reading lean than the error s/b 1.XX. If rich than the error s/b 0.XX.

3) You are not entering PE mode.....Some where's between here and your MAF tuning files the PE MAP enable went from 15 kPa @ 30% pedal to 75 kPa @ 30% pedal. Recommend you either change back to the 15 KpA or lower pedal to like 10% (my preference). As a note, in all logs never saw more than like 31 or 32% pedal. As its a transplant, you have verified the pedal to TB plate will go full scale. Not sure if you haven't pushed it yet or I just missed it.

4) Most important, if you are lean than you need to add to the VE (multiply by 1.XX), not pull (multiply by 0.XX)

Ed M

Haha, yeah I'm off the ledge. The great thing about tuning at your house is you can turn off the lights, close the door, and walk away. AND then go cry about your problems on HPTuners forum! LOL

1. I'm using the config called Matt Default.cfg. I'll re-attach it to this post. The error PID is on the bottom left above the LC-1 EIO. I know the EIO displays in gasoline scale (14.7), but I don't pay attention to it. I just use the EIO for my PIDs.

2. PIDs are good. You've had me second guessing my PIDs and wideband so much that my head spins.

3. This could be my problem! http://www.camaro5.com/forums/showthread.php?t=196696 < take a look at post #3. A few pictures down he says

Go to “Engine>Fuel>Power Enrich”. Under “Power Enrichment PE Enable”, set “MAP” to 75. This will help keep us out of PE mode so that we can get fuel trim data over a larger range, and is safer than disabling it entirely. Still, keep an eye on knock retard while you're data logging.

4. I may have made a mis-post here or there about pulling/adding in correlation to lean/rich, but I definitely know the difference and how to correct for it. I plot the lambda error VE (Histogram #8 in my config) and if a cell plots 0.90, i know that means i need to multiply the corresponding VE table cell by 0.90, which I loosely refer to as "pulling 10%".

Going to change the PE enable back to 15 and see what happens...

[bikewiz]

Changing PE enable to 15kpa did the trick! Thanks Ed!

Attached a couple logs and tunes. Got the VE lambda error histogram within about 0.95-1.04 with the car running in third gear on my lift. Then i forgot my cooling fans have to be manually turned on right now, so it hit 239*F and i shut it down. While it was cooling down, I adjusted the VE table. Loaded the VE table and ran it WITH FANS this time . Unfortunately, my histogram was now saying I was about 15% rich. Wondering if this was from the coolant temp I hit? Either way I called it a night and happy with the progress made today.

I did notice initial start-up when the engine was cold was lean on the lambda error histogram by as much as 20%. I'll poke around and see if there is a cold start multiplier I don't know about.

[mowton]

How does it start and idle at cold start? How long does it crank before firing up? Does it transition from crank to idle smoothly or dip and recover? The startup airflow is still high from the earlier days of tuning. I would try and back that down again (I think you added 50% on top of the 50% I initially added) first. The logs show a bit low in the transition idle RPM and a dip to 550 RPM or so before getting up to the 950 commanded.

I finally got straightened out on my end relative to your Error PIDS...for some reason on my end they were tagged as unsupported? The only advantage to tailoring the analog output from the LC-1 to the HPtuners software is to preclude having to perform several operations to arrive at the PID values. Each equation you perform to develop a PID which then is used in yet another equation takes up some time and slows down the logging process a bit. It's just cleaner. But for now I am on the same page.

Glad it is starting to fall in line for you.

Ed M

[bikewiz]

Cold start always starts and dies at least once or twice. Then it will fire up and idle pretty well. No throttling of the pedal is needed. Crank time is excessive and has room for improvement. The transition is not smooth, and is probably the cause of any start/stall attempts that occur.

I finally sat down and read this page: http://www.performancetrucks.net/for...lambda-524099/

Believe it or not, it was my 4th timing reading that thread. It finally clicked and I realized what you have been telling my about setting up my wideband for lambda. I swear I'm not usually this dumb! Essentially, the way I had it set up worked too, but as you said the scanner has to do more math and can slow down the data. I changed my whole config now. Sorry for being dense!

I'll go ahead and change the start-up airflow back to what you recommended in the early tunes and see where that takes me. Thanks again Ed!

[mowton]

Cold start always starts and dies at least once or twice. Then it will fire up and idle pretty well. No throttling of the pedal is needed. Crank time is excessive and has room for improvement. The transition is not smooth, and is probably the cause of any start/stall attempts that occur.

I finally sat down and read this page: http://www.performancetrucks.net/for...lambda-524099/

Believe it or not, it was my 4th timing reading that thread. It finally clicked and I realized what you have been telling my about setting up my wideband for lambda. I swear I'm not usually this dumb! Essentially, the way I had it set up worked too, but as you said the scanner has to do more math and can slow down the data. I changed my whole config now. Sorry for being dense!

I'll go ahead and change the start-up airflow back to what you recommended in the early tunes and see where that takes me. Thanks again Ed!

No problem Matt, tuning at times can become very frustrating but the end results are so satisfying. As I stated above, the written word can have multiple meanings to different people and sometimes it is tough to get a common thought across the boards.

Looks like you are on your way. I will check the PIDs tonight and update my files.

As a general tune cleanup, please consider the following modifications:

1) Open loop tables -- except Airflow gain, Injector temp and Air enrichment (set to 1.0 from factory), make sure all tables are stock to start and then using IVT (147 degrees and above) and ECT (176 degress and above) set all cells covered under these parameters to 1.0 in the balance of the tables. The idea here is to allow the enrichment levels to stay active while the car is warming up and then in the normal operating range command 1.0.

2) PE enrichment to 1.0. Will discuss the 15kPa and 30% TPS settings later in the tune process when you get the car onto the street.

3) IAT Spark correction -- put back stock table. cut 122, 131 and 140 degree columns in half. You want some protection should the IAT's rise higher than expected.

4) ECT Spark correction -- cut 230 degree column in half

5) DTC -- set P0300 (misfire) to no report. normally done with a cam

Ed M

[bikewiz]

I had a brain fart and missed the step in that wideband transfer spreadsheet where you change the equation to calculate lambda. I fired the car up today to see if the new config worked and the EIO I created was reading in gasoline AFR.

I went back to the spreadsheet, reworked the lambda formula for the LC1 in HPT, and now it should be good. I didn't go back out to the car to test it, but I'm confident I fixed it. New config attached.

[bikewiz]

No problem Matt, tuning at times can become very frustrating but the end results are so satisfying. As I stated above, the written word can have multiple meanings to different people and sometimes it is tough to get a common thought across the boards.

Looks like you are on your way. I will check the PIDs tonight and update my files.

As a general tune cleanup, please consider the following modifications:

1) Open loop tables -- except Airflow gain, Injector temp and Air enrichment (set to 1.0 from factory), make sure all tables are stock to start and then using IVT (147 degrees and above) and ECT (176 degress and above) set all cells covered under these parameters to 1.0 in the balance of the tables. The idea here is to allow the enrichment levels to stay active while the car is warming up and then in the normal operating range command 1.0.

2) PE enrichment to 1.0. Will discuss the 15kPa and 30% TPS settings later in the tune process when you get the car onto the street.

3) IAT Spark correction -- put back stock table. cut 122, 131 and 140 degree columns in half. You want some protection should the IAT's rise higher than expected.

4) ECT Spark correction -- cut 230 degree column in half

5) DTC -- set P0300 (misfire) to no report. normally done with a cam

Ed M

1. I changed all the open loop tables (except the 3 you told me to leave alone) back to stock. Most were set to 1.0 across the whole table. Inj Temp Gain is not 1.0 across the table from the factory, FYI. I then set IVT to 1.0 from 147+. I don't see an ECT table in the open loop section.
2. Changed PE EQ Ratio (Gas) to 1.0
3. I put back the stock table, but I cut all columns from 86-140 in half. It wouldn't have made sense to cut only those three columns because I would have had more spark removed at 86 than 122.
4. Cut 230 column in half
5. 5. P0300 changed.

Everything look good?

[mowton]

OK, so here is a case where I either type too fast or think too slow :-)...

What I meant about the ECT was based on it being an axis title, not a table, my bad! If you look at the Open Loop>Gas (Gear and P/N) tables you will see ECT is the top axis. So the idea here is to make all cells that fall on or above the IVT @133 and ECT @ 176 intersection to 1.0. These tables you can see the EQ slowly ramping down to 1.0 at that point and the balance of cells are 1.0....good. Do the same for Air Gain and Air Airflow. Again, we want the cold start enrichment to help us start and warmup (even during tuning) and set the operating temperature ranges with the tune. Can go back and tweal cold start area's if startup/warmup issues ocurr with these tables. At least you have a know baseline you are working from.

In another case where I am typing too fast or thinking too slow, I meant to say PE Enrichment Ramp In to 1.0.....sorry again. But the EQ table back to where you had it and make the change to the ramp in cell.

And finally, I guess I am not using the correct stock file. I screwed up the IAT Spark Advance correction table as well (perfect 3 for 3). based on your current file, zero the 86 and 95 degree columns and half the 104 column.........

My apologies again.

Ed M

[bikewiz]

Ed, I'm going to town on the tune now, hopefully I can get some results tonight and update.

Spinner, sent u an e-mail. Not looking to the steal the tune, but would like to have a compare file to possibly learn from. Thanks!

[Spinner]

sent it your way. like i said in the email, i'd copy entire and see how it works for you. you can reverse engineer it at your leisure afterwards from where you started. (you will have to dial in MAF vs STFT anyway) let me know how it works

[bikewiz]

Thanks again dude! I don't want to just copy it, I wan't to learn how to do this stuff the right way. However, having a good tune to compare to really helps.

Ed,

I made the changes you talked about and attached it. Unless I'm dumb, the Open loop> Gas table was already = 1.00 at > IVT 133 & ECT 176 intersection. Take a look and let me know.

[mowton]

Thanks again dude! I don't want to just copy it, I wan't to learn how to do this stuff the right way. However, having a good tune to compare to really helps.

Ed,

I made the changes you talked about and attached it. Unless I'm dumb, the Open loop> Gas table was already = 1.00 at > IVT 133 & ECT 176 intersection. Take a look and let me know.

Yes the Open Loop>Gas table was correct. Identified it to show how it gradually reduced to the 1.0 value at the 133/176 meeting area. Rest of changes look good.

Ed M

[bikewiz]

I tried copy/paste for most of Spinner's tune into another file for my car and loaded it. Granted it was a cold start, but it didn't run well, and it was lean.

I went back to my tune i've been working on and got the car warmed up. Its lean when cold started, whats the best way to richen it up when cold?

I also can't seem to get my fans to turn on when up to temp. They are wired up identical to a 2010 Camaro. I can command low and high speed fans on with the scanner manually and turn them on that way, but they don't turn on automatically.

Still have starting problems. In fact, its gotten worse ever since I pulled the startup airflow a few days ago.

On the log labeled fourth, it ran the best. I did notice the Bank 1 HO2S jumping around. Now, I know that I'm not going into closed loop mode right now with the tune, but just an observation. FWIW, my wideband is installed in the Bank 1 header collector.

[mowton]

I tried copy/paste for most of Spinner's tune into another file for my car and loaded it. Granted it was a cold start, but it didn't run well, and it was lean.

I went back to my tune i've been working on and got the car warmed up. Its lean when cold started, whats the best way to richen it up when cold?

I also can't seem to get my fans to turn on when up to temp. They are wired up identical to a 2010 Camaro. I can command low and high speed fans on with the scanner manually and turn them on that way, but they don't turn on automatically.

Still have starting problems. In fact, its gotten worse ever since I pulled the startup airflow a few days ago.

On the log labeled fourth, it ran the best. I did notice the Bank 1 HO2S jumping around. Now, I know that I'm not going into closed loop mode right now with the tune, but just an observation. FWIW, my wideband is installed in the Bank 1 header collector.

For the starting issue, try increasing Cranking FA by 40%, min idle air by 15% and put back the Startup airflow I had you remove last session.

The jumping around is a good example on how tight the narrowband O2's are around stoich or Lambda 1. As you will see, when your WB reads say .97 or less, the O2's go to >800mv, when the WB reads 1.03 or more, they drop to <100mv. This is basically the same thing that the PCM does in closed loop....varies the fueling up and down around stoich/lambda 1 to maintain proper fueling control. A great reason why you should NEVER tune WOT with the narrow band O2's like some suggest.

Ed M

[bikewiz]

Thanks for the explanation on the O2's. It's funny that I work on GM cars for a living, and given the level of technician I am, you would think I would know all of this. Truth be told, this isn't common day theory that I use in my job. I'm really enjoying that there is more to learn than I ever thought there was. GM teaches us the bare essentials needed to diagnose these cars. They don't go into depth on how the software really works. If they did, I feel like the technicians who took the time to learn the material would be better equipped to diagnose problems.

I'll make your recommended changes and report back. I wasn't opposed to reducing the startup airflow like you suggested before because the lower temp cells in the table were maxed out. When you max out a cell, it kind of makes you feel like you are going about this the wrong way, you know?