Constantly Moving Spiky VE Table

I'm fairly new to this but have read and learned a lot. I have recently changed to a boost reference fuel pressure regulator and changed out the injectors to a split pattern (still 60lbs) and I'm trying to get a good tune. I've set up the tune file for VE scanning (disabling MAF, etc.) then run a scan and the errors are all over the place. Really high in one cell and very low in the next cell. So I multiply the data back in by half, I run it again and it still wants more, and run it again and now it's way too much. Back and forth I go chasing my tail. The VE table is has some terrible spikes in it now after starting with a fairly smooth table. I just can't seem to make it happy, especially low RPM/low MAP.

Using a AEM wideband via serial.

Attachment 99567

Attachment 99568

I'm new to this as well but maybe look into these: Fail MAF in Engine Diag, Fail DFCO RPM, also with having some knock in your log file subtract 5 degrees out of the spark advance tables for now.

This guy does a great job at explaining the process. https://youtu.be/rmzpQBnTJ9w

Quote:

---

Originally Posted by KLew75

I'm new to this as well but maybe look into these: Fail MAF in Engine Diag, Fail DFCO RPM, also with having some knock in your log file subtract 5 degrees out of the spark advance tables for now.

This guy does a great job at explaining the process. https://youtu.be/rmzpQBnTJ9w

---

I appreciate your response. I've watched about every video he has and followed many of them to the letter. I have the MAF set to fail in Engine Diag, did his suggestions on the DFCO, and have no recorded knocking. I can't help to wonder if the boost reference is causing me an issue or something with the new injectors. Every scan shows that it's terribly lean at 400-800 RPM no matter how much I add. It's almost if the DFCO is still working even though I set the temp to 400.

sounds like injector data fight or transient fueling is causing problems.
try to tune in steady state if possible.

Sorry but you are showing knock retard in the log file. The worst is 5.2 degrees at 1:45:16.551. I'm wondering if this is due to Burst Knock kicking in

Are you turning off your vehicle before stopping the scanner? I noticed today when I log and shut down the vehicle while it's still scanning I get terribly lean areas down low. Now I just ensure to stop scanning before shutting down.

One thing to check out in the file is dynamic airflow. If I'm reading that correctly the High RPM re-enable needs to be bumped to something like 7900 rather than 90 as it reads, RPM must drop below this to re-enable airmass prediction calculations.

Please let us know if you get the overall issue figured out. Good luck!

Quote:

---

Originally Posted by cobaltssoverbooster

sounds like injector data fight or transient fueling is causing problems.
try to tune in steady state if possible.

---

I'm not sure what you mean by injector data fight or transient fueling. What is that and what would be causing it.

In one of the other posts, they noticed that I hadn't failed the MAF to disable it, and they are correct, but I had in earlier tunes and was still having much of the same issue. I don't know if the boost referenced fuel system is making it difficult. I would have assumed it would have made things easier. There isn't much information on how to properly tune with a BRFS. I may try opening the port to the regulator and plugging the vacuum line to the intake and see if tuning is easier, but that won't be a good tune to use in the end.

Quote:

---

Originally Posted by KLew75

Sorry but you are showing knock retard in the log file. The worst is 5.2 degrees at 1:45:16.551. I'm wondering if this is due to Burst Knock kicking in

Are you turning off your vehicle before stopping the scanner? I noticed today when I log and shut down the vehicle while it's still scanning I get terribly lean areas down low. Now I just ensure to stop scanning before shutting down.

One thing to check out in the file is dynamic airflow. If I'm reading that correctly the High RPM re-enable needs to be bumped to something like 7900 rather than 90 as it reads, RPM must drop below this to re-enable airmass prediction calculations.

Please let us know if you get the overall issue figured out. Good luck!

---

Good catch on the KR. Knowing the drive I took and where that is at, that was coming off the highway and turning into the driveway, the AFR at decel was pegged at 18 lean and as I turned in, it bogged and probably caused the knock. I will keep a better eye on that.

As far as the dynamic airflow, the wording is very confusing on how it works. So I am open to the idea of changing it to 7900 instead of 90. I thought we didn't want it to do airmass prediction (I'm thinking MAF). So the RPM would have to drop below 90 in order to re-enable the MAF???? Am I thinking of the wrong thing here?

My understanding is the High RPM Disable is for MAF and the Re-enable is for VE. If you look at Goat Rope Garage he has the Re-enable set 100 lower than his Disable so 7900 should be plenty high to keep you enabled in VE.

Quote:

---

Originally Posted by KLew75

My understanding is the High RPM Disable is for MAF and the Re-enable is for VE. If you look at Goat Rope Garage he has the Re-enable set 100 lower than his Disable so 7900 should be plenty high to keep you enabled in VE.

---

I made all these changes and I still get AFRs all over the place. I'm really thinking I have something else wrong outside of the tuning. I can hold the throttle at a fast idle, say 1500 RPM, and watch the AFR jump all over without even moving off the cell. How the hell do you tune when it's changing constantly within a cell? I'm going to disable the BRFS and see if that has any effect and then I'm going to switch back to my previous injectors. I went from a cone spray 60lbs to a split pattern 60lbs. I'm wondering if the injectors are not matched well or something.

Blows my mind how many different setups there are out there for doing VE tuning for the Gen 3 and people get good results but I try everything and can't even get one cell dialed in. There has to be something else going one.

I also noticed that even though I set the fail frequency to zero, my scan still picks up MAF data! I finally unplugged the MAF and that killed the data but also the IAT1 and it ran so bad I couldn't even attempt to tune it.

Even with the MAF "failed" it will still read. The computer shouldn't be making fueling adjustments off the MAF at that point. Could you post the new tune and log?

transient zones are the areas where data is changing rapidly. (ex: right when you depress the throttle, right when you let off the throttle)
you can filter this data out of your histograms by running filter limits on the derivative of the signal you wish to filter on. for this purpose its usually the map sensor or a load pid getting filtered. the filter function "slope" is how you do this.

Quote:

---

Originally Posted by cobaltssoverbooster

transient zones are the areas where data is changing rapidly. (ex: right when you depress the throttle, right when you let off the throttle)
you can filter this data out of your histograms by running filter limits on the derivative of the signal you wish to filter on. for this purpose its usually the map sensor or a load pid getting filtered. the filter function "slope" is how you do this.

---

That's very interesting. I will have to read up on that.

I made great progress today. I couldn't help to notice that my fuel pump was squealing a lot during the initial start up. I kept thinking I need to look into that some time and figure out what is going on. Well as turns out, I was running with a pressure of 90psi!
This is why the pump was making so much noise. I have a fuel pressure gauge mounted on my regulator, that's how I noticed. It wasn't this way when I adjusted the regulator right after install. I disconnected the return line just before going back to the fuel pump and directed it to a gas can and tested the pressure - 30psi (after loosening the adjustment). I pulled the fuel pump and noticed I could not blow in the return line. There is a small screen filter just before the return line dumps back into the basket of the pump. The screen was 90% clogged. This is a new fuel pump and only has a few hours on it. I suspect all the new fuel lines, regulator and filters must have had debris in them and was filtered out by the screen.

After cleaning the screen and adjusting the pressure back to 4 bar, the car immediately started running better. I was able to log some files and modify my VE table. I still have a few spikey areas, but not nearly as bad as before. I just assumed the table would naturally come out mostly smooth.

I'll do long drives and use that data with some small filtering to hopefully really dial it in. I just hope I didn't get a lot of the debris in the injectors and they are not spraying a good pattern or matched very well anymore.

That makes me feel better. Here is a current tune-in-progress. Check out the VE table. Does this look normal (in the cruising area)?

Attachment 99716

By the way, this is the vehicle I'm working on. These are photos from last year and they don't have everything installed yet.

Attachment 99717

Attachment 99718