Tip in throttle causes lean fuel spike and results in heavy knock

[pghalcrow]

I'm chasing an issue that I can't seem to resolve. I've been messing with the transient fuel table to mitigate most of the issue, but It's mostly guess work. I've seemed to resolve this issue the majority of the time, but I'm still getting heavy knock when I go hard on the throttle, usually enough to cause a downshift. The symptoms show as a lean fuel condition resulting in heavy knock.

E38 ECU, 6.0L with LSA 22% overdriven, 6L90.

I'll attach the log and tune. Issue begins at 10:38:55

20230317_3.hpl
adjusted transient fuel tables.hpt

[blindsquirrel]

Rect port heads or cathedral with adapter plates?

[blindsquirrel]

Injector data is wrong - flow rate table is flat, offset table is sloped. No way to know which, if either, is the right way. What fuel system, what regulator type, what injectors, and where'd the data come from?

[pghalcrow]

For the fuel system, I have a 535 LPH pump on 850cc FIC injectors. The table was originally copied from their website, but I'm not sure if the guy who tuned this the first time messed with this table. Fuel type is only 91 pump on a return style regulator. Im also using a VaporWorx pump controller that is set up to increase fuel pressure 1:1 based on the MAP signal.

Also these are LS3 style heads.

[GHuggins]

Don't change the transient settings anywhere near where the engine idles at - will cause false fuel corrections to the MAF
You need to tune in the VE table - most transient fueling starts here
Max map needs fixing
TM in the trans can cause it's own issues with this, so keep that in mind.

If you can post a more focused log - where you just have one pull that shows the problem - in this log you're leaning out bad, but not because of transient fueling - your MAF does have a weird setup, but it's wanting more fuel - sure you don't have a supply problem?

[GHuggins]

As Blind pointed out you need to fix the inj data for the setup - should be a flat setting for both the tables he mentioned.

[pghalcrow]

Let me add that the MAF curve looks a bit weird on top because I never get that far up it, so I never smoothed it out. I've done about 6 runs on speed density to dial in the VE table, but I'm guessing that might not have been enough. The final run I did have most values withing a few percent, so I just stopped there.

I don't have a way to measure fuel pressure right now, so that is an unknown. Fuel delivery issues are not out of the question, but I am running 535 LPH on 850cc injectors on -6AN fuel lines.

[pghalcrow]

I have 5 flow rate tables for the injectors - not sure what needs to be modified and how (like, how does it need to be modified to not be flat). I'm a little over my head here and I might just need a better tuner to clean this up. Was hoping to get a little insight here and learn how to fix this. I admit I'm very new to tuning.

[pghalcrow]

Here's another log where it went wacky at 8:11:17. I'd like to avoid trying to recreate this too many times so I don't smoke my rings (again).

20230316_1.hpl

[blindsquirrel]

If you have a 1:1 referenced regulator, and 58psi base pressure (key-on, or running with vacuum hose disconnected/plugged), then you'd use the injector data for the 400kPa delta column, in every column in both Flow Rate (12307) & Offset vs Press vs IGNV (12308).

If your base pressure is something other than 58, then use whatever column matches (43psi = 300kPa, 70psi = 480kPa, etc.).

If you do not have a 1:1 regulator ('returnless'/non-referenced), then both IFR & Offset should use the data as-is from the datasheet. i.e. neither one should be flat all the way across.

[Alvin]

IMHO the transient stuff is not where you need to be looking.

The VVE table looks like it's missing the LSA HUMP at 100+ KPA 0- to say 2500 RPM.

[pghalcrow]

If you have a 1:1 referenced regulator, and 58psi base pressure (key-on, or running with vacuum hose disconnected/plugged), then you'd use the injector data for the 400kPa delta column, in every column in both Flow Rate (12307) & Offset vs Press vs IGNV (12308).

If your base pressure is something other than 58, then use whatever column matches (43psi = 300kPa, 70psi = 480kPa, etc.).

If you do not have a 1:1 regulator ('returnless'/non-referenced), then both IFR & Offset should use the data as-is from the datasheet. i.e. neither one should be flat all the way across.

Good information here. I currently have a non-referenced regulator, but I have a VaporWorx controller that ramps fuel pressure 1:1 from 43 psi idle to 58 psi (100% duty cycle) using the MAP sensor as the referenced input. It's supposed to act like a vacuum reference, but uses PWM to control the fuel pump output. If I'm reading you right, I should be take the data from the 308 kPa column on table 12308 and copy/paste it all the way across the remainder of the table. Since table 12307 is flat all the way across, I won't touch it. I'm beginning to question my choice to go that route and maybe I should ditch the controller and go with a vacuum referenced regulator instead. The approach you mentioned above would still apply.

[pghalcrow]

IMHO the transient stuff is not where you need to be looking.

The VVE table looks like it's missing the LSA HUMP at 100+ KPA 0- to say 2500 RPM.

Ok, thanks. I'll log again in speed density to capture this discrepancy in the EQ ratio error data for this range.

[GHuggins]

Make sure to change your axis on the left so you're seeing the whole VE table in boost. You'll be able to see how it is after that. I still think you have something else going on as you're still leaning out 10 or more percent in the upper range of the MAF curve. I don't see a lean tip in - just lean continuing into boost in the logs you posted.

[blindsquirrel]

Good information here. I currently have a non-referenced regulator, but I have a VaporWorx controller that ramps fuel pressure 1:1 from 43 psi idle to 58 psi (100% duty cycle) using the MAP sensor as the referenced input. It's supposed to act like a vacuum reference, but uses PWM to control the fuel pump output. If I'm reading you right, I should be take the data from the 308 kPa column on table 12308 and copy/paste it all the way across the remainder of the table. Since table 12307 is flat all the way across, I won't touch it. I'm beginning to question my choice to go that route and maybe I should ditch the controller and go with a vacuum referenced regulator instead. The approach you mentioned above would still apply.

So it's not actually 1:1, it only does a linear ramp between a low pressure and a high pressure? If that's what it does, I don't think there is a way to set up the injector data for that.

The PressureWorx documentation says it's programmed for specific MAP sensors. The LSA MAP sensor has a full range of 43psi from min to max, but 43 to 58 is only 15psi. That's not 1:1.

[blindsquirrel]

Make sure to change your axis on the left so you're seeing the whole VE table in boost. You'll be able to see how it is after that. I still think you have something else going on as you're still leaning out 10 or more percent in the upper range of the MAF curve. I don't see a lean tip in - just lean continuing into boost in the logs you posted.

Duty cycle says it's asking for the fuel, it's just not getting there.

Attachment 129501

Or is the B2 PW not being logged throwing off the duty cycle math channel?

[GHuggins]

I would think with the pressure works starting at 43 PSI and keeping it there and then raising as it got into boost that you would want a linear flow rate table. Someone else can chime in on that if they like but that's how I would think you would want it. It would be nice if it would regulate pressure to something like 48 and then increase it to something like 65 or 68 instead of just running at 43 and then increasing to 58.

[blindsquirrel]

What is the base pressure, key on engine off?

Whatever the base is, it should be 10.1psi less at 30kPa MAP, 5.8 less at 60kPa, 11.6psi MORE at 180kPa, 14.5psi more at 200kPa...

If it doesn't have enough range to do those numbers, it can't use the same injector data as you'd use with a real 1:1.

[pghalcrow]

Ok, I think I've figured out the issue. Yes, you guys are right that it's just not getting the fuel it needs under boost. It turns out after a conversation with Carl over at Vaporworx, is that the controller only slows down the pump at idle and ramps up to full speed at or around atmosphere. After that, the fuel pressure is effectively decreasing at a 1:1 rate since I'm not boost referenced. I'm going to add manifold boost reference to the regulator. After that, believe my flow rate table should stay flat at this point, but I'd like to understand how the Offset vs. Press vs. IGNV table works. Fuel Injector Clinic provides some baseline tables, but their flow rate table is sloped and I'm assuming this is for a N/A setup or a non-boost referenced system for FI. I've attached this.

In HP tuners, [ECM] 12308 is described as "[ECM] 12308 - Injector Offset vs. Pressure Delta vs. IGNV: This value is added to the injector pulse width as the delay time it takes for the injector to operate. It varies with the pressure across the injectors and battery voltage." From what I understand by this description, is that this table should also be flat, all columns set to the "base pressure" column values and also as you mentioned earlier. The difference here is that the "base pressure" at idle will not be real as the pump is slowed down at idle due to the effects of the Vaporworx controller. I think I can manually kick on the pump with the controller not in the loop, and use whatever the pressure is when the pump is at 100% for selecting the column for offset vs. press vs; IGNV. I hope this all makes sense.

FIC 850cc HP Tuners Data.xlsx

[blindsquirrel]

If it's manifold-referenced, it only uses the one single column that matches your base pressure (58psi = 400kPa column) - for both the Flow Rate & that big scary-looking Offset table. Copy the whole column in the datasheet, paste it into every column all the way across.