Transient Fuel Question

[mn_vette]

I'm attempting to do some transient fueling tuning on my car. This is an LS3 Corvette with an ECS supercharger on it, no other mods. This is a MAP only tune. In an attempt to get rid of a lean spike I kept turning up the Impact Factor Gain.

I'm doing a low speed 50% throttle blip to try to get things figured out.


I started out with this. A lean area where the throttle starts opening, but then it goes down where it should be pretty quickly.
MoreFuel1.PNG



After turning the Impact Factor Gain up to 1.0 I ended up with this. Same kind of spike in the beginning, but then a large rich area when the transient fuel kicks in. Then back to where it should be.

WayMoreFuel.PNG


Any thoughts on what I can do here? or where I should be looking for a problem? Any advice would be appreciated. Thanks.

Here is the starting file.

23_08_21_FuelAdjust_o2_Off.hpt

[JasonHSV]

Bump up your enrichment ramp in rate to 1.9 and tune your VE where the lean spike occurs to get the desired result. In most cases you don't or shouldn't need to touch transients.

[mn_vette]

Bump up your enrichment ramp in rate to 1.9 and tune your VE where the lean spike occurs to get the desired result. In most cases you don't or shouldn't need to touch transients.

I took your advice, put the transient fueling back, increased the PE ramp rate to 1.9, and added a bunch more fuel in the 100kpa area. I went through and smoothed a lot of things out. I took more data and here is what I'm seeing.

Quick press of the throttle. I'm still getting the lean spike, but the before and after are right on for AFR.

QuickStab.PNG


I also did a very slow ramp of the throttle. I seem to be getting some lean spikes when the MAP readings bump up. But the flat MAP areas seem fairly rich. The VE table in this area is pretty smooth.
slowRamp.PNG

All the snips are taken at the same zoom level to make them comparable. I'm not sure if it helps to have the actual log files instead of the pics. Any advice would be helpful. Thanks.



Here is the new tuning file I used.
23_08_21_NormalTransiet_EnrichRampUp.hpt

[kingtal0n]

If the injection timing is off it will display these characteristics , did you change the cam or fuel injectors and forget to adjust end or start of injection?

There is also a table called "boiling time" or something like that, and a similar table across from it that may contribute additional fuel during MAP changes I think

I agree strongly that weight in the VE table is an easy way to deliver the appropriate fuel on-time for transient portions of the map that you will rarely steady state inside, but it wont work well without the correct injector phase

[mn_vette]

If the injection timing is off it will display these characteristics , did you change the cam or fuel injectors and forget to adjust end or start of injection?

There is also a table called "boiling time" or something like that, and a similar table across from it that may contribute additional fuel during MAP changes I think

I agree strongly that weight in the VE table is an easy way to deliver the appropriate fuel on-time for transient portions of the map that you will rarely steady state inside, but it wont work well without the correct injector phase

I'm running the stock cam with bigger injectors. I didn't think I wanted to move the injector timing with the stock cam, but maybe I need to do a bit more research into that.

I found an INJECTOR TIMING TOOL spreadsheet out there and I understand the math and what is going on with the timing. I'm really confused about what the "ideal" time to do the injection is. Do you want the injection to end before or maybe just after the peak piston speed? Or do you push it back to peak intake valve opening? or can you push it back further?

At idle it makes sense to start the injection pulse after the cam has cleared its overlap portion, and my stock numbers seem to do that.

Once I get over 2k rpm the injection doesn't occur when the intake valve is open at all.

[kingtal0n]

With bigger injectors (you don't say how much bigger, though) there is often an advantage to spraying into an open intake valve from say 300* btdcc (before top dead center compression) to 240* btdcc where the piston is reaching peak velocity (285*~ depends on rod stroke ratio).

The factory likely puts the fuel onto the closed intake valve 380-450* btdcc (sorry I am using Haltech end of injection terminology its what I tune the most)

Many say its trial and error and I agree. I always try injecting to a closed valve first, then to the open valve through the middle-ish of the intake stroke. As RPMS rise you can inject later and later as the chaotic swirling helps vaporize fuel and this gives you a larger window to inject with low duty cycles. Once duty cycle exceeds around 30% it matters alot less where you inject, which is why the larger injectors have a distinct advantage as they can fit more fuel into the intake stroke window while avoiding overlap.

With a big cam we like to avoid overlap near idle, spraying before or during overlap may lead to exhaust fuel smell in port injected engines. The stock cam you are probably fine.


At this point knowing you have the stock camshaft I would recommend you compare injector on-time (milliseconds) to wideband behavior during the transient conditionals where you have problem.

For example if you see a large spike in the on-time (ms injection pulse) while simultaneously a lean spike in the wideband, and no matter how much fuel is spiked in its still leaning out briefly, it becomes clear something with the way fuel is being injected isn't reaching the cylinder in time, it may be related to injection phase (Injector end of injection or start of injection setting). That would be my first diagnostic strategy here to determining whether you have a injection timing issue or simply a fuel starvation (need a larger pump shot) issue.

[GHuggins]

Adding 15 or 20 to the stock boundary can make a big difference on a stock cam for keeping fuel in the cylinder. Your VE table should be a lot smoother than that with a blower too, so recheck it after changing the EOI.

[mn_vette]

My injectors are now 60#/hr. I'm not sure what the stock ones are for an LS3.


I added the injector data on my last log file and here is what it is. I scaled the MAP(orange) and Injector PW(green) to line up with each other. You can see the pulse width doesn't seem to be tracking the MAP changes as quickly as it should. I can artificially bump up the VE in that area, but I'm not sure that is the correct answer. Do you think playing with the injector timing will fix this?

PWdelta.JPG



Edit:
I played around with the injector timing spreadsheet stuff a bit. I would appreciate it if someone could take a look at the changes I made and let me know how badly I'm off in understanding when I want my injectors to fire. I have the sheet with the stock numbers and one with my modifications. Thanks.

INJECTOR TIMING TOOL GEN IV - Stock.xlsx

INJECTOR TIMING TOOL GEN IV - Tweaked.xlsx

[kingtal0n]

The question isn't whether or not adjusting this or that will fix it. The question is, given the current configuration, CAN the issue be fixed at all? Does ANYTHING help or fix it? I would focus on whether any adjustments you make no matter how insane can actually fix the problem or if it persists no matter what.

The answer to that question will help you understand whether the issue is related to fuel delivery electrical/mechanical, or whether the issue is with the tuning software/ECU.

Next,
Here is the possible causes off the top of my head
1. Wrong injector wiring (wrong injector firing order) Diagnostic: Use scanner to disable injectors 1 at a time and unplug them on that cylinder to see whether they line up with what the ECU thinks.
2. Manifold leaking influence on fuel boiling/condensation, Diagnostic: Perform pressure test of intake manifold
3. Injector phase, wrong injection timing Diagnostic: Adjust injection phase experimentally and compare similar conditions
4. "Pump shot" problem , diagnostic: Adjust setting in the ECU which increase pump shot (there are a few I already mentioned)
5. Fuel system pressure issue, perhaps fuel pressure is dropping (unlikely but worth mention) Diagnostic : watch fuel pressure during the situation

You must also remember there is a transient delay depending how far the O2 sensor is from engine. If the oxygen sensor is a few feet the delay on what we see in the wideband output will be skewed from where the fuel was injected. Depending on RPM
Therefore it is hard for me to tell sitting here without looking at the engine and knowing how much power and distance of the oxygen sensor etc... to say whether the fuel is making it to the cylinder in time.

[mn_vette]

I was able to do more testing with more updated VE table and injector timing. I'm getting the same thing.

From your list, adjusting the "Pump shot" sounds like something promising. But I can't seem to find any instructions on how to do this on a 4th gen. Maybe someone has some suggestions......and yes I have tried searching, but I think I'm missing some key words or something.

The other thing I'm thinking is that at low RPM the supercharger is still blowing massive amounts of air. For example, if I unplug my card style MAF there is a ton of air that is blowing out of that hole when I'm just idling. The same thing would be true for any small change in the throttle body opening. Definitely different from an N/A engine trying to suck the air in.

I did check the fuel pressure with a gauge and didn't see anything odd with the throttle stabs. Pressure was pretty steady at the lower RPMs. I can try looking into some of the other things on the list as well.

[kingtal0n]

Hmm are you using a maf? Are you forced to use a maf? I would get rid of the maf if you can. MAF sensors are powerful, delicate, specific, precise, very good if they are installed and used correctly. But this is difficult to achieve unless you are very experienced with testing, calibrating, installing and tuning maf sensors in myriad applications.

If you must use the maf I have some suggestions. I have tuned hundreds of maf cars with forced induction and can help you with that. But I would prefer you avoid it altogether.

Since you are forced inducted you should perform a pressure test. Fill the intake/plumbing with 'boost pressure' using an air compressor while the engine is off. If you need a video I have one of me performing this test on a turbo engine.
During the pressure test you can also test the PCV valve. If you are using a stock chevrolet pcv valve it is probably leaking into the crankcase and blowing oil out of the engine and out of oil seals. Just a heads up.

I'll look in your file if you posted one and see if I can find anything related to pump shot stuff.

[kingtal0n]

Under your transient tab there is tons of stuff to play with.
'warmup transient fuel correct' play with the gain there. Make a big adjustment so its really obvious what happened/changed.

Only change 1 thing at a time.

'fuel from wall stabil' no idea what this does but its worth adjusting to see what happens.

'impact factor' mess with gains first.


Try turning down the 'enable delay' cut it in half or 1/4

If nothing there help play with the other settings. Something will wake it up. What you want to see is a nice spike in the injector on-time when you move the throttle quickly. If you get a huge spike of injector ms and the wideband still going lean it means the fuel isn't being delivered properly because of injector wiring or injector phase. I can't think of any else it wouldn't go into the cylinder quickly enough unless your injectors are way above the supercharger or something.

[mn_vette]

Hmm are you using a maf? Are you forced to use a maf? I would get rid of the maf if you can. MAF sensors are powerful, delicate, specific, precise, very good if they are installed and used correctly. But this is difficult to achieve unless you are very experienced with testing, calibrating, installing and tuning maf sensors in myriad applications.

If you must use the maf I have some suggestions. I have tuned hundreds of maf cars with forced induction and can help you with that. But I would prefer you avoid it altogether.

Since you are forced inducted you should perform a pressure test. Fill the intake/plumbing with 'boost pressure' using an air compressor while the engine is off. If you need a video I have one of me performing this test on a turbo engine.
During the pressure test you can also test the PCV valve. If you are using a stock chevrolet pcv valve it is probably leaking into the crankcase and blowing oil out of the engine and out of oil seals. Just a heads up.

I'll look in your file if you posted one and see if I can find anything related to pump shot stuff.

I am running SD only. MAF is disabled. The air is to turbulent to get a good MAF reading with the supercharger and MAF placement. But that did give me a good idea, I decided to plot the MAF just to see what it looks like. The MAF is spiking much faster than the MAP. White vs Pink in the bottom plot.

fast_throttle.JPG

This led me to start thinking about the VE airflow and the transient predictability coefficients. I think these will still do something even in SD. If I can get things to react a bit faster, maybe that will help.


I also tried playing with the Stomp Compensation. Maybe I did something wrong, but changing the base gain by a large amount didn't seem to do anything. The search still continues.

[kingtal0n]

If an engine has good air velocity in the ports and a well tuned VE map then even without a transient pump shot it should be very 'drivable' and respond well.

Does it physically kick, stutter, fall on its face during transitions? That is another sign of a injection phase issue or perhaps an air leak or PCV leaking issue

I would def. pressure test the intake system and play with injection timing if you have not done so
Did you disable injectors one at a time to make sure they are wired correct just a reminder

[mn_vette]

The car seems to drive fine. I don't notice anything out of the ordinary. But I'm trying to go through and tidy up some of the thing like the lean tip in.

I did find the PVC hose had come off the crank case. That explains the smell I noticed right after I started messing with the injector timing. That kind of messed with thing.

I will keep checking things and keep playing with knobs. I really wish some of these areas had better details. I have a feeling I'm mis-interpreting half the fields in the ECU.