Strange jump in VE between idle and rpms above idle

Hello.

I am tuning a turbo K24. I have tuned in steady state at 1250 rpm to 5000 rpm. This area of the tune looks normal to me. However, when I extrapolate the changes down to the idle zones (1000 rpm) and below, the AFR is reading rich. To get this area to match the target AFR at idle, I needed about 45% VE in those cells. What doesn't make sense to me is how the 1250 rpm row right above 1000 rpm has about 85% VE. Basically I have a jump of 40% VE between idle rpm and 1250 rpm.

I could increase short term fuel trim in the idle area of operation but that seems like a bandaid solution. Is it possible to fix this issue without excessive short term fuel trim?

Thanks.

Likely injector dead times may be a touch too high. If the cams are aggressive VE will be very low at idle.

Hello.

The engine is completely stock except for a 50 degree VTC gear and type s oil pump. I have configured the injectors on the ECU based on the data Injector Dynamics supplies as well as using a preset for the injectors that my Haltech ECU supplies. Both times I have had this problem. Could it really be the injector dead time? I feel like its unlikely that both sources would be inaccurate. Ill try messing around with the dead time though and see what happens, thanks for your insight.

As a quick sanity check get idle lambda to around 0.85 or 0.9 then multiply the VE up and down by 10 or 15 % and see what the actual lambda shift is. It should reveal if dead time is close or not.

Have you enabled any other correction tables? Usually only the coolant, and post start correction are commonly left on, and people either use air temp correction or turn it off and use the auto V.E air temp compensation.

And is the zero demand table enabled? This can kick in under idle conditions

I have corrections on but the output was 0%. The zero demand table is not enabled.

Do the percentage offset test (you need to do a block of cells around the idle point, you may need to tweak ignition if torque shifts a bit). If lambda output shifts more than VE proportionally, deadtimes are likely too small, if lambda shifts less proportionally than VE change dead times are too high.

I am currently testing the car. I find that when I make a 10% change up or down in VE, the resulting lambda is about 0.01-0.02 off my calculated result. For example, if the current lambda is 0.70, a -10% change (+0.07) should result in a lambda reading 0.77. Is this what you meant?

It may be worth mentioning that I noticed this also happens in higher rpms at idle load. When the car is not moving and I increase the rpm by slowly opening the throttle (staying at idle load), I am running rich at the higher rpms as well.

"I am currently testing the car. I find that when I make a 10% change up or down in VE, the resulting lambda is about 0.01-0.02 off my calculated result. For example, if the current lambda is 0.70, a -10% change (+0.07) should result in a lambda reading 0.77. Is this what you meant?"

Yes that's what I meant. Probably best to save the map as is then try some incremental changes to dead time. This is assuming fuel pressure is running at specification for the injector flow rating of course.

I have changed the dead times in the zones that it was accessing at idle. While it did help, lean it out, it is still very rich. I have 50% max disenrich for the short term fuel trim (the applied trim is always at this max) and could only get to about 0.87-0.90 lambda. Without any changes to dead time it idles at about 0.80 lambda. I don't think it's a dead time problem, but please correct me if I am wrong.

Also, can you please elaborate what you mean by "fuel pressure is running at specification for the injector flow rating"? I have a fuel pressure sensor installed and the fuel pressure is typically around 400-415 kpa. My flow rate table does have a zone at this pressure.

What nominal 3 bar differential flow rate are the injectors?

Is that absolute or differential fuel pressure?

If you are using pressure and voltage mapped dead times that's probably the best you can do, I wasn't sure if you were just using a 3bar differential array as opposed to a full table.

If you don't need 4 bar base to flow you target peak power you might find it easier to tune at 3 bar base or slightly under as you will effectively have more resolution.

You may find tweaking cam timing can assist in bumping VE also.

Nominal flow rate at 3 bar is 1065 cc/min (ID1050x). My maps are using differential fuel pressure. The dead times are mapped with voltage and pressure axes. I can't adjust my base pressure since I don't have a fuel pressure regulator. I can change the value in the ECU but I believe it doesn't affect fuel delivery since I have the fuel pressure sensor installed.

That would suggest the injectors are around 1250cc effective flow, should still be able to control idle on pump fuel with 500cc plus cylinders.

It may be work playing with offsets in the dead time table and cam timing to see if you can get it a bit happier. If it is otherwise running well I wouldn't stress too much if the ve/lambda shift test is reasonably close, that means the ecu should be able to compensate reasonably well for significant changes in barometric pressure and temperature if the fuel delivery is reasonably linear with those dead times and VE numbers.

Ok I found the problem. I looked at the voltage chart for my map sensor. The lowest value was -56.6 kpa which also happened to be my displayed load at idle. Turns out the sensor was reading below this value but I didn't have a value for voltage readings less than 0.5 volts so the ECU capped it at -56.6 kpa. Once I added a point at 0 volts, the lambda was on target with only 16% STFT which makes a lot more sense. I should expect a smooth VE change next time I go on the dyno and tune at the 2 load zones that I can now access.

Thanks for your help!