Lean transition into PE and lean spot during WOT

[Niemer]

I can't figure this out for the life of me, I've tried messing around with all the PE settings to get it to richen up quicker, but the thing that really bothers me is the little lean spike in the middle of a WOT pull, see the below screen shot of the log. I have tried to re-scale the MAF in that area but it really messes with the curve when I do and it doesn't seem to help, the picture of the MAF curve shows what a Multiply % - Half looks like if I use the error data from the log, that doesn't look like a proper solution to me.

G8GT with a TVS1900, headers, and cone filter. Suggestions are welcome.



SOLVED!

Engine -> Fuel -> General. Injector Control, Enrich Desoot Mode

Set to disabled.

[SSUte01]

Something doesn't seem right with that screen shot. If it were just the WB that read lean in a transient, then I'd blame transport delay or WB processor issues, but your narrowbands agree. The part that seems odd to me is that the IPW and IDC follow airmass nicely (from what I could eyeball at least). I am on a Mac right now, so can't view the log or tune other than the screen shot, but what kind of fuel system are you running because if the injectors are open (or at least commanded to) and you're confident your airmass/airflow modelling is correct, I would be suspecting something else like pressure. The initial transient delay is partially normal, but the 'lean' spot mid-WOT is odd. Is it consistently at the same airmass? Is it MAF only or blended and if both MAF/SD, is the SD tuned also?

[MikeOD]

At first glance, your MAF table isn't perfectly smooth...it's not bad, but right where you're seeing the lean blip, I see some funkiness in the MAF table...try smoothing that out some...DSteck made a spreadsheet that will fit a curve through your MAF calibration if you don't like using the HPTuners smoothing function. I also see you're running MAF only. I notice most cars wind up staying a little lean on PE entry without tuning VE and running it at lower RPM.

[Niemer]

I have ID850s and a CTS-V pump with Vaporworx constant pressure controller, http://vaporworx.com/speed_control_systems.htm. MAF only tune, the issue seems to be happening at approximately the same airmass at WOT, but I logged a part throttle pull that got into PE and it happened at a little lower airmass and Hz range, the odd thing about that is that it happened at exactly the same amount of time after PE was commanded as the WOT pulls, 2.1 seconds, not sure if that has to do with anything or is just a coincidence. Screen shots and logs of those pulls are attached. This only seems to happen in second gear at those Hz ranges, the WOT portion of the 4.0.6 log is a top speed run at an autocross so it has 3rd and part of 4th gear in it. If it were a fueling issue wouldn't it most likely get worse over the course of the pull rather than come back in line? I can try and stick a gauge on the rail to monitor pressure, but all I have is mechanical and I think it would be hard to watch and be safe at the same time.

Attachment 39938

Attachment 39939

[Niemer]

At first glance, your MAF table isn't perfectly smooth...it's not bad, but right where you're seeing the lean blip, I see some funkiness in the MAF table...try smoothing that out some...DSteck made a spreadsheet that will fit a curve through your MAF calibration if you don't like using the HPTuners smoothing function. I also see you're running MAF only. I notice most cars wind up staying a little lean on PE entry without tuning VE and running it at lower RPM.

If you're referring to the funkiness in the MAF table on the attached screen shot, that was just an example of what it would look like if I made an adjustment to it based on the log, the table in the actual tune doesn't have that flat spot. Can you point me to DSteck's spreadsheet? I tried to make one a while back with limited success.

[SSUte01]

I would certainly model the SD coefficients, it's not hard and you'll reap the benefits, this may or may not be the only contributor to your issue, but the other benefits are worth it. Given you have fairly extensively modelled the MAF, you've got most of the baseline for your VE coefficients anyhow.

I believe that'll make a difference for the initial entry to PE, there's always going to be some delay in that situation in MAF only setup depending on how the intake tract is setup. As to the lean out 2.1 secs into WOT, that I'm not so sure, in any case I'd like to see how it behaves in correctly modelled MAF/SD.

[Niemer]

I have been reading up on how to tune VE (SD?) but am still thoroughly confused at this point, I have some custom histograms set up and have putzed around with Bluecat's EQ <> VE program but am still not confidant enough to give it a go.

I believe I have pinpointed the cause of the lean spike at 2.1 seconds, like you thought I am loosing pressure, I tapped into the pressure sensor for the controller and logged that, sure enough there is a pressure drop right before the lean condition occurs, but it corrects itself after. I've contacted the manufacturer of the pump controller to see what they think about it.

Attachment 39950

[Niemer]

The more I look at it the more I'm not convinced that that is actually the problem. Those peaks and valleys are only .03s each, it hardly seems that a low pressure for that short of a time would cause this, plus there is a dip at frame 10223 that goes to 49.4psi with no corresponding lean condition.

[MikeOD]

If you're referring to the funkiness in the MAF table on the attached screen shot, that was just an example of what it would look like if I made an adjustment to it based on the log, the table in the actual tune doesn't have that flat spot. Can you point me to DSteck's spreadsheet? I tried to make one a while back with limited success.

No, I was looking at the table in the tune.

Here's a link to the MAF spreadsheet http://www.hptuners.com/forum/showth...ght=MAF+smooth

[Niemer]

Thanks, going to give that a try. Smoothed the curve in two steps, the max percent difference between what I had and the new curve from 4800Hz on up is +/-5%, seems reasonable. I left everything below 4800Hz alone, I'll report back later.

[10_SS]

How often does this happen? My first WOT run or two after installing my Whipple supercharger resulted in similar lean conditions... random without any explanation, I realized after this that it was due to air trapped in the large empty Whipple fuel rails... after the first few WOT runs, no random lean spikes returned... ever. The newer cars use return less fuel rails to stop the fuel in the tank being heated from the engine and creating excess emissions creating air pockets that have nowhere to go until high flow.. and even then... not all the air moves in one instant. .... it also causes problems for things like this (air and vapor lock). Sometimes taking a step forward requires a few steps backwards :/

[Niemer]

It's happened every time since I installed the ID850s, didn't do it with the stock injectors. I also have a lean idle and cold stumble that I am fighting that only showed up after the new injectors, but that'll be a different thread. I'm going to pull the rails off again and check for obstructions in the injectors and also take a look at the plugs.

[Niemer]

Thanks, going to give that a try. Smoothed the curve in two steps, the max percent difference between what I had and the new curve from 4800Hz on up is +/-5%, seems reasonable. I left everything below 4800Hz alone, I'll report back later.

This seems to have helped a bunch, thanks MikeOD. I still have a little error in one spot, but it's down to 4% vs. 10%, hopefully a couple more iterations will get that down to 1-2%, it also has seemed to have greatly reduced my lean spot entering PE.

Attachment 40031

[Niemer]

Alright, let's confuse this a bit more. I found it again, but this time in part throttle PE and a much lower MAF frequency, probably climbing a hill or something. Back to the drawing board?

Attachment 40049

[Niemer]

Another thing to note, it happened 2.1 seconds after PE was commanded, just like all the other times.

[Niemer]

Here is another part throttle PE with the spike at a different frequency. Same 2.1 seconds after PE is commanded though. I also smoothed the MAF curve down in the low frequency cells using DSteck's spreadsheet, I probably can't get it any smoother than it is. STFTs are +/-2%. I'm at a loss, here's another screen cap with the log and tune.

Attachment 40106

[MikeOD]

Is your VE dialed in as accurately?

[Niemer]

VE is disabled

[realcanuk]

The lean spike just after entering PE is common on these PD blowers. I have been fighting that for a while. To my surprise, when I finally gave in and tuned the VE properly and run it with the maf, the spike is gone. I always had it when running maf only.
What I'm not getting is the spike later on that you highlight in your pictures, seems to be well after you are in PE. I think that is a different issue then the lean spike just when entering PE.

[Niemer]

Yeah, so do I. I re-scaled the MAF in OL rather than using fuel trims and that really helped with the transition entering PE, but I'm still seeing the spike well into PE. This issue didn't exist before I changed out the pump and injectors, unfortunately I didn't get a log going into PE with just the fuel pump swap and stock injectors, so I can't say which might be the culprit.

Another thing I have been thinking about is that maybe something else is pulling fuel right there? Torque management maybe? What all has the ability to do that?