Help Tuning VVE / Start and run, but not idle

[Loose_Screws]

I am trying to work on VVE table and get it dialed in, but having some trouble after loading tune with revised VVE table. Car will crank perfectly, idle perfectly for just a moment and then die. Looking at the log, the IPW drops to 0.70, so I thought it was me being too aggressive pulling back on the VVE. However, it tried 3 or 4 tunes where I made slight adjustments and the results were the same...car will crank and run, then dies because the IPW drops so low.

I am following Goat Rope Garage's method from youtube on failing the MAF, disabling the LT and ST trims, and logging EQ error %. The AFR readout on the WB matches what I am seeing in the EQ error histogram, so I feel good about the changes I am trying to make to the VVE. I copy the EQ error % and paste to the VVE table (using %, 1/2%, whole table or just portions) in the tune setup for failed MAF. I can keep using the failed MAF tune and alter the VVE table there, correct?

So here is the tune setup to fail the MAF and log EQ error - 5.3.24 - RBs Nova Tune_Fail MAF, Log EQ Error.hpt

Here is the log after loading tune and cruising to dinner - 5.3.24 - RB Nova, MAF Fail, cruise to dinner.hpl

Here is the tune with VVE adjusted using log - 5.3.24 - RBs Nova Tune_Fail MAF, Log EQ Error, Pass 1.hpt

Here is a quick log of the car starting, but not idling - 5.3.24 - RB Nova, Load VVE, won't idle.hpl

I reloaded the first tune, cranked the car and drove home and logged - 5.3.24 - RB Nova, MAF Fail, Cruise Home.hpl

I am also curious why there's a cliff at 60 kPa. Makes no sense, but the car runs just fine (when MAF not failed).

Engine is swapped into a 1963 Chevy II wagon, was built with a mix of everything...so there's not a good base tune I can compare too and try to figure this out. Below is the setup with part numbers.

5.3 Gen IV with truck accessories
E38 ECM (VIN 1GC2GTDG1A1175124) from Chevy Express Van
T42 TCM / 4L65 with VSS and slightly higher stall torque converter
Custom built complete wiring harness
RME87 Throttle Body (gold blade) (Delco / bought new)
15865791 MAF (LS7 card style) mounted 15" from TB in 4" aluminum intake tube, cone filter (Delco / bought new)
12594942 MAP (1.3 bar) (Delco / bought new)
12613411 Injectors in TBSS Intake and fuel rail (bought slightly used)
PCV fresh air connected to port on intake tube between MAF and TB
PCV crankcase air connected to top of intake (full vacuum), 1/16" restriction inline on PCV hose
In tank fuel pump / external Vette filter+regulator (returnless), confirmed ~58 PSI with manual gauge on fuel rail
TSP Stage 2 Low Lift Truck Camshaft, 212/218, .550/.550 112 LSA, LS6 valve springs (no DOD/AFM or VVT)
Mid-length headers 1 7/8" primaries into 3" collector
2 1/2" dual exhaust with X pipe, no cats
Innovative LC-2 wideband

[blindsquirrel]

You have it switching to 'MAF-only' above 2,000 RPM.

HOW TO FAIL MAF:
Dynamic disable/reenable: 8,000/7,950
MAF fail high/low: 2/1 or 0/0
DTC P0101/P0102/P0103: MIL on First Error
edit: And also do not log any channels that directly read from the MAF, when you've specifically asked the ECM not to use the MAF.

You're also in open loop and I guess planning to use the wideband to tune the normal driving areas, even though after you get done the narrowbands will be the ones in control of things? If you have a race car and plan to run it all the time forever in open loop then that's fine. If you want it to run in closed loop then tune it in closed loop using the sensors that are in charge.

[LS ROB]

According to that log, it's super lean. Are you wanting to run this in Open Loop or Closed Loop? I'd suggest turning the LT & ST trims back on and running it in closed loop that way the ECM can make corrections.

[Loose_Screws]

Appreciate the info!

I'll change dynamic disable/reenable as suggested, MIL on for all 3 DTCs and also stop logging MAF. I'm curious about using the NB for correction, I thought they would only show rich/lean? Is the voltage variance enough to calculate a percentage change needed?

Lastly, any idea why the tune I loaded with VVE altered would drop the IPW to 0.07?

[Loose_Screws]

According to that log, it's super lean. Are you wanting to run this in Open Loop or Closed Loop? I'd suggest turning the LT & ST trims back on and running it in closed loop that way the ECM can make corrections.

I thought it needed to be in open loop with LT and ST trims disabled to calculate % change needed for the VVE?

Now that I think about it, I could log LT+ST in histogram and apply to VVE? Disregard using WB?

[LS ROB]

I thought it needed to be in open loop with LT and ST trims disabled to calculate % change needed for the VVE?

Now that I think about it, I could log LT+ST in histogram and apply to VVE? Disregard using WB?

That's correct. You can use the Wideband for WOT corrections, but everything else; I'd use the NB's to tune it.

[blindsquirrel]

It'll actually be using fuel trims for your tuning, not the direct O2 voltage. LT/ST trims are how the ECM controls rich/lean when it's in closed loop. Positive numbers are adding fuel, negative numbers are subtracting.

[edcmat-l1]

VVE step.jpg

First off all, you didn't apply your WB correction you multiplied the entire table by roughly .9 for a 10% reduction in fueling.

Second, the VVE is not at all correct in shape or numerically. You have a step in fueling (or airflow) of a roughly 100% increase, from 57.5 Kpa to 60 Kpa. So, at the tip of the throttle from idle, the fueling roughly doubles.

You need to change that first thing. Either manually shape the table into something that makes sense or hit the smoothing button about 50 times.

Next, set your dynamic airflow disable-reenable to 8000 and 7900.

Then turn your trims on and let the PCM tell you want it needs. Forget about the wideband for now.

Lastly, I've seen this VVE before in another thread. Told you then it needed to be cleaned up.

[Loose_Screws]

Which LT and ST PIDs do I want to log for VVE correction? Should I combine them into one histogram? Or use only LT since the VVE table is so rough right now?

[blindsquirrel]

There's a pre-made math thingie for STFT+LTFT in the scanner. Just make a graph that uses the same axes as your VE/VVE table.

[Loose_Screws]

Here is the tune I loaded with MAF failed, CL enabled - 5.4 - Nova, fail MAF, Log LT and ST for VVE correction.hpt

Here is a test log with engine warmed up. - 5.4.24 - Test Log.hpl

I used RPM control to raise/hold RPMs, however it wouldn't go above 1200 RPM for some reason, so I used pedal.

The WB is NOT matching the NB O2s, but I am understanding now that WB can't be trusted at lower RPMs.

I did find the predefined maths for LT+ST...thanks!

So, reading the results of the histogram, it's rich and I can confirm by the smell LOL! I am making a pass at the VVE table now, multiplying by 1/2%. Will post up comparison shortly for review.

If you see anything squirrely (besides the VVE table), please let me know!

[blindsquirrel]

Uh, how is it going to be rich if it's in closed loop which corrects it back to stoich automagically?

[blindsquirrel]

Raise & hold RPM, for what? Are you just tuning idle? If so you need the idle tuning guide.

[Loose_Screws]

Uh, how is it going to be rich if it's in closed loop which corrects it back to stoich automagically?

Very good point...I forget the corrections were in play.

Raise & hold RPM, for what? Are you just tuning idle? If so you need the idle tuning guide.

My thought was to hit some other cells around the idle zone. Where is the tuning idle guide? I'll read through it...being new still to HP Tuners I am sure I'm missing something.

I adjusted the VVE using the test log. Whenever I select calculate coefficients, I'm not liking the results. Here is the adjusted tune. 5.4 - Nova, fail MAF, Log LT and ST for VVE correction, pass1 at VVE change.hpt

Here is the log after "Pass 1" tune was loaded with adjusted VVE. Still showing 10% correction being applied. 5.4.24 - Pass 1 VVE Loaded.hpl

[blindsquirrel]

Add the filter string: [50118.239]=1&[50091.156]>0

That is a quick'n'dirty way to exclude everything where EQ isn't 1 and APP is at idle.

[blindsquirrel]

Think maybe you should clean up your DTC list for stuff that isn't being used anymore?

This all seems very random poking at things, I think some studying and finding a methodical approach to follow would give much better results.

[edcmat-l1]

smoothing.jpg

First thing you need to do is learn how to follow directions. You need to make the VVE table look like this before you go any further.

[Loose_Screws]

smoothing.jpg

First thing you need to do is learn how to follow directions. You need to make the VVE table look like this before you go any further.

I don't disagree at all. I'm just trying to find a solid starting point of reference before I scale/smooth the table. Unless you have a suggested table to start from? I have a 2010 5.3 Silverado stock tune, but the engine setup is quite different.

[Loose_Screws]

...I'm going to take a stab at smoothing the table as is.

[edcmat-l1]

Dude the table is horrible as it is. Nothing wrong with hitting the smooth button 20 times. You aren't going to make it any worse that's for sure.