Quick guide to simultaneous VVE/MAF calibration for gen4+

[eXo3901]

The use of average is optional. I use an average of 200 for closed loop in all my logs as it seems to line up the GMVE data with the O2 sensor swings nicely. For open loop I do not use average. Play around with it to suit your own tastes.

I was going to ask how you got your GMVE trace to be so docile. Do you also mess with the logging rate?

Never had great luck using slope so I have experimented with true/false switches. In scanner terms, True=1 False=0. example: ((GMVE_STFT)*(EQcmd=1))+((GMVE_WB)*(EQcmd<1)) Obviously depends how the car is setup and what you are looking to log. Should be able to add multiple switch events as long as it multiplies out to 1. The graphs would then require less filtering if it is already in the math.

[PGA2B]

Closed loop filter
[2517.161.avg(1500)]=0 and [2517.161.avg(-200)]=0 and (abs([50090.156.slope(1500)])+abs([50090.156.slope(-500)]))<2 and ([6310]=9 OR [6310]=14 OR [6310]=15)=0

Open loop filter
[2517.161.avg(200)]=0 and [2517.161.avg(-200)]=0 and (abs([50090.156.slope(250)])+abs([50090.156.slope(-250)]))<2 and [6310.avg(250)]=9

What are 2517.161 and 50090.156?

When I check 2517 it comes back as Torque Management Advance and I don't have 50090.156 that I could find. It is Throttle?

6310 for me is Fuel Trim Cell.

I use fuel trim cells in my filters to limit data.

[hjtrbo]

I was going to ask how you got your GMVE trace to be so docile. Do you also mess with the logging rate?

Somewhat. Usually log 1 notch faster than preset. Currently settings are:

• Mass Airflow = 10Hz
• STFT Bank 1 = 5Hz
• LTFT Bank 1 (Optional depending on tune config) = 5Hz
• STFT Bank 2 = 5Hz
• LTFT Bank 2 (Optional depending on tune config) = 5Hz
• Wideband Bank 1 = 10Hz
• Wideband Bank 2 (Optional) = 10Hz
• EQ Ratio Cmd = 10Hz
• Engine Speed = 10Hz
• Intake Manifold Absolute Pressure = 10Hz
• Volumetric Efficiency = 10Hz
• Manifold Air Temp = 5Hz

Never had great luck using slope so I have experimented with true/false switches. In scanner terms, True=1 False=0. example: ((GMVE_STFT)*(EQcmd=1))+((GMVE_WB)*(EQcmd<1)) Obviously depends how the car is setup and what you are looking to log. Should be able to add multiple switch events as long as it multiplies out to 1. The graphs would then require less filtering it is already in the math.

Nice, I can see what you are doing there. I'll have a play with that for sure. Cheers for passing that on.

Slope is derivative. Its interesting to create a slope math parameter of something (i.e throttle), muck around with the time value and graph it to see how what it calculates to. You will get positive and negative slopes whose steepness changes with throttle movement. After mucking around with the slope numbers for long enough and graphing the changes against your data you get an idea on what is a smooth throttle transition and what is too aggressive.

Ideally slope is 0 which is perfect steady state, but that is impossible and the vehicle can never accelerate so we have to add in a little allowance for throttle transitions. Hence why I look forward in time and backward in time at the chosen time domains to ensure that it (i.e. throttle) falls within a boundary acceptable. Making them absolute values makes it easier to filter as all the slopes become positive values.

[wildfan1976]

Call me stupid but how do you set this up in the graph channel? Cant seem to get it to work. Got all the user math done in the math parameters.

[Leva]

Setup a maf and ve graph like you typically would copying over labels/axis from the editor. Use the gmve math as the parameter in the ve graph and stft as the parameter in your maf graph. Your graphs will then log fueling errors like they typically do using any of the other tuning methods.

[marksrig]

So since I have dual WB's, I can setup WB1 to log error for MAF and the setup WB2 to log error for VE at the same time. Do you still shutoff DFCO/CFCO? The only issue I see with comparing MAF and VE using this method on my setup is Bank 1 and Bank 2 of my engine are never even. One is always about 2-3% leaner or richer than the other like JSLLC said previously.

[hjtrbo]

That's typical. Mines around that roo. Shouldn't have to turn DFCO. Your filter should get rid of those states. Look into fuel trim cell parameter.

[07GTS]

u could also add in a WB1 + WB2 x .5 so u get the average for both banks for the calculation, do not use / it dosnt like it use x .5 is better for all this stuff

[hjtrbo]

How weird. Anyone familiar with 8 bit micros etc knows the old trick of >>1 aka bit shift right to divide by 2. Very low overhead. Apples and oranges with floating point so not applicable to the scanner.

[Capt Chewy]

So after nearly 8 hours of trying and failing over the course of days to get the math expression correct, i finally got it figured out. Now the gmve stft and the stft are very close. I'm not even sure if the injector flow rate is 100% accurate as i only had 2 data points to work from and doing so as a rookie trying to tune it, and its still close. I wound up using smokeshow's original formula but used MAT instead of IAT. Once i got it right it showed that what i had been doing with the car in ve mode and maf mode was way off. Again a total rookie.

IMG_0388 (2).jpg

This is after only three or 4 small test runs as the car is on a two post lift. i could only drag the brake to simulate resistance. The car would run horrible in blend air mode but now i was able to make a mini simulated pull. I do have it set to divide to average the fuel trims i may change it to the X .50 instead and see what it does. Also using the filters HJtrbo provided. Awesome stuff guys!

[NoDoze]

No problem. The math parameter is meant to be adjusted anyway, depending on your specific vehicle and whether you're working in OL or CL. This procedure assumes the user is comfortable with calibration enough to edit the math parameter to their liking.

What unit is the MAF data supposed to be pasted as, I assume g/s? I paste using Multiply by % - Half using g/s.

Any recommendations for filtering transient data on the street? I am trying to track down a flat spot in the log file. I haven't ruled out Idle Coastdown as the culprit due to all the metric lining up for that table when the log goes flat.

forza-3.jpg

[smokeshow]

What unit is the MAF data supposed to be pasted as, I assume g/s? I paste using Multiply by % - Half using g/s.

Any recommendations for filtering transient data on the street? I am trying to track down a flat spot in the log file. I haven't ruled out Idle Coastdown as the culprit due to all the metric lining up for that table when the log goes flat.

forza-3.jpg

That data should be filtered out. Not only is it not closed loop, it isn't even a fuel trim. There is an intrusive O2 response test running in the background because you have far too many diagnostics disabled. Saw this posted on FB a little while ago...there are a number of other things that need to be cleared up in that file before dialing in the airflow.

[hjtrbo]

Cheat sheet has had an update. Still located here
https://forum.hptuners.com/showthrea...l=1#post665159

[NoDoze]

That data should be filtered out. Not only is it not closed loop, it isn't even a fuel trim. There is an intrusive O2 response test running in the background because you have far too many diagnostics disabled. Saw this posted on FB a little while ago...there are a number of other things that need to be cleared up in that file before dialing in the airflow.

I am going to start with the stock file I sent to the tuner originally and get it running myself. It is an incredible disappointment to have paid a supposed pro in the TSP\HPT world for something that doesn't even run. Some people IMd me from that post and said they would fix it all, but who can be trusted?

[D Man]

Closed loop filter
[2517.161.avg(1500)]=0 and [2517.161.avg(-200)]=0 and (abs([50090.156.slope(1500)])+abs([50090.156.slope(-500)]))<2 and ([6310]=9 OR [6310]=14 OR [6310]=15)=0

Open loop filter
[2517.161.avg(200)]=0 and [2517.161.avg(-200)]=0 and (abs([50090.156.slope(250)])+abs([50090.156.slope(-250)]))<2 and [6310.avg(250)]=9

Can you explain these filters a little bit more. my biggest issue is at low rpms and 1-2 and 2-3 gear changes where it bogs down and trims go all over the place. i think some of these filters are blocking out some of the data im trying to find. the other thing i cant seem to find is the actual "GMVE" math so i can set up GMVE-af model logging the correct units.

now i know this is more for GM V8 applications but i am trying to apply it to my turbo 4cyl, which is a gen V e39a ecm, and another thing i cant seem to find anywhere, is how to tune the VVE on a dohc VVT application. with every cam angle change, the whole VVE table changes, and i cant set the car to have a constant cam angle. so with all the knowledge and wizardry in this thread, can someone point me in the right direction so i can get my little 2.0t Cadillac to have some normal driveablilty again. my WOT setup is great, but my closed loop sucks. attached is a light throttle log file incase its needed.21-11-11 05-54-54.hpl

[hjtrbo]

If your chasing gear change fuelling then my filters will not work. The philosophy of my filter is to only allow steady state conditions to pass.

A break down of the closed loop filter:

2517
Shortly before a shift and a bit longer after the shift, ignore.

50090
Looking at the rate of change of the throttle both before and after a point in time. If it is not a large change then the filter lets it pass.

6310
Ignore anything open loop or DFCO

Open loop filter is basically the same but only used when in power enrichment (6310 = 9)

[D Man]

If your chasing gear change fuelling then my filters will not work. The philosophy of my filter is to only allow steady state conditions to pass.

A break down of the closed loop filter:

2517
Shortly before a shift and a bit longer after the shift, ignore.

50090
Looking at the rate of change of the throttle both before and after a point in time. If it is not a large change then the filter lets it pass.

6310
Ignore anything open loop or DFCO

Open loop filter is basically the same but only used when in power enrichment (6310 = 9)

Thanks for that input, so the 2517 i will have to remove and I stay out of PE but when you say ignore DFCO, do you mean any time the throttle is 0% or does it specifically ignore DFCO. Depending on which way that sways, the shift filter would be the only thing to ignore on my end. I will disable DFCO if that filter does not work. But great work on the throttle filter. Can you help with the GMVE formula, I've searched and searched but can't find it, all I find is it being referenced or used for another calculation.

[hjtrbo]

Ignores DFCO. Because your not an e38 or e67 I have no idea if your fuel trim cells are the same. You need to verify this yourself with testing or have it confirmed by someone who knows.

Do you not have a volumetric efficiency PIDs to select? There should be 2. One is straight GMVE and the other is GMVE airflow.

If not, the clue for the GMVE formula is in the units.
mg.K/kPa
Fuel flow (mg) x manifold air temp (degrees kelvin) / manifold air pressure (kPa)

I have a spread sheet attached to this thread a few pages back that will build the formula for you if you have the GMVE PID.

[D Man]

Ignores DFCO. Because your not an e38 or e67 I have no idea if your fuel trim cells are the same. You need to verify this yourself with testing or have it confirmed by someone who knows.

Do you not have a volumetric efficiency PIDs to select? There should be 2. One is straight GMVE and the other is GMVE airflow.

If not, the clue for the GMVE formula is in the units.
mg.K/kPa
Fuel flow (mg) x manifold air temp (degrees kelvin) / manifold air pressure (kPa)

I have a spread sheet attached to this thread a few pages back that will build the formula for you if you have the GMVE PID.

Wow I'm such a noob I do have both volumetric efficiency pids. I have that spread sheet and I was able to copy paste your maths and they lined out except for stlt and ltft bank 2 since its a 4 cyl so I just removed them and divided by 2 since I have less cylinders. I believe I have the airflow model lined out now I just have to get the CL model lined out. Im sure I'm missing something simple. Thanks for clarifying that.

[hjtrbo]

For 4 cyl the easiest way so things don't get confusing is to make both fuel trim banks the same parameters. The formula will still work without modification.

(2+2)/2=2