Posted on HPT forum too, hope to generate some good discussion on this.

Some of you may have read some of my recent posts but for those that haven't some back ground. 88 TransAm GTA conversion, 98 LS1. Went through the VE tuning with MAF fail set to zero, dialed in VE table. Turned maf on and logged and corrected Airflow tables. LTFTs are still adding a ton of fuel with adjusted tables and MAF turned back on. Trouble shooting found that I had the wrong injectors installed, they are 99/00s which are 2lbs lower flow rate. OK, I corrected this and back to square one, need to tune VE again. MAF should be dead on because we were measuring actual airflow against reported aiflow when I tuned that.

Now here's where the question comes in. Fair warning, this goes against everything I've seen posted or documented.

Assuming you are going to run a MAF, why would you want to turn it off to tune your VE tables? If you've taken the time to check the actual flow against reported flow and adjusted the MAF calibration couldn't you just go out and log you LTFT/STFTs and adjust the VE table with the data logged?

Here's what got me started on this. I updated the IFR tables yesterday to reflect the actual injectors in the car. Reset LTFTs and went out and drove around for just a short drive. MAF was enabled and plugged in. LTFTs were still postivie but much lower than before the IFR adjustments. Did a little more driving/logging today while running errands. Had about 30 miles total on it. Not enough I know but STFTs were all pretty low (+/-3) in the cells I was hitting and LTFTs were fairly settled. Decided to go ahead and spend some time and do some logs with the MAF turned off (fail frequency 0), DTCs set so I know it's off. Guess what, after about 40 miles the LTFT/STFTs look almost identical to the ones logged when the MAF was still connected.

All theory at this point but makes me wonder if we can skip the MAF disconnect and tweak order we tune. Tune your MAF then do your VE table?

Heat shield on flame away! Hoping some of you senior tuners can chime in and prove/disprove the theory.

 

because the PCM does not completely fuel off the VE tables if the MAF is enabled....

and if you "tune your MAF" first, then its skewed whenever you change the rest of the tune.

 

Hmmmmmm...well, all I'm gonna say is see if it works. If it does, great. Since I have LT headers, I don't trust my fuel trims for ****, so doing this wouldn't work for me. Then again, I run an open loop SD tune, so I have already misplaced my narrowband O2 sensors and the MAF.

 

I agree, the PCM does not completely tune off the VE when the MAF is enabled. But if you're going to run a MAF and not go SD wouldn't you want the VE table to be calibrated in the same mode that it's actually going to be used in, i.e. w/ MAF connected and it (VE) being only part of the equation.

I do see how you might want to do the the MAFless tune to dial in the secondary VE which is used in case of MAF failure. Of course you could probably get by with leaving that set to stock. Hopefully you want be out hammering it if you get a code. Stock tables should be close enough in theory to get you home. Unless of course you've done good heads, large cam, headers, etc.

On the second part of your response, "then it's skewed", is it really skewed or was it skewed before you calibrated and now it's accurate?

Don't take this wrong, not wanting to argue just wanting to get some ongoing discussion.

 

well, skewed is always a relative term.
however, the end result is one number, that the computer considers to be the amount of air in that cyl..
now with the VE table, its a mathematically calculated constant..(the VE table is calculated to begin with...) and in an ideal world, with everything matching the math model, a 70 in the VE table means the engines cyl is 70% full..
with the MAF... you're getting a frequency signal.. this frequency is looked up on a table, and thats the number used. there are alot of things that can effect it, but its generally close.. however you'll notice theres 1000s of OEM cars that use MAP based speed density tuning from the factory. there are no OEM ones that use just a MAF, because the MAF is slightly off for various reasons.. it is a physical measuring device after all.. and it is indirectly measuring the air mass...
meanwhile the pressure sensor in the manifold is highly accurate in comparison.. but its is calculating the air indirectly.

so on on hand, you have a accurate measurement, but a calculated number derived from it.
on the other hand you have a less accurate measurement, looked up at a table.

both are more then accurate enough to make a engine run well. however because of how the PCM looks at everything, in theory and thru the logical steps developed by alot of tuners on here, we go by the VE table first.. get it doing what we want, then we make the MAF match it... im sure its possible, but i dont know how you would make a accurate VE table based off the MAF readings.. not to say you cant do it the other way, but you'll be on your own figuring that mess out.. lol.

did that make sense, or did i just ramble on?

 

in a nutshell what im trying to say is:

if both are on there, how do you know what one the PCM was using that instant and know what one to correct??

and then how do you make changes from there??

 

I see what your saying, now as to understanding it I may need to read it a couple more times. I did understand the "you'll be own your own part", already throwing me to the wolves... LOL.

As to your part two. Exactly, if both are on there how do you know which one the PCM was using. I understand a lot of guys have come before me on this one and they some how decided you need to use SD tuning to get the VE dialed in. I guess I'm just questioning how they came up with that and is it really the best way. Probably is but again thought it worth discussing.

I'm not sure we can come up with the answer. If we could see the actual PCM logic it I'm sure that would shed some light on this.

 

well, you need the VE table to idle and run the car.
you dont need the MAF

so since you need to eliminate one to figure it out, its easy to see what one goes.

 

There are 3 unknowns: IFR table, VE table, MAF table.

If you change injectors, then the IFR table becomes unknown.
If you change cam, heads, intake, headers..., then VE table becomes unknown...
as also may the MAF table (MAF calibration is sensitive to airflow changes).

The IFR table can be made known by calculating it.
To make the remaining 2 unknowns known (...lol...), you have to eliminate one of them...

You can't eliminate the VE table, because when you run with the MAF, the PCM uses a combination of MAF and VE table.

So it's just easier to eliminate the MAF and the PCM runs exclusively on the VE table...
This allows the VE table to be dialed in
(and then this can be used to dial in the MAF table).

$0.01

 

MrDude, in your earlier post you said "we go by the VE table first.. get it doing what we want, then we make the MAF match it."

Am I understanding this correctly, you do an SD tune and get the VE table dialed in to achieve LTFTs as close to zero or slightly negative as you can. You then hook up the MAF and scale the aiflow tables to get your LTFTs back to those same zero/negative numbers?

If I read and understood that correctly that means you're not calibrating (or at least using that calibration) the MAF using frequency/Dynamic airflow. You're just adjusting the MAF table to get the LTFTs back to what they were when you completed the VE tuning?

Seems to me if you calibrated the MAF and then had to change it to get the LTFTs back to the desired reading then you're "faking" out the PCM instead of reporting an actual measured (however accurate) number.

I guess that raises the question, can you accurately calibrate the MAF with only MAF in Hz and Dynamic Airflow numbers? If you can and it's accuracy could be considered fairly or highly accurate then that would imply another factor/variable in the air/fuel formula is incorrect. Question is which one?

Starting to get a headache now...

 

If your VE table is dialed in (based on WBO readings), then the VE table derived airflow is accurate and can be used to dial in the MAF, or...

with the MAF enabled, if the WBO shows deviation from the commanded AFR, then the MAF table must be out because the VE table was just dialed in.

 

joecar, thanks for chiming in. Good basic logic to what you posted. Back to the basics kind of thing, I need air, fuel and spark to make it run.

I would refer back to my reply to MrDude though. The premise of my theory is the MAF can in fact be accurately calcualated and thus is not an unknown. If this is true, then it means we only have one unknown and can solve the equation.

Back to my original most basic question. If you plan to run a MAF why wouldn't you do all your tuning with it in place working and as part of the equation.

Now MrDude in his lastest threw me one I need to ponder. Car will run without a MAF but not without a VE. I need to science that one out some.

 

Here's some info from over on HPT forums in reply to this same question that might be of interest.

http://www.hptuners.com/forum/showthread.php?t=5242

Read the link within the post to about the sponge theory.

I can buy in to all that's being said but I still have some questions about how this VE tuning might work out if you left the MAF hooked up. Again assuming you've taken the time to calibrate the MAF before hand.

I guess the only way I'm going to satisfy this in my own mind is to try it out. Someone has to be the guinea pig don't they?

Anyway here's my plan, hopefully I'll be able to do this testing tomorrow and report back.

Logging run - Dynamic air and MAF Hz; going to do this one more time for 50+ miles to make sure I've got the MAF as near to dead on as I can get it. I'm going to log histogram data and run it through RHS spread sheet too. Results should be exactly the same but I've got two other sets of data in the spread sheet already so it will give me more reference data.

If I have to adjust the MAF calibration at all then I will reset the LTFTs. If the calibration was already correct I won't bother with a reset.

Logging run - LTFT/STFTs; I’ll also be logging air flow data, tps, MAF lbs/min, etc. to see if I can get enough of the right data to be able to watch things like transition from steady state. If I had to reset the trims I'll try to get at least 100 miles of logging. If I don't need to do a reset then I'll drive until I hit all the cells I want and the STFTs are steady +/-1 or 2.

Modify VE - Change VE tables based on data gathered in the histogram. Reset LTFTs, rinse and repeat until LTFTs are +2/-4.

When I think the VE tune is as close as I'll be able to get it then I'll reset the LTFTs again. Turn MAF frequency to zero and go do another logging run.

Check the log data and see what the LTFTs look like. If they're all still in the +2/-4 range (each individual cell within +/-2) then I would consider this a successful initial test of VE tuning with the MAF active. If they're off, well I had fun burning a lot of fuel and I get to go do it some more to get the VE table where it really needs to be.

 

You realize that dynamic airflow is a PID derived from Mass Airflow and engine speed. You can't tune your maf with your maf

Hope you don't think I'm trying to bust your ***** or anything, as all of us could stand to gain some learning.

 

So are you starting out with a dialed in MAF table and a dialed in VE table...?
(How else did you dial in your MAF table...?)

 

joecar, Not sure the VE is dialed in. It was dead on in SD mode when I was running an IFR for 28.6 injectors even though I had 26.6 injectors installed. So I'm guessing the VE table is not dialed in. I'm assuming the MAF is as dialed in as I can get it but will do one more check on that before I start on the VE tables.

Only way I know how to calibrate the MAF is to log the Frequency vs. Dynamic Airflow and adjust based on results. As I mentioned I've got two sets of log data stored in RHS's spread sheet so I'll use it again.

Muncie, maybe that's the snag in my theory. If you can't calibrate the MAF accurately then this all falls apart. Are there any other fields you could log for MAF calibration? I haven't seen any other way (on the car) suggested.

 

There's no 'direct' way that I know of to calibrate the MAF in HPT.
Calibrating the MAF via the VE table is an indirect way, but you have to have faith your VE is dead on.

 

Yes, agreed, we're not being hard on anyone, just trying to learn something, wrap our minds around something.

 

Well remember I said it would go one way or another. I enjoyed the driving and burning fuel. Short version, I guess you need to turn off the MAF to get your VE dialed in. I've got several hours worth of data with and without the MAF. I need to really do some studying on this but I think I've killed my original theory about tuning VE w/ MAF enabled.

I'll be traveling this week so I'll have some time to really go over this stuff and try to figure it out. I'll try to update some time during the week.

Thanks to all those who've read and shared their input.

 

crewchef, thanks to you also, this is good, it's making us think; keep us posted.