Simple MAF/DYNAIR pid question.

[acomp917]

RHS,

About the Me and Jeff post.

I mapped SAE.MAF/GM.DYNAIR. The ONLY discrepancy was during tip in/out. This indicates either the VE needs more slope(inversely prop. to map) or the SAE.MAF is not using the TauX (wall wetting) parameter. I have yet to think enough to check that. I guess I could bash the TauX and see what happens. A GUESS IS MAF IS NOT USING TAUX.

Jeff, 2 days later: well, maybe MAF alone cannot account for every change that occurs in the airflow or fuelflow, so I'll start changing IFR to make up for it!
Wonder what he found exactly?



me: so you're adjusting IFR to make up for the MAP/MANVAC based changes. Why don't you also add RPM based changes to it, and just call it the VE table?

You sure have a problem with IFR adjustments. I have been an electronics tech for 30 years and a mechanic since a child. I know that balancing an inductive load considering vac differential/transient and moving mass is going to take a very long time. According to what I know of your standards... IFR is basically crap. For a give (constant) situation, I think I'll trust the MAF calibration.

S

[acomp917]

Shawn,

Changing the IFR will just change the entire VE table a corresponding % directly proportional to IFR. I kept it simple, that assumes changing the entire IFR a given %. Of course the VAC.psi(stupid) and VAC.kPa would have to be converted.

I don't know why you say all of the air flow settings must be changed. All that was done was an adjustment to the calculated fuel mass. I have to ponder this. My trk is still re-learning, I will check this when it stabilizes.

^ Ah Ha, My intention was not to change the VE overall, it was to change the LTFT from +6 to ~-2. The contour will change to reflect timing changes and AFR changes under PE. GM's tune was timing and fuel safe. If a person has .5 brains... they could not make it much worse.

There is no mystery to the basic VE table. It is just a 3d lookup table(with interpolation) that reflects the engine's ability to fill the cylinders, based on RPM(speed) and VAC(density)... hence SD. Of course the retrieved number is but one part of a much bigger algorithm. That is the mystery when dealing with these dam GM systems.

S

[acomp917]

A reconsider,

About that not being able to wreck a factory tune... Has anyone looked into some the modified tunes on holden crazy? Brings to mind the anvil and rubber hammer analogy. Does anyone think those tuners ever looked at the fuel tables after they were modified? Looks like non-filtered autoVE and NO knowledge.

S

[WeathermanShawn]

No, I understand what you are doing.

IMO, a 10% change in your Injector Flow Rate (IFW) is no different than a 10% across the entire MAF Calibration Table. Of course, with a IFW adjustment there is no load adjustment..spark tables. MAF there is.

I can run a 2002 stock Camaro VE Table in my car (other than idle) with similar Trims, Commanded AFR, etc., as long as my MAF is calibrated. Where all of this gets fun is when you purposely fail the MAF. If you have a 20-30 % airflow difference, I bet your car won't run. So, no I am not faulting you. I am just preparing you for the answers you are about to hear from others. You have to occasionally fail your MAF to at least prove you have the correct airflow.

I know you are a smart guy with a background in mechanics. Others without the experience may want to try the same thing. Just be prepared. I barely challenged current thinking. And, I was not right about everything. Saying IFR adjustments on this forum will get you a lot of responses.

[acomp917]

Does anyone have an accurate dyno based tune for a stock LM7 5.3 truck. I'm not asking for the data, just a sanity check. I believe that hump in the VE should be pushed toward the lower RPM's. I know everything above 6k is a waste. This is the first hack. Waiting for the truck to re-learn before I can really find it's true personality.

Any opinions?

S

[joecar]

...

Changing the IFR will just change the entire VE table a corresponding % directly proportional to IFR. I kept it simple, that assumes changing the entire IFR a given %. Of course the VAC.psi(stupid) and VAC.kPa would have to be converted.
...

S,

No, changing the IFR leaves the VE table as is (you have to edit the VE to change it)...

Changing the IFR changes the fuelmass delivered by the injectors, changing the measured AFR.

Same applies to IFR in relation to MAF table.

Also, as Shawn pointed out, changing the IFR doesn't change the load (which indexes into the spark tables)... changing the VE or MAF changes the load (airmass/cylinder).

[acomp917]

Shawn,

I don't mind, I am a challenged thinker I mean I don't mind being challenged by thinkers. About that modifying load/spark tables: I have not found a basic running table that reflects spark with reference to MAF. I believe "load" is a function of MAP vs. RPM, ie. VE based. Would anyone please educate me about this if I am wrong?

Found my oversight.
joecar,
First, my writing is not very good.

No, changing the IFR leaves the VE table as is (you have to edit the VE to change it)...
I should have stated in order for fueling to remain constant, VE will need to be changed...

Changing the IFR changes the fuelmass delivered by the injectors, changing the measured AFR.
IFR changes affect the calculated fuel mass and only affect AFR in an OL system, else LTFT.

Same applies to IFR in relation to MAF table.
Don't understand your statement, though I do understand they are ALL relational.

Also, as Shawn pointed out, changing the IFR doesn't change the load (which indexes into the spark tables)... changing the VE or MAF does change the load... if the VE or MAF are correct, you get the correct spark timing (if the spark tables are correct).
I think I understand how load is calculated(correction above, maybe here too). F!@# the timing table values... GM did. just kidding. I do know what you mean. Remember, we should try to maintain atleast one constant(I KNOW! I'm the person that neglected to do that ).

S

[acomp917]

A reconsider,

My apologies to joecar and Shawn. I did find that the spark table is referenced to GM.DYNCYLAIR vs. RPM vs. timing. This is the first system that I've tuned that used that g/cyl axis. As a matter of truth, this is the first system with a MAF that I have tuned.

Stands to reason. This allows the MAF to have access to spark timing... Now for the big question! What does a CL SD/MAF system use to determine base spark. I guess GM.cylair(MAF) and GM.DYNCYLAIR(SD). Any blending, precedence, or other?

While I'm askin... does it apply any additional modifiers? Oh yea, there is entire list. This thing will FU UP!

Gotta have somethin to do,
S

[joecar]

See my inline comments in this strange blue ink:

...

No, changing the IFR leaves the VE table as is (you have to edit the VE to change it)...
I should have stated in order for fueling to remain constant, VE will need to be changed...
ah, ok, yes that is correct;

Changing the IFR changes the fuelmass delivered by the injectors, changing the measured AFR.
IFR changes affect the calculated fuel mass and only affect AFR in an OL system,
changing the IFR changes the pulsewidth calculation which changes the delivered fuelmass, it does not change the calculated fuelmass (which is calculated from airmass and AFR);

else LTFT.
my understanding is this: in CL, the PCM first computes the required fuelmass and then uses O2 feedback to trim to stoich (i.e. the trim adds/subtracts fuel on top of the calculated fuelmass).

Same applies to IFR in relation to MAF table.
Don't understand your statement, though I do understand they are ALL relational.
If you change the IFR the MAF will also have to change to keep fueling the same.

Also, as Shawn pointed out, changing the IFR doesn't change the load (which indexes into the spark tables)... changing the VE or MAF does change the load... if the VE or MAF are correct, you get the correct spark timing (if the spark tables are correct).
I think I understand how load is calculated(correction above, maybe here too). F!@# the timing table values... GM did. just kidding. I do know what you mean.
Yes, cylinder airmass is considered to represent load... this can be found in most ICE textbooks;

Remember, we should try to maintain atleast one constant(I KNOW! I'm the person that neglected to do that ).
No problem... in order to learn we must take a few steps and possibly make mistakes... it's the mistakes we learn our best lessons from...
we're all still learning this stuff... learning is like a closed loop algorithm that requires a few occasional open loop excursions.

[WeathermanShawn]

Every big of logging I have done with MAF-Enabled has used the MAF calculated airflow (GM.CYLAIR). I know Table B5913 says "DYNCYLAIR", but if you hold the cursor over it, it lists multiple PIDS. I have not seen any influence of DYNCLYAIR affecting spark load when running MAF.

When I tested the IFR % change, my fueling (TRIMS) were fixed, but my load stayed the same, and my DYNCYLAIR and CYLAIR airflow values were off by the same % I had applied to the IFR.

The last week I have been running CLSD to compare against a MAF-based tune. With the MAF disabled, Table B5913 references GM.DYNCYLAIR. Of course with MAF disabled, I am running off the Low Octane Spark Table. With a custom OS, you have more options.

Those are some of the issues you face with a IFR adjustment. Sometimes Idle and decel can be 'rich' when IFR and airflow do not match. Meanwhile Trims and Commanded AFR look good. Something to think about.