To Polish up AFR

[gr8vet]

When tweaking AFR in non WOT areas, do you guys adjust in VE or MAF. Do you go back to the VE and keep polishing or just do a MAF adjustment and let it be?


Thanks in advance
tt

[kamikazematt]

i too would like to hear some opions on this.

[WS6FirebirdTA00]

It depends, if you are in SD then change the VE. If you are MAF and the VE is good, change the MAF. If you are MAF and the VE is bad, fix it then do the MAF again. MAF is used above 4k, so if it is off above there then change the MAF.

[eficalibrator]

You can change the thresholds to force it to use either MAF or VE exclusively. In a perfect world, you would do a complete calibration in each mode independent of the other. Within either mode, the delivered AFR should be within a couple percent of the commanded AFR before any closed loop trims. If you get this right in both MAF and VE modes exclusively, then you will end up with really good operation when you restore it to normal "blended" operation as it was from the factory.

[viperbluelx]

You can change the thresholds to force it to use either MAF or VE exclusively. In a perfect world, you would do a complete calibration in each mode independent of the other. Within either mode, the delivered AFR should be within a couple percent of the commanded AFR before any closed loop trims. If you get this right in both MAF and VE modes exclusively, then you will end up with really good operation when you restore it to normal "blended" operation as it was from the factory.

Is there a reason GM made it a hybrid system instead of MAF only like Ford?

What you're saying makes great sense, I'll have my HPT here in a few days and I'll have to look into changing the thresholds. Can you change the settings and make the car MAF exclusively from idle to redline? Is there a problem with leaving it that way?

[WS6FirebirdTA00]

The problem with MAF only at low RPM and transients is it's ability to accurately read the airflow. Low flow conditions can cause erroneous MAF readings. The MAF is just a flow meter so all the rules still apply for accurate flow readings. It may do a better job at transients than I think, but I am sure Greg may be able to comment on this a little further.

[eficalibrator]

Simply put, GM opted for the best of both worlds. A MAF is great for steady state conditions and high flow rates (within the range of the sensing element/table), so it gives a really good value at cruise and WOT. The MAF is really nice since it's already compensated (for the most part) for weather, temperature and baro changes without extra exotic math in the PCM.

During transients, the "spring" of air between the valves and air filter acts like a Slinky toy so to speak and can mislead the MAF if you're not careful. In this case, a speed density system based on a more stable MAP reading works better. So anywhere the delta-MAF or delta-pressure is large enough, they look to the SD model and in places where it's stable, they look to the MAF.

[viperbluelx]

Simply put, GM opted for the best of both worlds. A MAF is great for steady state conditions and high flow rates (within the range of the sensing element/table), so it gives a really good value at cruise and WOT. The MAF is really nice since it's already compensated (for the most part) for weather, temperature and baro changes without extra exotic math in the PCM.

During transients, the "spring" of air between the valves and air filter acts like a Slinky toy so to speak and can mislead the MAF if you're not careful. In this case, a speed density system based on a more stable MAP reading works better. So anywhere the delta-MAF or delta-pressure is large enough, they look to the SD model and in places where it's stable, they look to the MAF.

with that said couldn't you enable MAF at say 1,000rpms which wouldn't effect idle. Then use the RAF table to dial in a consistant idle and let the MAF do all the rest or is this overcomplicating things?

[WS6FirebirdTA00]

It's oversimplifying is what it is. 1000 RPMs is low flow and there are too many transients at that low of an RPM range.

[eficalibrator]

with that said couldn't you enable MAF at say 1,000rpms which wouldn't effect idle. Then use the RAF table to dial in a consistant idle and let the MAF do all the rest or is this overcomplicating things?

RAF is a setpoint, not a measurement. The actual measurement that confirms that the engine is operating near the setpoint (RAF) at idle comes from either an SD model based on known VE (lookup table) and a MAP reading, a direct MAF measurement, or a blend between the two.

It's entirely possible to make a GM vehicle operate in strictly MAF state by setting the RPM threshold to something lower than idle. It essentially operates like a modern Ford EEC like this. I do this when calibrating the MAF curve to avoid confusion with the VE model and tune as far down the curve as possible. After I've calibrated the MAF curve, I then repeat in speed density for the same region(s) to get that model correct as well. It's a bit time consuming, especially with the newer equation based GMVE models, but the end result is better.

[viperbluelx]

Good to know, it seems that MAF would be much easier to dial in to get the car in a nice running driving state. So I'll start with that and then dial in VE.

[WS6FirebirdTA00]

No, start with the VE then go to the MAF. If the VE is off under the MAF threshold, there is a blend of the two and it will throw off your data.

[viperbluelx]

No, start with the VE then go to the MAF. If the VE is off under the MAF threshold, there is a blend of the two and it will throw off your data.

I was basing it off the following.

After I've calibrated the MAF curve, I then repeat in speed density for the same region(s) to get that model correct as well.

[eficalibrator]

No, start with the VE then go to the MAF. If the VE is off under the MAF threshold, there is a blend of the two and it will throw off your data.

Or you could just set the threshold to a speed lower than idle and have ZERO blending from the VE model... It's almost always easier to start with the MAF since that falls into place really quickly if you've got the right fuel injector data.

Then, if you're really slick, you can calculate VE based on actual MAF, MAP, rpm, and IAT and plot that in a histogram against RPM and MAP to "auto-generate" the VE table...

[viperbluelx]

Or you could just set the threshold to a speed lower than idle and have ZERO blending from the VE model... It's almost always easier to start with the MAF since that falls into place really quickly if you've got the right fuel injector data.

Then, if you're really slick, you can calculate VE based on actual MAF, MAP, rpm, and IAT and plot that in a histogram against RPM and MAP to "auto-generate" the VE table...

I haven't got to the GM section of your book yet, but so far it's a really good read.

I wish I was slick enough to calculate VE based on the other stuff. Maybe Marcin will make a spreadsheet to do it all.

[WS6FirebirdTA00]

A MAF is great for steady state conditions and high flow rates (within the range of the sensing element/table), so it gives a really good value at cruise and WOT.

Then why did you say this? I don't know any of the detailed specs on the MAF sensors, but I assumed that they would need a little higher velocity through them to be accurate at idle. Not only that, but there are a lot of slight transients I think would skew the data at that low of an RPM. I see your point, but my comments were based on the assumption of the inaccuracy of the flow meter towards lower airflow. From the sensors I use of that size (not on cars), ~4-5 ft/sec is a low velocity to rely on the data. JMO.

[eficalibrator]

The MAF sensor can and should still be able to get decent data at idle. Look at any Ford vehicle, for example. They use almost the exact same Hitachi slot MAF element on the 05+ Mustang as the LS3/7. The only difference is that one is analog (0-5v for Ford) and the other is digital (0-15,000Hz) output to the PCM.

If the engine speed and actual airflow are stable (even at a low number), the MAF should still give a decent reading as long as there isn't any noise from standing waves in the inlet tract. This is why you see the Helmholtz resonator tubes/chambers on inlet systems, BTW.

The engine just has the ability to rapidly differ from the stable idle reading if the throttle is opened/closed quickly. *THIS* is what's difficult to correlate with MAF data as the column of air acts like a slinky/spring and potentially shows a different value at the MAF vs at the cylinders for a brief period.

If everything is stable: MAF airflow = VE airflow = cylinder airflow pretty much anywhere.

I haven't got to the GM section of your book yet, but so far it's a really good read.

I wish I was slick enough to calculate VE based on the other stuff. Maybe Marcin will make a spreadsheet to do it all.

I didn't put that exact equation in the first book, but I cover it in my Advanced GM classes.

I will also have a variant of that equation in my next book since that will focus much more on speed density systems. That should be on the shelves sometime later this fall.

[Gh0st]

I am running OLMAF right now because I am about to upgrade injectors.

I can tell you right now I have no issues with fueling and tuning the MAF always seems to be much easier for me than working with SD(which is affected by the weather heavily ).

I love how MAF gives me the same fueling no matter the weather(or so it seems!) and my idle is beautiful with it.

Upgrading injectors tomorrow, will be starting with a full MAF tune, then I will work on the VE when time permits.

[viperbluelx]

I am running OLMAF right now because I am about to upgrade injectors.

I can tell you right now I have no issues with fueling and tuning the MAF always seems to be much easier for me than working with SD(which is affected by the weather heavily ).

I love how MAF gives me the same fueling no matter the weather(or so it seems!) and my idle is beautiful with it.

Upgrading injectors tomorrow, will be starting with a full MAF tune, then I will work on the VE when time permits.

so right now you're running MAF only, correct? Why open loop? If you've got MAF coming in below idle and the MAF transfer is correct the 02's should work properly, correct?

[Gh0st]

so right now you're running MAF only, correct? Why open loop? If you've got MAF coming in below idle and the MAF transfer is correct the 02's should work properly, correct?

Because I feel like it. Once you mod a car, 14.7 AFR isn't always the "magic" number. OL let's you do what you want.

I was in the middle of redoing my airflow tables and said to hell with it I'm getting new injectors, i'll just chill in OLMAF for another week.