DME Motronic - O2 sensor questions

[Fast951]

The 3.2L did not come with a MAF (Mass Air Flow), it came with a VAF (or Air Flow Meter - AFM; or barn door). If the VAF is pegged, why not convert to a Mass Air Flow (MAF) which if sized correctly will not max out and you get the flow even at higher RPMs.

[Fast951]

Originally Posted by hcoles

yes, by all means... I think Steve W. sells a system that has all that figured out. Actually measures the maf up to the limits of what the engine can eat and adjusts the AF to what you want... there's a concept... that apparently was missed back in 1983 or so by Porsche.

Originally Posted by hcoles

when I say Steve's product I refer to
http://www.911chips.com/HFM1.htm
This is what the factory should have done IMHO but the technology may not have been there quite yet.... If I had an extrie $2,700 I would buy this in a second.

I'm not familiar with this particular system.
There are other True MAF systems on the market for less than $1300.

[ischmitz]

Well, both a barn door (metering air volume) or hot film MAF (metering air mass) work while the engine is under part-throttle conditions. The MAF has a couple of advantages: Mainly, it is less restrictive, quicker in its response time, and has no moving parts. On the down-side the filament on earlier types can get contaminated and/or can burn out. Newer versions have an auto-clean function that burns contaminations off. However, while a MAF is a better system during fast load changes in part throttle it really doesn't do anything for your mixture under WOT or highway cruising.

Beyond a certain point (on the BOSCH DME it is 58% load) the amount of fuel is derived simply from the volumetric efficiency of the engine multiplied by the RPM (corrected for intake temperature). The metering device is only a restriction in the intake path at that point. Only its NTC or PTC reading goes into the calculation of fuel.

Cheers,
Ingo

[Fast951]

Ingo, With all due respect, I disagree with you. The AFM signal or (MAF with appropriate DME software) is used even at WOT!

[ischmitz]

Fast951 - I guess it depends on which system you are talking about. I have worked on only two, Bosch and Megasquirt. Megasquirt does not even have a WOT switch implementation and simply uses a 2D map for the fuel calculations all the time - even under boost. Load is simply determined by intake manifold pressure. However, the Bosch Motronic code on the 3.2 and 3.6 has dedicated WOT tables and the WOT switch.

There are several reasons why I believe (and don't know for sure) that the Bosch system does not use the metering reading at WOT:

1. The WOT table in the Bosch system is 1-dimensional: fuel = f(RPM) while the part throttle tables are 2D: fuel=(RPM, load). The axis definition on the Bosch code is pretty straight-forward. Calculating the bin labels shows that the 2D map load axis covers a range from 10% to 58%. See an example here of a 964 Motronic map:



2. If the Bosch Motronic was to use metering readings beyond the 58% level they would have extended that 2D map.

3. I wonder how the DME would use information from the metering device once we it is past the 58% level. Maybe there is another part in the algorithm where the metering output flows in as a linear factor. While I don't know this for certain I strongly doubt it. It would make remapping a car from a linear to a logarithmic metering device nearly impossible without re-writing entire code segments in machine language. The RPM information should be sufficient to determine how much fuel is needed.

4. Thinking back to CIS here the metering device has no information about engine speed and thus has to rely entirely on the CIS sensor plate to determine fuel quantities. My conclusion is that one or the other is good enough at WOT. This is why the WOT switch was implemented. It makes tuning for performance much easier.

It is my understanding that the Bosch unit pretty much uses a speed-density algorithm (metered air, RPM, pressure, intake temp.) for part throttle and reverts to an Alpha-N algorithm (throttle position, RPM, intake temp.)

I guess the ultimate answer would be to run a car on the dyno and simulate a low AFM reading while the WOT switch is engaged. I wonder if the mixture would change......

Cheers,
Ingo

[Fast951]

Ingo, We are talking about the Bosch Motronic (3.0, 3.1) as used on the 911 3.2L, 944/951 some BMWs.. I do know for a fact that the code uses the VAF signal under WOT.
We actually rewrote the Motronic code. We have used a MAF on the 911 for years, we change the DME code and not use a signal massager to convert the MAF signal to look like the AFM signal so the DME understands it.
I'm not here often, but if you look up some of my posts on the RennList (951 forum, screen id: fast951), I have a few posts that will clarify things. Email me directly if you have any questions.

[JasonAndreas]

Originally Posted by ischmitz

However, the Bosch Motronic code on the 3.2 and 3.6 has dedicated WOT tables and the WOT switch.

There are several reasons why I believe (and don't know for sure) that the Bosch system does not use the metering reading at WOT:

The three throttle states; idle, partial and wot are only correction factors. The base injection times are still calculated from the AFM and are then modified by the correction maps.

Originally Posted by ischmitz

2. If the Bosch Motronic was to use metering readings beyond the 58% level they would have extended that 2D map.

If you were to express the right side column of the partial throttle (fuel & timing) correction maps as a fixed percentage of load (Bosch doesn't) it would not be 58% (you don't divide by 255) and the left side is not 10%. All of the PD/Freeware and most of the commercial editing software that I've seen is wrong when it comes to units of measurement.

[scarceller]

Originally Posted by ischmitz

I guess the ultimate answer would be to run a car on the dyno and simulate a low AFM reading while the WOT switch is engaged. I wonder if the mixture would change......

Cheers,
Ingo

I have a wideband O2 monitor on my 3.2L and have done some testing. I have logged RPM against AFR at WOT and have done 2 runs:
One run was with the AFM in stock config mode and I have a AFR of about 13.2 across the RPMs 3-6K.
Then I slid the pointer up in the AFM (making the AFM read richer) and this drove the AFR down to 12..5.
I do not know that much about the chip program but making the AFM read richer drives the WOT AFR richer, this much I know from testing.