FAST! AE fuel "test question"

[AEkrot]

I have a question about how AE Fuel parameters work:

Car A
AE MAP limit (kPa): 315
dMAP Noise threshold: 255
AE MAP Decay: 2040

Car B
AE MAP limit (kPa): 111
dMAP Noise threshold: 3
AE MAP Decay: 96

Both cars are identical and have a turbo that makes 200 kPa maximum boost.

Assuming the rest of the tune is correct and the same, what is the difference in how the FAST ECU will treat the AE Fuel specified in the various tables between these two cases?

Alex

 

[JeffB]

when you say "200kpa boost" i'm taking that to mean approx 29psi of boost and a total pressure reading of around 300kpa...

in any case:

AE MAP Limit: this sets the highest load that AE will function at... so on Car A, there will be accel enrichment all the way up through the boost range... on Car B, the accel enrichment pretty much stops as soon as boost comes on (actually at about 1.5psi boost)...

dMap Noise threshold: there is conflicting documentation on this one... the FAST online help says the units on this are "kpa", but the software itself says "counts"... i'm guessing that its actually "counts" with the MAP input being an 8-bit quantity (0-255)... either way this parameter is supposed to say "ignore pressure changes less than X when doing accel enrichment"... so if its set to 20, no matter how quickly the load reading increases, if it only goes up 19 (either kpa or counts, depending on what we figure out there) it won't trigger any AE... i've never seen a value as high as 255 for this... it would seem that Car A would basically be disabling all AE with this setting...

AE MAP Decay: this is the rate to "fade out" the extra fuel after the acceleration condition stops... a bigger number gets rid of the fuel quicker... in this case Car A will get rid of the fuel much quicker than Car B...

so in general, Car A seems to be set up to basically eliminate AE fueling, wherease Car B has a pretty "stock" AE setup... the only thing that goes against this is the high AE MAP Limit value for Car A... if they were trying to eliminate AE fueling altogether then they should have set that limit much lower... perhaps Craig can step in and settle the units question on the dMAP noise parameter...

 

[AEkrot]

I was amazed that this post didn't generate more responses.

Car A tune was the reason for my post - Jeff is correct that car B setup is pretty much considered a 'stock' setup - these are the parameters from the example turbo buick demo.gct file that comes with the FAST system.

Car A settings are from Chris Derrick's when he ran an ATI ProCharger F3 at the World Ford Challenge. Rumor has it that Wayne Young tuned it but I haven't confirmed this with Chris or Wayne.

The way I interpreted this tune to work and "why" it was done this way follows:

I agree with Jeff 100% on his explaination of the AE MAP limit - Car A will always get AE fuel no matter what the boost level is.

The dMAP noise value was what I was hoping Craig or someone would explain better. While I agree with Jeff's explanation of this parameter I disagree with his interpretation of the end effect that a value of 255 will have on Car A:

Jeff wrote:
dMap Noise threshold: ...snip... this parameter is supposed to say "ignore pressure changes less than X when doing accel enrichment"...

I agree with Jeff's above explanation.

Jeff wrote: I've never seen a value as high as 255 for this... it would seem that Car A would basically be disabling all AE with this setting...

Based on my observation of how my car is running with this value set to 255, I believe it always allows AE fuel any time the MAP changes in a positive direction as determined by the AE MAP rate of change table. Based on my observation of how the car runs, I think the value of 255 has the same effect as 0 - this might be because of how the FW treats signed and unsigned 8 bit values... Craig should be able to clear this up.

In any event, I believe the reason that these particular values were purposely chosen was for pure throttle response on the drag strip with an over-powered true 10.5" car due to the TPS changing between 100% and 0% often during a pass due to tire spin and close encounters with the centerline or wall. For example, you shift into 2nd at ~100 mph (PG) and the car busts loose spinning. You completely get out of it and then punch it immediately afterwards. The motor will go from vac to 20+psi of boost almost instantly which requires a LOT of extra fuel, otherwise the car will hesitate. I think this is the desired effect of this "tune".

The high AE MAP decay rate supports this theory since the VE table should take care of things once the AE MAP rate of change fuel has done its job.

Comments or other explanations of why Car A tune works and what it was intended to do are very welcome.

Reverse engineering how something works or why someone did something is quite challenging.

Alex

 

[JeffB]

i agree that it would make the most sense if the 255 value was the same as 0 (ie. max sensitivity)... and yes, 255 is the unsigned representation of -1 for an 8-bit quantity... interestingly enough, when i go into CComWP, i cannot enter any values outside of the range 0-30... if i put in 255 it turns it into 30... if i put in -1 it turns it into 0... what version of the software is displaying the 255 for you?

 

[AEkrot]

Chris uses the old DOS version which probably allows the wider setting range.

I read in his tune and haven't "changed" those values on that file. I'm running the WP version and I experience the same thing wrt entering the value of 255 or -1. But if I read in his file and don't change those values, it displays 255 in the WP version.

You must be a "bit twiddler" as well...

With the understanding that 255 acts the same as 0 on this tune, do you agree with my interpretation of why someone would do this?

 

[JeffB]

i'm not really sure that i see the benefit of 0 vs. the standard 2, 3 or 4... my map readings tend to bounce up and down slightly even with continuous WOT... a combination of non-shielded wiring and a $30 map sensor i suppose... with a value of 0 every little jump of 2kpa or so, even if it is actual noise, would trigger AE... this may make it tough to keep a consistent AFR throughout the run... if you are actually "pedalling" the throttle, the kpa readings would be going up and down WAY more than 1-2 kpa so i wouldn't think you'd need this much sensitivity... that's just my opinion though... it would still be nice to hear from FAST...

 

[Craig Smith]

255 and 0 for the noise threshold would have profoundly different effects in the tuneup file. As suggested way up above somewhere, 0 would make it so that any increase in manifold pressure would allow for some shot of AE vs. MAP fuel, and 255 would basically disallow it.

If you use a current version of C-Com or C-ComWP with a current ECU, the max you can set the noise threshold to is 30. With older versions of ECU firmware and/or PC software this parameter is not even there. I think this probably explains why you see no difference between the minimum and maximum settings for the noise threshold. I'll bet that if you take a tuneup right now with current everything, set the noise threshold to 2, do some experimenting, and then change it to 30 and repeat those experiments, you will see a difference in a data log. I know that I did just sitting here doing this on the engine simulator.

As far as a recommendation on how to work with these numbers, I would say to set the MAP AE limit so that no AE vs. MAP fuel can be applied under boost. It can throw things off if the MAP signal bounces around at all under boost, which is not uncommon. I would keep the noise threshold as low as you feel you can keep it without having a noisy signal to the MAP sensor being a problem. Usually between 2 and 5.

It seems to me in reading all these posts that everyone here correctly understands how these parameters work and what they are supposed to do. The confusion, as best I can tell, simply stems from the differences between old and new versions of C-Com. "Old" doesn't even have the dMAP noise threshold parameter in there, so if you look in a "old" tuneup with "new" C-Com it will be an uninitialized variable, and the value of that variable would be 255.

Have I solved this mystery or shall we chat some more?

 

[JeffB]

you've solved it in my mind... the parameters all seem to work as i would have thought they should and the way that i first posted that they worked... it was only the fact that a known "good tune" contained the unusual "255" that caused some other theories to surface... 255 was in effect like a 0, not because of any wierd numbering scheme, but just because the parameter was not being used... thanks!

 

[tazgn]

Craig,
I loaded a new program in my ECU today and noticed that my threshold was set at 255 (probably been that way since I installed the system). So this means I was never getting AE enrichment? That would explain A LOT!!!!!! Why no matter what I seemed to do, if I stepped on the gas quick the A/F would immediately jump to 15.94 (wayyy lean) and then slowly recover after. I reset it to 20 in this program, and now it does just the opposite. A/F quickly jumps down to about 9.8. Damn, now I have more screwing around to do with the curves....

But am I on the right track here? Was this the problem?

Thanks,
Derrick

 

[Chyke]

The answer is "B" the train heading to LA will arrive two hours earlier due to the alignment of the moon. It was a trick question.....

 

[Ormand]

Train to LA?

I got lost at Anaheim. But still, I agree with Chyke. Where would I find a "primer" to read up on the factors discussed above? There have been a lot of these discussions that look interesting, but I don't understand all the acronyms. I can understand why fueling parameters for a supercharged Mustand would be very different from those for a turbocharged Buick, or any other car with a turbocharger and an auto transmission. Just want to read up on the terminology.