MFO Question: Resetting the adaptives?

[CoolVanilla]

A fellow member asked me this in private and I think its a great question, worth discussing publicly. I'll add it to the Q&A as soon as we've come to an understanding here.

Would it be worth it to reset adaptive memory (pull fuse #11) for every test? If one vehicle had -3% on long term and you test an intake that has better air delivery might you push the envelope of the computer?

This has been a lingering question that has played a roll in every MFO to date. As our understanding of these systems have evolved, so too has our understanding of the roll adaptives play. Here is our very best understanding as of today:

Adaptives exist. Both long term and short term adaptives play a significant roll in how our LXs perform. But here's the key point when it comes to our tests: adaptives are in effect at wide open throttle (WOT) only until 3600 RPM.

Buried in our PCM is a setting that specifies an RPM at which the adaptives are to stop being applied. This value is changeable by a tuner via the CMR software; some tuners leave this value alone, deciding to have total control over the WOT table themselves, never having to worry that adaptives will mess with their tuning changes. Other tuners utilize this value, saying that the PCM can best adjust for external changes and thus more often nail peak performance at WOT. The difference of opinion here is well beyond the scope of this discussion, but in my opinion both points of view have significant pluses. Ultimately for my money I prefer the former method; I like knowing that my WOT tune has been fine tuned on a dyno and will always stay consistent in terms of AF, if not in peak performance. But I digress...

All MFO tests have run with non CMR tuned PCMs and/or that default 3600 RPM value has not been touched. So when viewing a chart we produce, keep in mind that the values up to 3600 might have adaptives applied. But after 3600, the PCM stops using adaptives and defaults to a static WOT table. Above 3600 RPM the adaptives are NOT a variable in our tests; all intakes above 3600 are on exactly equal footing when it comes to the (lack of) adaptives.

[Blown7]

Adaptives exist. Both long term and short term adaptives play a significant roll in how our LXs perform. But here's the key point when it comes to our tests: adaptives are in effect at wide open throttle (WOT) only until 3600 RPM.

And you know beyond a doubt that is the only table in all the various Chrysler calibrations whether it be a NGC3 or 4?

I'm not bashing it's just I myself don't know all the calibrations and when adaptives run and store.

[CoolVanilla]

And you know beyond a doubt that is the only table in all the various Chrysler calibrations whether it be a NGC3 or 4?

I'm not bashing it's just I myself don't know all the calibrations and when adaptives run and store.

No sir! You're looking at things from a 100ft view and no question when you get down there your concerns are dead spot on.

But as you point out, can we know for sure? Or are we limited by the tools we have available to us? And that tool is currently the CMR software and the CMR software tells us there's 'a value' that defaults to 3600. Follow me?

For our purposes I maintain we can view things from the 1000ft or maybe even 10000ft view, and still make some definitive statements (like I did above). Definitive enough, anyway, to enable us to run our tests and draw some confidence from our results.

[JonW]

I've recently done some DYNO time with my new Vortech setup and I can tell you that it is changing while on the DYNO because of adaptives. First and second pulls are usually 504 down to 501 and then 3rd pull and beyond stays around 495. I've repeated this scenario about 3 times so far.

[CoolVanilla]

Jon, what makes you positive its the adaptives alone? Could it be temperature related? What kind of dyno? Even the best dynos have an error of around plus/minus 1%. 1% of 500hp is 5hp. Meaning you could swing 10hp from worst to best and it all be in the error of the dyno.

Just throwing it out there. My experience over maybe 600 carefully studied pulls says its far more likely temperature related (heat in the oil, coolant, trans fluid, diff fluid, etc) and/or dyno error swing than anything else.

[CoolVanilla]

PS: Another thought pointing at it not being the adaptives, the adaptives we're talking about are the long term adaptives. They take a good long while to formulate themselves and to have an effect (how long, I don't think anyone knows, but my experience says its easily in the dozens of miles range). You're not on the dyno long enough for those adaptives to have an effect.

Again just more thoughts...

[JonW]

This was on a Dynojet. All the runs were back to back when the car was already at full operating temperature (around 190). I had my fans on plus a huge fan in front of the car and the hood up. The temp did not fluctuate very much. I'm not saying it couldn't be something else, but after reloading the tune and doing it again, it sure seemed consistent to me.

[JonW]

PS: Another thought pointing at it not being the adaptives, the adaptives we're talking about are the long term adaptives. They take a good long while to formulate themselves and to have an effect (how long, I don't think anyone knows, but my experience says its easily in the dozens of miles range). You're not on the dyno long enough for those adaptives to have an effect.

Again just more thoughts...

What about short term adaptives? Yes, the LT's need significant driving to set them, but it is cell based and that is to hit all possible cells multiple times. On the dyno, you're only hitting a couple of cells - repeatedly. I'm no expert on this - it's just a thought.

[CoolVanilla]

Gotcha. Did you log oil temps Jon? I'm willing to bet you'll see a steady increase in those temps from run to run, even though your ECTs were consistent. Thats something we specifically looked out for at this MFO, and we found it pretty difficult to balance oil and ECTs actually. We found (in 40deg weather) that around 7 minutes between pull was required to get them *close* to the previous run.

[seanol]

I have a few runs under my belt as well and when testing centrifugal superchargers they are very susceptible to heating up the intake path. Does the number stay at the llower levels aftyer the initial passes? Log your IAT and see if you reach a soak point.

I don't know how the adaptives work with the Hemi and I also don't know what the scalars are for timing and fueling at WOT. Voltage and IAT can have a very large impact on the top end of the curve.

When testing anything for gains I have always found it best to start at a number you can get to (180deg coolant, 200 deg oil, whatever) and make three runs.

Look at the runs and you will find that runs in the series will be comparable to other runs in the series I.E. 3 runs 001, 002, 003 compared to runs 004, 005 and 006. Compare 002 to 005 for greatest accuracy.

Not running in the vehicle is one of the bigger reasons for poor repeatability. It takes some time to get the rear gear oil up to temp. When I ran dry sump cars we would have to run up for about 5 to7 minutes with the Oil sump at 180 deg to circulate the oil in the motor and we routinely heated the rear diff oil to 180 deg as well.

As for the fan, as long as you use the same fan all the time and you have the dyno intake temp probe reading inlet air you will be fine. A dyno can't give a final HP figure but an inertia dyno and a good operator can give a very repeatable (within the ability of the vehicle) result needed to find the Next Great Thing!

Hope this helps,
Sean

[CoolVanilla]

What about short term adaptives? Yes, the LT's need significant driving to set them, but it is cell based and that is to hit all possible cells multiple times. On the dyno, you're only hitting a couple of cells - repeatedly. I'm no expert on this - it's just a thought.

To my understanding, STs affect part throttles, not WOT, directly. To my understanding they're used in calculating LTs. LTs are used in the WOTs until 3600.

By all means someone correct me if I'm wrong about that. I don't have ANY definitive proof. These are just my thoughts based on observation and discussions with the "expert"s.

[CoolVanilla]

I have a few runs under my belt as well and when testing centrifugal superchargers they are very susceptible to heating up the intake path. Does the number stay at the llower levels aftyer the initial passes? Log your IAT and see if you reach a soak point.

I don't know how the adaptives work with the Hemi and I also don't know what the scalars are for timing and fueling at WOT. Voltage and IAT can have a very large impact on the top end of the curve.

When testing anything for gains I have always found it best to start at a number you can get to (180deg coolant, 200 deg oil, whatever) and make three runs.

Look at the runs and you will find that runs in the series will be comparable to other runs in the series I.E. 3 runs 001, 002, 003 compared to runs 004, 005 and 006. Compare 002 to 005 for greatest accuracy.

Not running in the vehicle is one of the bigger reasons for poor repeatability. It takes some time to get the rear gear oil up to temp. When I ran dry sump cars we would have to run up for about 5 to7 minutes with the Oil sump at 180 deg to circulate the oil in the motor and we routinely heated the rear diff oil to 180 deg as well.

As for the fan, as long as you use the same fan all the time and you have the dyno intake temp probe reading inlet air you will be fine. A dyno can't give a final HP figure but an inertia dyno and a good operator can give a very repeatable (within the ability of the vehicle) result needed to find the Next Great Thing!

Hope this helps,
Sean

Sure does Sean! Glad to see we're on the same page. Again this is why we went out of our way this time to nail ECT and oil temps. Consistency for both was our goal and we did the best we could under the circumstances (the 40deg shop temps allowed us to do so this time, where when the shop temps are in the 110s I dont think we could manage it given the time gun we're always under). And I think that result is pretty informative, actually.

[monty1269]

Morpheus would know.

?

[seanol]

Jason,
I agree, you nailed it!

This is an unbelievable service to the LX community and you, your staff and SVS should be commended for all the hard work and time to put it all together.

This has been more professional than a lot of the organizations I have seen attempt this type of test. It is very easy to color a test to show a given result even if the bias is unconscious.

While the experiment theory may be simple in description, it's a bear in execution!

I think your earlier statement in one of the other threads said it best and that is look at the area under the curve. This is real and repeatable.

Peak is for salesmen, not racers.

Thanks again and look forward to MOFO7 "Thunder in the Tunnel"!

Sean

[Frozen_]

and what about KR? i swear just by looking at the dyno sheets some KR was def there.