Stock FR-S Tune Just Done

[Zach3794]

Trying to answer all questions at once here. I use Shell 91 exclusively for fuel. If you'll look at the grey line, that was from the first batch of our cars to hit US shores. The tuner said it looks like they changed the way they tuned the ECU for my batch to be more conservative. He uses the ECUTEK software. When I go this Saturday I will get the before and after AFR curves printed.

Finally, the numbers read low because as stated, this dyno is the queen of Heartbreakers.

And Tim, we are both in Vista! Let's go drive sometime soon!

[Zach3794]

Originally Posted by Xdragonxb0i

Looks good, a tune really unlock the potential, Theres probably some discrepancy in the first stock run, its really low. After a tune i expect our cars to be in the 160-175whp range depending on the fuel and aftermarket parts. 200whp for e85.

I'm not sure what you mean by really low for the "first stock run".

As I explained above, the pink line is my car before the tune, and is the middle/average of 3 runs, which were all very similar.

The grey line is an entirely DIFFERENT FR-S, stock, no tune. The only reason I included it is because it was done on the same dyno, but this FR-S was one of the first to hit U.S. shores. I am including it here because when you compare the curves, it's not just a difference in power, it's a difference in how the curve actually behaves. This suggests that at some point, the factory tweaked the ECU tune to what it is now, my car, pink line.


As for the numbers being low in general, I will again remind you all that this is first and foremost a Dyno Dynamics, which already reads lower than a Mustang. Secondly, this is a much older model Dyno Dynamics, which reads even lower than the more recent ones. What we end up with are numbers that are MUCH lower than the industry standard Dynojet would read.

I hope this clears up some of the confusion.

[Tim_Asphalt_FRS]

Originally Posted by Zach3794

And Tim, we are both in Vista! Let's go drive sometime soon!

I may have seen you driving around. Saw a Raven last week near RBV. You?


Sent from my iPhone using Tapatalk

[ST185RC]

I'm liking how the torque dip got narrower and the initial delivery of the torque below 3K rpm has flattened out a bit.

Anxious for AFRs now.

[brzmaybe]

Originally Posted by RandomHero

That's very impressive all stock gains. Perhaps you had a faulty sensor and its reading got tuned out? Otherwise why would this car be drowning itself in fuel up near redline? The O2 sensor should have been shouting shut the f***ing fuel off to the ECU lol

Either way that's stellar performance my man.

OK, I'll bite...

Given that there's no free lunch...

Aside from the good money put down to achieve the tune, what other characteristics of the stock tune were sacrificed to enable those nice gains? That's a back-door way of wondering why Toyota didn't offer the performance tune from the factory?

[Kodename47]

Originally Posted by Zach3794

Trying to answer all questions at once here. I use Shell 91 exclusively for fuel. If you'll look at the grey line, that was from the first batch of our cars to hit US shores. The tuner said it looks like they changed the way they tuned the ECU for my batch to be more conservative. He uses the ECUTEK software. When I go this Saturday I will get the before and after AFR curves printed.

Is it not possible that the ECU's learning cycle put it through some crap parameters that caused it to perform badly. I very much doubt that a revised manufacturers ECU would come out with any less power, let alone that big a drop. Imagine if that made the news on such a highly publicised car!! Seeing as the Torque line starts similar then just falls off a cliff then it would suggest that during the car's lifetime poor grade fuel and/or something else that has caused the ECU to restrict power.

TBH what matters is the end result that is good!

[robispec]

Get in line quick!

Robi

Originally Posted by Tim_Asphalt_FRS

I may have to give him a visit.


Sent from my iPhone using Tapatalk

[suaveflooder]

WOW! That is an amazing difference! I might have to take a trip down there. Thanks for posting!

[WatchmaN]

Drive the car 2 weeks and go back to dyno, and see what happens to your gains, I will be posting very interresting discovery soon about how the ecu react to differant tunning avenue.

[sw20kosh]

Originally Posted by Zach3794

I'm not sure what you mean by really low for the "first stock run".

He is talking about the shape of the curves. Your "stock" runs are not typical. Typical stock runs don't plummet that soon and that drastic. It has nothing to do with absolute numbers but rather the shape of the curves.

Something was going on with your stock tune. That is why the new tune looks so good.

[Shankenstein]

Originally Posted by WatchmaN

Drive the car 2 weeks and go back to dyno, and see what happens to your gains, I will be posting very interresting discovery soon about how the ecu react to differant tunning avenue.

Could you give any more specifics?

Modern ECUs have a fair number of adaptive variables, and any number of them could increase or decrease the tune's effectiveness. Most require specific circumstances to be met before the adaptive algorithm starts. From there, it's usually a "prod" that produces a response. The real response is compared to the expected response, and table values (coefficients) are adjusted so that it better fits the real response values in the predicted response's shape.

AFR_stoic is a major one. If you're tuning it to always run rich, the computer will slowly adjust the target AFR to match it's perceived stoichiometric value. Since we don't have ethanol sensors, it's calculated recursively from the wideband O2 values during certain operating conditions. The coefficients on the prediction algorithm are slowly adjusted until the mean-squared-error (or whatever they use) is minimized or the operating conditions move out of range.

Spark timing is another main one. Knock sensor feedback is pretty crappy, but with enough filtering and careful adjustment you can design a spark timing controller that adapts.

[ahausheer]

Is it possible the stock runs were not wide open throttle in the higher RPM range thus causing the curve to drop off? Or does that not even matter?

[WatchmaN]

Originally Posted by Shankenstein

Could you give any more specifics?

Modern ECUs have a fair number of adaptive variables, and any number of them could increase or decrease the tune's effectiveness. Most require specific circumstances to be met before the adaptive algorithm starts. From there, it's usually a "prod" that produces a response. The real response is compared to the expected response, and table values (coefficients) are adjusted so that it better fits the real response values in the predicted response's shape.

AFR_stoic is a major one. If you're tuning it to always run rich, the computer will slowly adjust the target AFR to match it's perceived stoichiometric value. Since we don't have ethanol sensors, it's calculated recursively from the wideband O2 values during certain operating conditions. The coefficients on the prediction algorithm are slowly adjusted until the mean-squared-error (or whatever they use) is minimized or the operating conditions move out of range.

Spark timing is another main one. Knock sensor feedback is pretty crappy, but with enough filtering and careful adjustment you can design a spark timing controller that adapts.

Well I'm not gonna give away the punch of the thread I will make, But I can say that I took, Perrin tune, FA20club, and stock tune. Did runs right after the flash pass them trought virtual dyno, then did the same things 2 weeks after and the result were shocking.

I'm not a tuner to analyse why it did what it did, so tuners will be invited(PM) to comments and analyse.

[Zach3794]

Originally Posted by Tim_Asphalt_FRS

I may have seen you driving around. Saw a Raven last week near RBV. You?


Sent from my iPhone using Tapatalk

I drive past RBV every weekday coming home from work