daveak05

they wouldn't be able to sell it if it only claimed 10 HP : )

RI A6

I understand that results can be manipulated. Lets assume that in a specific set of tests on the same day, on the same car, under the same atmospheric conditions that they are not, and the dyno operator performs dyno runs in good faith using the best practices.

Now, given this, what say you about the relative and absolute accuracy of multiple dyno runs on exactly the same engine chassis configuration, and a modified configuration?

Bob W.

choose a MB shop to tune an Audi...great idea.

daveak05

just because the car is tested same day, under same conditions, doesn't make HP reading accurate. it will still be affected by the conditions(as opposed to other conditions elsewhere) and the car itself(is it running optimally?).

the only true HP and torque numbers are those set by the mfgr, strictly following SAE net(and other) guidelines. the conditions are very controlled, the engines are new, the dyno is a bench dyno.

basically, while you may be able to achieve consistency on your car's numbers, on your dyno of choice, the number you come up with proves nothing. it amounts to your number, for your car. then someone else comes along and posts a better HP number on this or that dyno halfway across the country, and you boggle your mind trying to figure out what went wrong...is that what you're saying?

think about it...everybody who chooses to dyno their car(stock or w/mods) coming up with varying power figures(naturally)...where does it end? what does it prove? bragging rights to run around and say your baby got 360 WHP? it's all pointless IMO.

daveak05

what are you trying to achieve by constantly having your car dyno'd? are you waiting for a magic number to jump out at you?

i simply recognize and go with the official HP number..and even those can be less than perfect(as past history indictates in some instances). but that's the only number i feel comfortable tossing out there when discussing my car(an of my cars) with others.

if i do an engine mod, then i get into a very gray area of what the actual HP may be. i would then say that it should have more power due to an engine mod that is designed to add power. but, i'm not going to foolishly start throwing out bogus HP gains based on the tuners "claimed" gains. we all know how that goes.

case in point: notice how most tuners will say (based on chip, etc) that it 'adds' 20-40 HP(example)...or in the case of the A4 turbos, 220, 230, 240...depending on what program you're running on the chip/software? those dubious ranges alone should tell us something, don't you think? they aren't measuring anything, they are estimating. and that's because they can't get a repeatable number for each identically modded car, each time....car to car, tuner to tuner, dyno to dyno, day to day, ANYWHERE.

RI A6

Wow, you do have a problem with generalization. I'd hardy say that having my car dynoed on two days would be considered "constantly". And my purpose has been stated numerous times.

Dyno 1 (day 1) - was to establish a baseline on one dyno that can be used as a point of comparison to other dyno runs, and to understand how the RS4 ECU adjusts A/F mixture and timing. I ran 9 dyno runs on the same day, learned the following: A/F mixture is adjusted in real time for all throttle settings; timing does not adapt quickly, so it is a good thing to reset the ECUs before running dyno runs to re-advance timing.

Dyno 2 (day 2) - was a before modification dyno run. Three runs that were substantially the same were run after ECU reset. In this run, I learned that "something" was causing a loss in torque/hp above 4500 rpm. Otherwise, comparison plots between Dyno 1 and Dyno 2 could be explained by atmospheric differences.

Dyno 3 (day 2) - was an after Milltek exhaust modification dyno run. 3 runs that were substantially the same were performed after resetting the ECU. In this run, I learned that below 4500 rpm, on this particular day, the exhaust makes more torque. This holds up, whether compared to Dyno 2 or Dyno 1 plots. Based on that I can safely say that the Milltek downpipes and non-res catback exhaust make about 5% to 10% more torque below 4500 rpm than stock. Above 4500 rpm I learned that torque and HP were still running low.

Subsequent reading and analysis of possible causes for the "something" that is causing a loss in torque and HP lead me to hypothesize that the intake butterfly valve is the culprit. One day, I will install a switch to control the intake butterfly and dyno again, with the valve in the open and closed position and monitor the differences.

daveak05

why can't you just accept what is "generally" known? and that is that dyno results showing specific power/torque numbers can't be validated. it's as simple as that.

it's fun to dyno your car, sure. and sometimes a dyno can isolate problems that aid in diagnosis.

but, why people have to run around and waive their dyno results as some big revelation, is beyond me. they should know better.

stick to your flap operation diagnosis via the dyno, instead of making all kinds of excuses for these widely varying dyno numbers.

some time ago, you asked for examples of dyno accuracy problems. i gave you at least 3 links(by reputable sources), that all showed the same thing. basically, that they are an unreliable method of accurately measuring power. you didn't accept that then, and it appears you still can't.

believe what you want, but the overwhelming evidence out there does not support what you're saying. and it never will.

RS4saloondrivr

That's what I'm doing, only had the car 2 months and have 5k miles already.

RI A6

I'd agree that without a good calibration standard, it is near-impossible to come up with accurate absolute horsepower numbers. However, as long as the machine and car under test is repeatable within some margin of error, relative differences can be gleened from multiple dyno runs, as long as the difference is greater than the margin of error.

This is why I will run 3 runs on one configuration, and then 3 runs on a new configuration. Lets call them runs 1 through 6. There will be some amount of measurement error in the runs with the same configurations. If we take the average of Run 4,5, & 6, and subtract them from the average of Run 1,2,3, and if that difference is significantly greater than the maximum difference between 1,2,3 or 4,5,6, then we have a measurement where we can say that we have a relative gain or loss from the configuration change. Obviously, the larger the run-to-run variation for each specific configuration, the less resolution we have in making a relative measurement.

If I were a tuner, trying to benchmark a particular modification, I'd run a considerable number of runs under controlled conditions and statistically process the measurements to characterize the error in the dyno, and the error in measurement of (operation of) a particular car. Once that is known, it's fairly easy to perform relative comparisons of engine modifications.

RI A6

I cannot speak for others.

I agree that people can game dyno tests. I've said nothing to indicate that I believe that specific power/torque numbers can be validated. I agree with you on that. I'm not arguing about that.

I do believe that with proper care and feeding relative percentage variation due to a modification can be measured and reported, if it is done right, and only within the error margin of the dyno and variability of the engine operation itself. If either of these are too large, any measurement made is worthless.