Dyno in Greensboro, NC area? I'm willing to dyno a Steve W. chip...

[KFC911]

Although I haven't chimed in on the most recent round of chip bashing by Loren (and others), I'm willing to attempt to put this issue to rest. Although I purchased my chip before Steve's web page was available (based upon the experiences of folks on this board), I can absolutely see why someone might be skeptical of any "claims" without proof. However, I (at least in my 'mind'), can definitely "feel" a difference with Steve's chip, and there's no doubt that my car feels more responsive, particulary in the lower rpm ranges. However, now I'm willing to 'quantify' this with a few dyno runs if that is feasible. Does anyone know of a qualified shop in the Greensboro, NC area (Speedwerks perhaps?), that does dyno runs for a reasonable rate? I seem to remember a recent 'group dyno' going on in Charlotte...has that already happened? I'm willing to spend a few $ to see which of these knowledgable guys is correct on this issue (and make no mistake...both Loren and Steve Wong know more than I EVER will about these issues). I'm not sure WHY I'm wanting to do this (particularly since the data posted by Mike Wong's G-tech runs have already quantified an average HP increase of 13.5 with Steve's chip), but now the analytical side of me is taking over, and I want to see the actual dyno numbers for MY car with MY mods. My car has a sport muffler, euro premuffler, and I always burn 93 octane...now I'd like to let the dyno numbers speak for themselves between the stock chip and Steve's customized chip. Can someone in my area point me to a shop?

ps. Loren...would you like to make a small friendly wager as to what the results are? Of course I would never bet on anything that I 'thought' I was going to lose...but make no mistake...I have been wrong "at least" once before !

[1980SC]

I've used "TurboTune" in Greensboro. They've got a Dynojet dyno. I think it was $60 for three runs the last time I was there.

I've also used "Carolina Automasters" in Durham. They also have a Dynojet, and the price was $65 for a half hour.

-Rob
1980sc

[racemor]

This info comes from Central Carolinas Region SCCA's web site. Please excuse the all caps, I'm just pasting in.:

TAYLORS SPORT-N-IMPORT SERVICE COMPANY IS OFFERING DISCOUNTS TO SCCA MEMBERS FOR TIME ON OUR NEW ALL WHEEL DRIVE CHASSIS DYNO.

FROM MORE INFORMATION PLEASE CONTACT KEITH TAYLOR OR JASON SMITH, TAYLORS SPORT-N-IMPORT SERVICE CO
724 OAKRIDGE FARM RD HWY 150E
MOORESVILLE, NC 28115
(704) 664-4099

[JeremyD]

I'll send you a twenty to help pay for it just to shut everybody up too.

and I have lots of feelings

[Halm]

Ok, Ok... I am several weeks late in going over there and checking them out. Maybe tomorrow but I promise I'll get there by the end of next week.

Keith, Tom Morgan, Paul Thomas, Ken Charnock and I plan a Saturday with them or someplace in the near future. You are more than welcome to join us.

[Embs]

If you have a third party (not a buddy) do the runs and meet the following criteria, I'll bet you $50.00 that you DO not see more than a 5% increase in RWHP at 5000 RPM.

• 3 runs stock chip no other changes
• 3 runs AM chip no other changes
• every other run will require a chip change
• Final score will be the average of each of the 3 runs respectively
• honestly report the data

[KFC911]

Thanks to all who've responded! Hal, That would be perfect...I'd love to join in with you guys when the time comes...just let me know! Plus, I'll get a chance to drool over some of the numbers you guys are going to be posting (...now what's a hot rod twin plugged 3.5 MFI gonna do?). Todd, I was really just kidding w/ Loren when I made my offer to bet...
However, you're now changing your tune a bit aren't you if you're conceeding 5% at 5K (...what about the 'conservative' increase in the rev limit)? What I mean is this...I've NEVER seen anyone promote Steve's chips as being an earthshattering boost...quite the contrary! I think all of us 'you CAN "FEEL" difference in Steve's chips' crowd have been dismissed as spouting bs by Loren et al while all we've been saying is "try it...you'll like it"...I have NO DOUBT that the dyno runs will quantify that there IS a performance increase to gained by a chip upgrade, but I'll let the numbers speak for themselves. Just to let you know, I'll probably only do one dyno run each with the stock chip, and two different versions of Steve's chips (one is mapped for 92 octane, the other for 93 with a few other tweaks). I know that's not statistically valid (neither is a sample of three runs), but I'm just telling you up front what I'll be doing. Mike Wong's G-tech data that has already been posted was comprised of 7 or 8 runs each (still not enough for a valid statistical sample) but more of what you're looking for. I will without a doubt be posting the numbers here for all to see when the time comes. Let the games begin!

ps. I do not 'think' that any of Steve's tweaks are causing any pre-detonation issues, etc. (particularly w/ 93 octane), but since I don't know how to quantify that...I'm willing to 'roll the dice' with Steve's expertise in that matter. If I am (in fact) shortening my engine life, then it's my $ and I've simply followed the Junior Johnson school of thought and "ran it 'til I blew the MF up"!

[Embs]

Originally posted by KC911
However, you're now changing your tune a bit aren't you if you're conceeding 5% at 5K (...what about the 'conservative' increase in the rev limit)? What I mean is this...I've NEVER seen anyone promote Steve's chips as being an earthshattering boost...quite the contrary!

No tune changing, 5% is less than the minimum of 6% that 911chips.com claims "throughout the powerband", I would say that 5000 rpm is within the powerband.

From the 911chips website:
"You can expect a 6-8% increase in horsepower and torque throughout the powerband on to redline, with a 10-20% increase in part throttle torque, horsepower and engine efficiency."

Once again, I have nothing for or against any of the people selling things, incuding chips. But when someone is making claims of increased HP "throughout the powerband", then LET"S SEE IT!!

What exactly is part throttle engine efficiency?? Because if I can gain 10-20% efficiency (without damaging my engine) I want a chip right now.

[JeremyD]

So all that pschyo babble bs was over 1% If it hits 6% Steve's not a snake oil salesman but at 5% he's a friggin liar? Give me a break! At 8% I guess he's a genius. And why a 3rd party? - I can see you ask this of Steve Wong - conflict of interest if Steve does his own dyno runs but come on.

KC911 - I'm with you on this - but I'm afraid - even if you did it in black and white - hell even if you videotaped it and had Alan Greenspan sign off that it was real you'd have naysayers. I can hear it now - not enough runs to be statistically relevant, airt temp was too low - everybody knows engines make more hp in lower temps, better gas available in NC, tires were not warmed up properly, sun was in my eyes, that sort of thing. Not sure it's worth the effort.

[KFC911]

Jeremy, At this point I'll be doing this to satisfy my own curiousity more than anything else (plus it'll be a blast to see some of the other NC guys cars on the dyno!). Many months ago (on Rennlist), Loren & I discussed this, and I pointed out to him then that in NC we get better gas than is available in CA. Loren's (and IMO irrational) argument at that time was that if you didn't quantify the chip(s) with dyno runs, then any gains were 'percieved gains' (i.e. not real because they haven't been proven). Well, I don't NEED a thermometer to tell me a fire's hot, and I don't NEED dyno numbers to 'FEEL' the difference in Steve's chips, but now I AM curious ! I'll do the runs (it'll be fun), and we'll let the numbers speak for themselves.

[Jim Smolka]

Keith,

Sounds like a fun project, and thanks for effort put forth.

Please fwd me the raw data. To accurately analyze the data requires more than just averaging, more than standard deviation too.

IMHO, since this a 1 change experiment, a random drawing from a hat should be used (such as a blue marker for SW Chip and Green for Stock chip) to determine run order (i.e. blue blue green blue green green green blue green blue). Time btwn runs should be a constant also. Naturally, each run should be of the same length. A min of 5 runs of each color, and more the better (up to a max of 30 runs each). Any unusual items should be noted.

There is a bunch of great boring statistics that must be used to analyze the data (OK, I will admit it, I was awake in grad school)

[JeremyD]

Sorry for the rant - Jim - it would be great if you can be involved - and KC911 - I applaud you for doing this. I have the same set up and would be curious to see your results too - I also think there is more than just feeling the results.

[danmatar]

I've been to Turbo Tune a number of times, although not with the Porsche. Nice guy(can't remember his name). He seems to do alot with the Grand National/GNX crowd. $60.00 for a session seems about right if memory serves. They are right around the corner from the coliseum.

[Lorenfb]

Horsepower Comparison of Performance Mods

There are errors which are inherent in any comparison testing:

1. the error of the measurement device, e.g. G-Tech, a dyno,
2. the statistical error of the two data samples, i.e. the variance
of the data samples
3. and the methodology of the actual test, e.g. G-Tech - having someone driving on the street with a device mounted on the dash
or windshield versus a car fixed to a dyno.

The G-Tech probably uses some form of an accelerometer with various
other inputs and constants to determine HP. It really doesn't matter
about the actual functioning. The key is its measurement error and
repeatability of measurements, e.g. Is there a recovery time or bias
which occurs between measurements? I doubt that the G-Tech has an
measurement error less than 5%. Who knows what the repeatability is?
The measurement errors can possibly be cancelled in a comparison test,
but not when there's a repaeatability problem. When using a dyno,
it will most likely have much less error problems than the G-Tech.

The statistical error will always manifest itself anytime data is
gathered from two different data samples. Thus, one can not just
take the average of each sample and compare the averages to determine the better of two performance tests. The key is whether
there exists a significant difference in the two data samples to conclude
that one mod is better than the other. As an example:

Run #1 160, 170, 168, 178, 175 average = 170 HP

Run #2 185, 165, 180, 170, 190 average = 178 HP

Although Run #2 has a higher average of 8 HP, when one compares
the two data samples based on the variance of the data points,
one might find that there's really no significant difference
between the two runs. This is only an example and I didn't
do a statistical comparison test which must be done.

Third item mentioned is the most problematic especially for the
G-Tech device since you have a driver driving on a street trying
to always maximum each run while shifting and accelerating.
You have this human element which comes into play which you
don't on the dyno, i.e. "pedal to the floor" in one gear & no
other external or psychological factors.

Therefore, when all is considered, the G-Tech probably has less
reliable results when comparing mods than does using a dyno.
In either case, though, one must consider all the errors before
making any conclusions.

Good Luck
Loren
'88 3.2

[Lorenfb]

Sorry for the previous bad format.

Horsepower Comparison of Performance Mods

There are errors which are inherent in any comparison testing:

1. the error of the measurement device, e.g. G-Tech, a dyno,
2. the statistical error of the two data samples, i.e. the variance
of the data samples
3. and the methodology of the actual test, e.g. G-Tech - having
someone driving on the street with a device mounted on the
dash or windshield versus a car fixed to a dyno.

The G-Tech probably uses some form of an accelerometer with various
other inputs and constants to determine HP. It really doesn't matter
about the actual functioning. The key is its measurement error and
repeatability of measurements, e.g. Is there a recovery time or bias
which occurs between measurements? I doubt that the G-Tech has an
measurement error less than 5%. Who knows what the repeatability is?
The measurement errors can possibly be cancelled in a comparison test,
but not when there's a repaeatability problem. When using a dyno,
it will most likely have much less error problems than the G-Tech.

The statistical error will always manifest itself anytime data is
gathered from two different data samples. Thus, one can not just
take the average of each sample and compare the averages to
determine the better of two performance tests. The key is whether
there exists a significant difference in the two data samples to
conclude that one mod is better than the other. As an example:

Run #1 160, 170, 168, 178, 175 average = 170 HP

Run #2 185, 165, 180, 170, 190 average = 178 HP

Although Run #2 has a higher average of 8 HP, when one compares
the two data samples based on the variance of the data points,
one might find that there's really no significant difference
between the two runs. This is only an example and I didn't
do a statistical comparison test which must be done.

Third item mentioned is the most problematic especially for the
G-Tech device since you have a driver driving on a street trying
to always maximum each run while shifting and accelerating.
You have this human element which comes into play which you
don't on the dyno, i.e. "pedal to the floor" in one gear & no
other external or psychological factors.

Therefore, when all is considered, the G-Tech probably has less
reliable results when comparing mods than does using a dyno.
In either case, though, one must consider all the errors before
making any conclusions.

Good Luck
Loren
'88 3.2

[KFC911]

Thanks Loren, Jim & everyone for their response! Loren, I understand exactly where you're coming from on this. Way back when... I also sat through (or skipped !) many stats classes, etc. in one of my majors (Quantitative Analysis), and math has always been a very strong subject for me. No doubt that it would be nice to have 30 dyno runs (technically, 30 would be the minimum to be 'statistically valid' as I recall), alternating chips at random, etc. as Jim has suggested, but that's just not feasible (at least for me). Since I'll be going to the dyno with the other NC guys, I don't know how long each car will spend on the dyno, number of runs, time delay between runs, etc. that will factor into the equation. That's why I've already stated that nothing I'm going to do will come close to providing any kind of "valid" statistical data. I'll simply plan to do some runs (as many as is feasible) with the stock vs. SW chip (if it's feasible to randomly select the run order and swap chips in between, I will). It 'might' be just one run with each chip variation (1 stock vs 2 different SW chips), but I'll certainly try to make it "more" statistically valid than that. Jim, if you want to come join us (...got anything you want to dyno?), you're more than welcome to assist, otherwise, I'll be glad to provide the raw data for you (and everyone else) to analyze for themselves and draw their own conclusions. I'm looking forward to hearing the debate on these conclusions ! Let the fun begin...

[Lorenfb]

The key point I was trying to make was the HP gain (small 6 to 8%)
versus the possible errors, e.g. G-Tech >= 5%, based on methodology.
I agree that a large number of dyno runs (> 3 to 5) would be costly
and a waste. All that is needed is a good planned dyno procedure
and runs (minmum of 3 each).

Good Luck
Loren
'88 3.2

[kjb]

Loren,

I think you can assume that the Dynojet measurement error is independent of the chip installed in the DME.

How may runs you have to make with each chip depends on this measurement error and the smallest difference in horsepower that you'd want to be able to call significant.

In other words, a G-Tech should be fine for doing this as long as the number of runs with each chip is sufficient.

/ J

[1987C1]

From the G-Tech Web Page:

Absolute Horsepower and Torque measurements are within 3% and consistency within .5%!

I see no independent tests listed on their web page, just this statment.

[Lorenfb]

Thanks 1987C1 for your good input. At least we have some idea about
accuracy even though no real confirmation of sure.

Have Fun
Loren
'88 3.2