04 vs 03 Dyno test on vtec.net
Registered User
Joined: Sep 2003
Posts: 1,877
Likes: 0
From: Murrieta
Regarding your dyno, I made no intentional crack at you or your equipment, but you do happen to run the dyno which frequently puts out bigger numbers than the other(s). While it's true that in the end you are tuning for the relative gains and whether the ultimate listed horsepower is X or Y is neither here nor there, these "inflated" numbers can lead to some uninformed posts down the line by folks who don't understand the common variances in the dyno makes.
Anyways, I thank Shawn for taking the time to see this... but i would like for him to elaborate on it and swap ECUs with his personal car or something to see the reaction that the 02 has.
By the way Shawn, this is Jose.. I had that 01 GTI VR6 that came with Hoegner and lampley when they tuned their cars... I am now a 03 S2000 owner...
Registered User
Joined: Feb 2001
Posts: 2,895
Likes: 1
From: You wish
Steve,
You've stated that we've seen 10-20 hp differences in same model year cars when tested on other dynos. I disputed _that_, since none of those tests is done with the level of control that we are able to achieve on the dynapack with proper monitoring equipment - thus those results are not likely representative. My data, which includes some 50-60 S2000s, many of them stock for their first dynos, says otherwise. Now, that may not be a big enough sampling to publish a paper, but its enough for me to say that there is relatively little variation within model years, and a consistent 5 hp average difference between the early and late model cars. There is data to back that up.
In the case of the 04 model, as you have already noted, I said:
"Combined with changes in ignition timing, cam profiles, etc. the power gains are quite reasonable - although we are looking for a consumer purchased model (vs. our press car) to test as a verification. (Contact shawn@vtec.net if you're in Southern California and want a free dyno test.) "
What more can you ask for? I don't plan on buying 5-6 cars to test. As for those who choose to use this information to justify their purchases, that is their issue. I did not post the article here, I simply performed the testing and wrote the piece. I will say, however, that I have tested some 10-12 press cars from Honda over the last 2 years on various dynos including my own. Among them a Civic Si, Civic Hybrid, Acura CL-S, Acura TSX, Acura RSX, S2000, Accord (3 different engine/tranny combos), Insight, etc. And I have also tested at least one example (and usually far more than one) of each of these models as purchased by a consumer. In each and every case, the results were pretty close to what the press car put down. Sometimes the consumer cars were a little higher, sometimes they were a little lower. So, based upon past experience, Honda doesn't seem to go in for ringers.
But in the end, it was a dyno test. The results were published as is. They were surprising. And yes, my dyno reads higher, but what's relevant is the relative difference, which was noted in the article. I've also stated that the gains will be smaller on a roller dyno due to wheel weight differences. But that doesn't make a roller dyno a "real" dyno. The test team was comprised of 3 MY00 S2000 owners, so if there was any bias it would be against the 04 since none of us plans on selling.
UL
You've stated that we've seen 10-20 hp differences in same model year cars when tested on other dynos. I disputed _that_, since none of those tests is done with the level of control that we are able to achieve on the dynapack with proper monitoring equipment - thus those results are not likely representative. My data, which includes some 50-60 S2000s, many of them stock for their first dynos, says otherwise. Now, that may not be a big enough sampling to publish a paper, but its enough for me to say that there is relatively little variation within model years, and a consistent 5 hp average difference between the early and late model cars. There is data to back that up.
In the case of the 04 model, as you have already noted, I said:
"Combined with changes in ignition timing, cam profiles, etc. the power gains are quite reasonable - although we are looking for a consumer purchased model (vs. our press car) to test as a verification. (Contact shawn@vtec.net if you're in Southern California and want a free dyno test.) "
What more can you ask for? I don't plan on buying 5-6 cars to test. As for those who choose to use this information to justify their purchases, that is their issue. I did not post the article here, I simply performed the testing and wrote the piece. I will say, however, that I have tested some 10-12 press cars from Honda over the last 2 years on various dynos including my own. Among them a Civic Si, Civic Hybrid, Acura CL-S, Acura TSX, Acura RSX, S2000, Accord (3 different engine/tranny combos), Insight, etc. And I have also tested at least one example (and usually far more than one) of each of these models as purchased by a consumer. In each and every case, the results were pretty close to what the press car put down. Sometimes the consumer cars were a little higher, sometimes they were a little lower. So, based upon past experience, Honda doesn't seem to go in for ringers.
But in the end, it was a dyno test. The results were published as is. They were surprising. And yes, my dyno reads higher, but what's relevant is the relative difference, which was noted in the article. I've also stated that the gains will be smaller on a roller dyno due to wheel weight differences. But that doesn't make a roller dyno a "real" dyno. The test team was comprised of 3 MY00 S2000 owners, so if there was any bias it would be against the 04 since none of us plans on selling.
UL
Registered User
Joined: Feb 2001
Posts: 2,895
Likes: 1
From: You wish
Luis, note the transmission correction factor (TCF) on the graphs. This is something Dynapack provides so that you can estimate back to crank hp. I believe such a concept is flawed and do not use it. As long as the TCF is 1.0, you're seeing exactly what was read at the hubs, so it isn't flywheel hp. I don't like the label, but I'm stuck with it.
On the acceleration, we went into 4th gear and the 04 kept pulling. If you like, I would suggest doing a thrust comparison based upon those torque curves. I think you'll find that when the MY02 is at 80 mph in 3rd gear its turning near 8000 rpm. And if the MY04 is in 4th gear at 80 mph, its turning about 6500 rpm (you'll have to do the exact sums). The overall gearing advantage (3rd to 4th) for the MY02 is about 20%, but its making 17% less actual torque, so the thrusts will be similar. Furthermore, as the MY02 climbs to its limiter, its torque is dropping while the MY04's curve stay flat to slightly up!
In the lower gears, the differences between first and second gear are enough to offset the increased torque of the MY04, but as speeds climb and the gear ratios become more closely spaced, its game over.
UL
[QUOTE]Originally posted by Luis
On the acceleration, we went into 4th gear and the 04 kept pulling. If you like, I would suggest doing a thrust comparison based upon those torque curves. I think you'll find that when the MY02 is at 80 mph in 3rd gear its turning near 8000 rpm. And if the MY04 is in 4th gear at 80 mph, its turning about 6500 rpm (you'll have to do the exact sums). The overall gearing advantage (3rd to 4th) for the MY02 is about 20%, but its making 17% less actual torque, so the thrusts will be similar. Furthermore, as the MY02 climbs to its limiter, its torque is dropping while the MY04's curve stay flat to slightly up!
In the lower gears, the differences between first and second gear are enough to offset the increased torque of the MY04, but as speeds climb and the gear ratios become more closely spaced, its game over.
UL
[QUOTE]Originally posted by Luis
Registered User
Joined: Mar 2002
Posts: 2,133
Likes: 0
From: Northern California
Out of a pool of 50-60 pre04 S2000 dynos, you should be able to calculate the uncertainty (say 95% confidence interval) about the true mean value of your data. As far as I can recall, 50 measurements is enough to consider a z-distribution. It will definately help you check the quality of the measurements your taking. The standard deviation (or a rough estimate of +/- 5HP) isn't quite sufficient in proving that your data (or the instrument, in this case the dyno) is acceptably accurate and precise.
I don't doubt your ability as a tuner or your knowledge; it is the Dynapack that I do not have faith in. The TCF, as you said, seems like unnecessary factor that only adds to the mystery of Dynapack dynos.
Anyways, a couple quick calculations and the uncertainty will tell you alot about your dynamometer.
I don't doubt your ability as a tuner or your knowledge; it is the Dynapack that I do not have faith in. The TCF, as you said, seems like unnecessary factor that only adds to the mystery of Dynapack dynos.
Anyways, a couple quick calculations and the uncertainty will tell you alot about your dynamometer.
Banned
Joined: Jul 2003
Posts: 425
Likes: 0
From: Heaven,
As what I've always expected, the 04 will dyno more than any previous model even though Honda still rates its hp the same at the crank. No matter what type of dyno you put it on, MY04 will put out more power than any previous model. My instincts are always right. 
Registered User
Joined: Apr 2003
Posts: 1,987
Likes: 0
From: Hawaii
I look at published numbers as a reference to go by, not as the absolute performance level.
Automakers can claim any HP figure since there's no law mandating a 3rd party to check out their claims. Honda released 3 different models of S2000's all with the same peak HP claims, but as a few select tuners have learned, there are some differences.
The car everyone loves to hate is the SRT-4, and if you're not up to speed about this, the HP claims are way short of what it's actually producing. This isn't a new thing, Chevy has done this long before I was able to drive and I'm sure there's others. Some overrate their car's performances causing yet another uproar.
The goal of UL's article was to compare A versus B, it doesn't matter what the exact power value is at the wheels. His goal is to find the value of the difference between car A and B, not what the ultimate HP output is. As long as the testing parameters and environments are controlled to their best and as UL points out time and time again, monitoring coolant temps and such to make sure the testing procedures are as close as they can be, then we're being as fair as possible. Whether the difference in power is for example 15HP (255HP versus 240HP) from a dyno that produces high numbers and another dyno that also measures a 15HP difference (235HP versus 220HP), the difference is still the same.
Automakers can claim any HP figure since there's no law mandating a 3rd party to check out their claims. Honda released 3 different models of S2000's all with the same peak HP claims, but as a few select tuners have learned, there are some differences.
The car everyone loves to hate is the SRT-4, and if you're not up to speed about this, the HP claims are way short of what it's actually producing. This isn't a new thing, Chevy has done this long before I was able to drive and I'm sure there's others. Some overrate their car's performances causing yet another uproar.
The goal of UL's article was to compare A versus B, it doesn't matter what the exact power value is at the wheels. His goal is to find the value of the difference between car A and B, not what the ultimate HP output is. As long as the testing parameters and environments are controlled to their best and as UL points out time and time again, monitoring coolant temps and such to make sure the testing procedures are as close as they can be, then we're being as fair as possible. Whether the difference in power is for example 15HP (255HP versus 240HP) from a dyno that produces high numbers and another dyno that also measures a 15HP difference (235HP versus 220HP), the difference is still the same.
