BillK

Just a reminder with respect to the graph above; R&T stated in their RX-8 supplement that they had to drop the clutch at 7500 RPM to get the 5.9 0-60 time they reported, so Speed Racer's graphs match that pretty well...

Speed Racer

I don't think that the Gtech will give HP/Torque numbers that will match a dyno because they are calculated rather than measured. If you have a slow run the numbers will be significantly lower. I did one run last night where I was mindful of shifting smoothly into second gear at a lower rpm, about 4k, and pulled all of the way to redline to try to get a more accurate shape of the HP and torque curves. The actual numbers are low because it took me 8.2 seconds to get to 60 MPH but the shape should be accurate. So please don't get all worked up about 156 HP. The faster run (0-60 in 6 sec) had much higher HP/Torque values (~240HP) but the graph is pretty jagged and extremely short because peak HP was reached in first gear after my 8k launch.

StretchSJE

Also, for the G-tech to calculate your torque, the known weight of the car must be 100% accurate, and you must be on completely level ground.

RobDickinson

3174llb is probably damned close.

3000ish is with 1/4 tank of fuel, as per some other threads.

Speed Racer

I agree completely with your comments. In normal day to day driving the RX-8 is not capable of doing a 6 sec 0-60. To get the car to launch off of the line you have to beat the snot out of it by dumping the clutch at 8k. When I dropped the RPMs to 6k my time also dropped to 7 sec for 0-60. So my question to the other people who have tried to measure 0-60 times would be, what RPM did you use for your launch?

I don't have access to a 350Z to use as a bench mark but if anyone else (in the NH area) wants to try their car you should send me a PM. I'll setup up the Gtech but you'll have to do the driving because I don't want to take the risk of damaging your car.

As for accuracy of my new Gtech Pro Comp I can only tell you what the manufacturer claims.

"There are several significant improvements over the original G-TECH/Pro Performance Meter that have been implemented in the new COMPETITION model. First of all, there are 3 accelerometers and they are fully temperature compensated. This in itself is a huge improvement in accuracy. Secondly there is a very sophisticated new calibration algorithm that allows much higher precision. Also the system now has 32 times higher resolution which speaks for itself. New Noise-Correction algorithms have improved overall accuracy as well. With all of these valuable features we are very proud to announce that the accuracy is now within 5/100 second. With the consistency at 5/1000 of a second! Absolute Horsepower and Torque measurements are within 3% and consistency within .5%!"

Speed Racer

I have a 6spd GT and I used the following figures to come up with the weight.

3029 Mazda's offically listed weight
+145 My weight
+25 Misc junk in car
-25 Tank was about 3/4 full

3174 Estimated total weight

I'll have to put the car on a scale at some point to validate those numbers.

Shamus

Performing a google search for something like "G-Tech how.accurate" - yeilds plenty of hits from guys who have taken their cars to the tracks to compare the G-tech w/reality. The G-techs are fairly consistent, usually under 1/10th of a second off in the quarter. Sports Compact Car also did a test of this device (long-time ago) and were very impressed with the accuracy. They aren't usually off by 8%.

Dynos might remove the driver from the equation but they add several more assumptions to the mix. They certainly don't tell us squat about the real hp at the crank, without knowing what the normal parasitic losses are on a brand new vehicle, whether it was cooled properly, if the DSC was engaged or not, wheel-slip, etc.... Each dyno is different as well.

All I am saying is that one must make several assumptions in order to conclude that the car is down on power to extent people have been postulating. But these speed runs remove many of the factors that dyno runs bring into the data, either the car is performing or it's not. The dyno's do seem to point to a loss at the higher rpms relative to the rest of the run, but that doesn't change the fact that someone is damn-near nailing the test times with his US Spec RX-8.

This is good stuff.

RobDickinson

So numerous Dyno's where the DSC (off or not, who knows), rich fuel etc

vs

G-Tech with a car running on real roads as its meant to be driven.

One way is for Speed Racer to dyno his car too.

From what I've seen there seems to be a mix of results, some claiming good high rpm power and better gas milage/no sooty exauhsts, others not.

Speed Racer - hows your gas milage, and exauhst cleanliness?

Speed Racer

My gas mileage has ranged from 13 to 24 but out of 3,500 miles the norm seems to be 16-18 with mixed city/hwy driving. Oil consumption is very high at about a quart for every 1,000 miles. Plus the exhaust is covered with soot.

If anyone wants to set me up with some dyno time (in NH, VT, or MA) send me a PM.

takahashi j

iTrader: (2)

btw...on what date did u get ur RX just for the record...?

Elak

Let me point out that it is not "a single Gtech vs multiple dynos" to indicate that the car has all ponies (at speed). The top speeds achieved by P00Man and others are not possible with only 175-190whp (assuming Mazda isn't lying about the 0.31Cd).

It seems quite likely that the dyno runs are not feeding enough air into the ducts. However as pointed out by others, the SAE methods for crank power measurements does not allow external air induction (turbo and compressor are OK). With this explanation the dynos and test runs are consistent, and the only real answer will come from an engine dyno.

/Elak

Speed Racer

I picked it up on 7/16 from Wellesley Mazda, S/N 1102, rework 03F02 & 03F04 on 7/8

pelucidor

This may not be a valid statement. The Lexus IS300 with 215hp at the engine and 185rwhp easily gets to 144mph (speed limiter, achieved by many people). A much lighter RX-8 with similar Cd should easily get to 145mph with similar (i.e. low) rwhp.

Elak

Could you post the TQ/HP curves for the fast run, even though the graphs are a bit jagged?

Thanks,

/Elak

Elak

Sorry but weight has nothing to do with it, - the front area and Cd values do. Also it may be quite possible that the IS300 also benefits from Ram-Air effects, and thus produces more than 215hp at top speed.

The air drag effect on a vehicle is physics, not religion. Unless the RX-8's Cd is lower than 0.31, it requires ~250whp to reach 148mph.

/Elak

Speed Racer

I don't have access to the graphs right now but I'll post them tonight after I head home.

P00Man

there have been numerous (well, maybe three or a few more?) reports of people achieving the test numbers seen in CD and RT, most recently Speed Racers results. as i said before, some guys freind got a 14.5 in the 1/4 and has the time slip to prove it (ill look through the s2k vs rx-8 thread to find the guy and send him a PM asking to post his freinds time slip) my run to 145mph (and there was power available for more, or at least there seemed to be, either way, 145 is damn close to 148)

these numbers are not possible to achieve with a car making less than 250hp, or at least not with a car similar in dimensions and weight to the rx-8

however, the only way to get numbers like these is to drive "properly" with quick semi-abusive shifting and a damn heavy foot, if more people go out and start beating the **** out of there cars (after proper break-in, of course!) then we will definately see more numbers like this, as long as the person driving knows how to drive (ie, doesnt take 3 seconds to shift, mashes the gas, etc.) a good way to judge, i find, is that if the ride is incredibly uncomfortable, youre getting close

pelucidor

Elak - I probably know a little more physics (my hobby) than most and I am also a Mechanical Engineering who specialized in Applied Maths. I know plenty about drag force (see my posts in this thread).

I will also laugh at your statement 'it requires ~250whp to reach 148mph' unless the car has a Cd under 0.31. Hahaha. Honda S2000 (stock): several people have got over 161mph on the speedo (true GPS verified speed over 150mph). Cd = 0.38 soft top up, Cd=0.42 top down. About 200whp. Hahaha.

See these threads:
http://www.s2ki.com/forums/showthrea...threadid=60091
http://www.s2ki.com/forums/showthrea...threadid=40244

P00Man

ok, found the guy (not the s2k thread, the zoom zoom not thread where the 350z guy cries and whines for a while) anyway, i pmed him asking if he could get the timeslip and scan it onto here, hopefully hell be able to get a hold of it

Elak

It is never fun to be wrong and this time I am. Somewhere I lost the 0.5 factor from the drag formula, which means that the drag for the RX-8 is about 125hp at 248mph. Not the 250hp I've been arguing.

My appologies for adding to the confusion.

/Elak

pelucidor

Yep - 1/2*rho*Cd*A*v^2=Drag Force.

No problems.

ChurchAutoTest

Two things,

The G-tech results posted support the dyno data shown to date. While the hp/torque levels don't look accurate in the absolute sense, the shape of the curves is relevant.

Note that between the torque peak and 8750 rpm, the g-tech data shows a loss in torque of 19% (from peak to 8750).

I just pulled the data on the RX8 I dyno'd and it showed a loss of 21.5%. Considering variations in car, testing procedures, etc. and the fact that according to existing sample data the RX8 I tested was down about 12-13% in peak hp from where it should have been - I say the accelerometer results tend to support the dyno results to date (from the various dynos used) - even down to the dips and bumps in the torque curve above 6000 rpm.

Now, for the comments about dyno accuracy/relevance/etc....

1. The accelerometer tells you less about crank hp than the dyno does. Think about it.

2. There don't have to be any assumptions about DSC, temperature, etc. on the dyno - at least no more than there have to be in a street test.

The only area where a dyno is lacking is in the simulation of ram air effects on a car in motion. Everywhere else it is superior for measuring changes in power and performance simply because it _eliminates_ more variables than you can on the street. The more variables you can eliminate, the more sure you can be about the performance of the engine - which is what dynos for. This is also why engine dynos are superior to chassis dynos for measuring engine performance - they eliminate the chassis variable.

SC

RotorMotor

Ugh. You guys make me sick.... :D

Keeper

A 3g Eclipse tops out at 145mph at about 5500rpm in 5th gear. It has 50hp less than the 8.

My car is an automatic, sends about 160hp to the wheels, and I've personally had it up to 138 before I backed off (the speed was still climbing, but I could see a car in the distance and the floaty feeling developing in the rear was making me nervous).

So I'd say that a car making 175-190whp is quite capable of hitting 145mph.

Disclaimer: I don't recommend anyone else try and duplicate my stupidity. I also don't recommend that someone jab on the brakes at 120+ either like I've noticed a few people mention -- if the brakes don't pull evenly you are going to have a very bad day ...

Speed Racer

The software with the Gtech automatically locks into the area where it thinks that peak HP was made. In the case of my fastest run I supposedly made the most power during 1st gear. The graph is very short due to the high RPM launch. I'll also post another graph of the HP during the full run. I don't have a lot of faith in these numbers because:

1. A Gtech is not a dyno
2. Lots of wheel spin in 1st gear
3. Rough power shifts may confuse the Gtech's accelerometers