This thread is intended to detail the *facts* about dynos. It is however not intended to discuss such things as drivetrain loss, preferences of one dyno brand over another, what hp/tq really mean, or any individual's relative dyno numbers. Please keep posts factual.

1. A Dyno is a Tool
A dyno is a tool, nothing more. Use it as one would any other tool. Think of a dyno as a gauge inside your car. It represents some metric of the car(in this case hp/tq), at a given time. That point in time includes environmental conditions, such as altitude, humidity, barometric pressure, dew point, and temperature. As such, and two samples may vary wildly, or be spot on.

2. Not All Dynos Are Equal
Principally, there are two types of dynos: engine dynos and chassis dynos. Engine dynos measure the output of an engine outside of the vehicle. Chassis dynos "bolt up" to the physical car, and measure the output of the drivetrain. When most people refer to dynos, they are referring to chassis dynos.

In the world of chassis dynos, there are major differences. Some dynos use a roller/tire interface, while others mount directly to the hubs of the driven wheels. As a result, each type of dyno measures output differently. Additionally, some brands of dynos allow for "load factors" and/or "adjustment factors." The adjustment factors allow for one brand of dyno to approximate its readings to that of another brand. All dynos offer correction to "standard" atmospheric conditions, i.e., SAE.

3. Dynos Are Merely One Way to Measure a Car's Power
Dynos produce synthetic numbers. As in any scientific process, conducting a dyno requires the removal and/or control of as many variables as possible. As a result, some variables must be approximated to project real-world values. For example, the amount of air flowing over your car's radiator during a dyno pull is unlikely to be equivalent to the real amount of air your radiator would experience at the same speed on the open road. Logically these approximated values result in another approximation. In other words, what your car produces on the street, in the real world, will not be the exact same numbers collected from a dyno.

4. Choose a Dyno & Get a Baseline
If you intend on dynoing your car to keep track of the progression of your mods, do yourself a favor and get a baseline. If you don't have a baseline, then future dynos are pointless, other than serving as a new baseline. Once you have obtained a baseline, stick with the same dyno. If for some reason you must change to another location, your previous dynos have now been invalidated.

5. Dynos Do NOT Reflect Your Car's Worth & Do NOT Represent ***** Size
At least weekly on this forum and others I read the same thread about low dyno numbers. It usually starts out with an optimistic author that learns his car is putting down numbers less than the average as dictated by "conventional wisdom." The author typically will whine and cry, and b!tch and moan about an incredibly high drivetrain loss, or that his car is a dud, or that he now must spend more money on mods. As has been previously mentioned, one's dyno cannot be compared accurately with another person's dyno at another location. SAE corrections attempt to help solve this problem, but the corrections are themselves approximations of dyno numbers that are approximations. At some point, the original data have lost so much precision that they are now meaningless. If you don't like your dyno, get over it. If someone across the country with the same year, transmission, and mods dynos higher than you, then so be it.

6. It's NOT About the Peak Numbers
Peak numbers are for magazines and arguments with coworkers. The real numbers are your median hp/tq numbers. The greater area under the curve(i.e., the greatest amount of power applied to the ground over the longest period of time) is what matters most. So if your brother has gained 5hp from 6200 to 6400 rpms because of his shiny new intake, but lost 10lb-feet of torque from 3k to 4k rpms, then that shiny new intake didn't do jack.

7. Supplemental Information
For anybody interested in a scientific look at dynos, read this article(props to GEE_PASTA for finding the link):
http://www.dinancars.com/whitepapersFile.asp?ID=9

Also, if anyone intends on using a dynojet, download the Winpep software here. When you go have your car dyno'd, have the operator email you the dyno run files. You can use winpep to load the run files, and plot anything your brain can imagine. You can plot your runs historically, or you can even graph time versus speed. It's a fabulous tool, and it's free. It's orders of magnitude better than looking at a static printed graph.

 

Wow. Good post. I agree with everything you posted. I hope a lot of people read this and become more informed that a dyno is just a tuning tool. A G35 that makes 270whp at the Church's in probably no quicker or faster than a G35 that puts down 235whp on a typical Dynojet.

 

Nice write up!

I agree with everything but the above statement. The area under the curve is important as stated. The median is just the midpoint between low and high and is as meaningless as any other single data point. The "mean" on the other hand or "average" hp does convey more useful information.

So my question is a GTech-Pro RR just or an OBD scan tool just as good in documenting the effect of mods as a dyno? I have always believed so, but don't really know how accurate or precise one method is compared to another...

 

^ Thanks for the correction ^^
As for the gtech, I believe they are. I don't care how accurate they are, as long as they are consitent and repeatable. One might show times a full second faster than reality, but as long as it's cosistently off by the same margin, it's cool. I imagine it wouldn't be difficult to "calibrate" the times from such a meter against a real track.

 

one other correction: i think dyno's are reflective of ***** size-

the more one touts dyno results, the smaller the *****

 

246 at dynojet and 262 at churchs. so it is about 15-20 hp difference.

 

Great Points. I think it should be sticked.

.

 

my ***** > your *****.

 

thats what she said...

 

Nope - it's just faster than the other VQs that didn't dyno that high the same day.

BTW, 270 at Churches is closer to 250 at typical Dynojet (at least with my car).

And my ***** size is about avg (maybe slightly bigger )

 

I disagree with much of this. While I agree that peak numbers are pretty worthless, suggesting that median (or I assume you mean "average") numbers are the "real numbers" makes the assumption that your #1 priority is overall average acceleration over the entire rev range. While some may want that, others may actually want to make their car as fast as possible, and therefore optimize power in the high RPM range at the expense of a little acceleration in the low RPM range when driving normally.

 

Ideally you want increase the power throughout the powerband. Increases that elevate the entire powerband are far more important than just a couple quick blips of increased HP that raise peak power for a mere 200-500rpms. I think that was point Trey was making.

 

Define "powerband". Clearly nobody is going to argue that more power everywhere is better than more power just over a couple hundred RPM. But my argument is that a mod that increases power from 5000-7000 RPM at the expense of power from 2000-5000 RPM isn't necessary useless to someone who wants to maximize acceleration from gear to gear.

 

Powerband is the rpm range which is used during a race. With the VQ35, the powerband is from ~4,400-6,500rpms and fractionally higher for the RevUp motors. Elevating this area will always improve performance, but the entire powerbad needs to be elevated to make any real difference. A 10whp blip in power from 6,200-6,600rpms will do absolutely nothing because the motor spends so little time accelerating through his rpm range. People need to realize this when they're looking over dynos and see UDPs or some other crap mod that makes a minor surges in power because you calculate the average power across the powerband, it's the gains are insignificant.

 

I dunno, I wouldn't necessarily call gains over 20% of the powerband insignificant.