My argument is simple. I think. Lets say my buddy and I took our stock EVO's to separate local dyno shops and my car tested at 230whp and his at 260whp. NO conclusion can be made that his car makes more whp.

If we dynoed both cars on the same dyno, then a reasonable conclusion can be made. That is my argument.

Speedlimit...

 

SILVER SURFER writes...
"Accelerating and drifting through corners" involves slip angles where the tire is rotating slightly out-of-line with the direction of the car. It does not involve significant, if any, fore/aft wheelspeed differential. The only time that would happen is during perhaps on low friction surfaces (dirt, ice, snow, etc,.) or during hard first grear launch where the front wheels would spin slightly for less than a fraction of a second. If you don't beleive me, touch your transmission case after a couple of dynojet pulls and compare it to what it feels like after even a full hotlap on a road course.

Yes I can. The jaggedness you are referring to is high frequency noise caused by Dynojet hp calculations. The jaggedness I'm referring to is low frequency which is mechanically-induced.

A Mustang dyno samples in 20-odd locations during the pull and does a "Best fit" to connect the dots. This is why graphs from Mustang dynos look unique (smooth and broad) with breakpoints evenly spaced every 250rpm. If you think Mustang dyno results "look" anything like those from AWD dynojets, please let me know what you are reffering to since that would be news to me.

Could you present them clearly so I don't have to dig through the thread again?

Ahem... STis have made between 211 and 237 wheel hp and EVOs have made anywhere from 165-202 wheel hp. Also, in what plane of physics does driveline loss double as output doubles? Contrary to popular belief, there is no linear correlation between driveline loss and hp output.

Read the my earlier comments again.

Weights of known mass are placed on each of the load cells (one for each axle) to see if they are reading correctly. The load cell is mounted in tagent from the spinning eddy current brake with an arm of known length. From this, the computer caluculates tractive effort which it then uses, with engine speed and/or gearing info, to calculate torque and then hp. It's really pretty simple. The eddy current retarders do nothing but controll the ramp-up rate as dictated by the dyno operator. Conceptually, I think this is a more honest way of measuring hp/torque than seeing at what rate it takes to spin big heavy rollers by looking at a "look-up table" (which the dynojet does). If the dynojet was truly capable of measuring (not estimating) hp and torque, it would be able to provide readings during steady-state conditions which it, as we all know, cannot do.

No it doesn't.

Cheers,
shiv

 

Shiv writes:

You mean like the type of environment these cars see in rally racing?
It's not my cup of tea, but let's also not forget the up and coming drifting/donut competitions.

So would that be one blister for the road course and two for the dyno jet pulls? Seriously though, maybe a temperature probe would be more accurate? Have you done some testing in this regard that your not telling us about?

As you well know, low frequency dips and peeks (Jaggedness) can be caused by a variety issues. I can easily show you a number of 2wd dyno plots that have just as much if not more "low frequency jaggedness". Take a look at Pruven's AWD Dyno Jet plots, they are smoother than yours, so what does that prove?
I am not arguing, nor do I have any doubt that AWD Dyno Jets that use two different size rollers put more load on the center diff. I just do not see sufficient evidence to indicate that it significantly affects accuracy or causes damage. Obviously some tuners have done an awful lot of pulls on these types of dyno's. Have any center diffs been damaged from dyno pulls that we know of?
Dyno pulls on any dyno put unusually high thermal loads on the engine and drive train. Your simulating load conditions for speeds in excess of 100 MPH, but the airflow the car gets is barely a fraction of that. I try not to do more pulls than is absolutely necessary, and then follow up with tuning on the street (deserted road).


You misunderstood my observation, I was not talking about how pretty they look compared to each other, but the actual data that each graph indicates. If you compare the actual HP/TQ values that these two very different dyno's show for any given RPM point, they are remarkably similar.



So from these numbers are you trying to say that the average STI should make 224 WHP and the average EVO 184 WHP? I do not see how this contradicts my statement, if anything it strengthens it. With 40 more WHP (or a 22% increase ) , I would think the real world acceleration should be far different from what we actually see. Especially when Dyno Jets (and the like) show the WHP from these two cars to be more like 8-10%, IMO this is far more realistic.



Sorry, I haven't got a clue which comments your referring too?


So it is calibrated using a method very similar to the Dyno Jet method, how interesting

Unless I am mistaken, the “eddy current retarders” are what provide the load on your dyno, this load coupled with RPM is what gives you your HP and TQ numbers. If you change the ramp rate and/or load through operator entry or electrical component drift, you can get a huge margin of error. Looking at your EVO dyno plots compared to Dyno Comp, proves this to be true.


Since your system is calibrated using the same method that the Dyno Jet uses for actual measurement, I would have to disagree.

I also suspect that your ramp rate/load factor used for the STI may be incorrect since the TQ curve below 3500 RPM does not show a significant difference from the EVO values measured on your dyno. Out of all of my observations, I would really like to get your feedback regarding this.


Yes it does we may just have to agree to disagree on this one. I just do not think that you can in any way accurately compare your numbers (percentage or otherwise) to the generally accepted standard (Dyno Jet).

Kind regards,


Eric

PS. I hope your not offended by my arguments, that was not my intention. I also have nothing to gain proving you right or wrong, I'm just a number crunching Geek searching for the truth.

 

I don't understand your point. Especially considering differentials are often-replaced items in completition. Like motors, turbos, wheel bearings, etc,.

No, just 10 years of dyno testing on a variety of dynos with a variety of cars.

I take it that you've never actually used a dynojet. One of the software features is the ability to smoothen out all the noise/mis-calculation artifacts generated by the Dynojet. It's on a scale of 0 to 5, five being the smoothest. 5 is the setting just about everyone uses. You give up a bit of information (which I would agrue was never really there in the first place with Dynojets) but get a smoother reading. Our has no such feature for obvious readings.

Hmm... what does a damage center diff feel like? Ever wonder why some AWD drive cars spin their front wheels longer than others during launches? Well, that's a sign of a degrading center dif. Don't expect to hear clunks and rattles as center diffs go bad. My goal in this discussion is not to convince you of anything. I don't have the time or in the inclination to to a statistical survey of center differential integrity. I'm merely presenting the facts as I see them.

Different strokes, I guess. I do as many as 30 dyno pulls in one hour when mapping a car. And it can be argued by those who have driven with me that I put cars through a lot more thermal stress on the road than I do on the dyno (which typically seens a nice 15-20 second cool down between runs).

Except for the fact that Dynojets don't have the resolution to pick up transient knock retard very well. It get obscured with all the other noise. As for looking "similar" in other respects, okay, I agree I guess.

You are entitled to your own opinion. We have considerable dyno experience with both cars and feel otherwise.

Could you explaint to me how you believe Dynojet calibrates their dyno?

As I said in my last post, the eddy current retarders provide load but do not provide any sort of reading. That's the job of the load cells which are mounted externally and in tangent to the spinning retarders. As for ramp up rate, yes, a 5 second dyno pull will give different numbers than a 10 second dyno pull. As would doing a dynojet pull in 2nd gear instead of 3rd or 4th.

[QUOTE]I also suspect that your ramp rate/load factor used for the STI may be incorrect since the TQ curve below 3500 RPM does not show a significant difference from the EVO values measured on your dyno. QUOTE]

Okay.

Best Regards,
shiv

 

Sorry I missed an important one.

Shiv writes:

I do not see where I may have implied this idea, because I agree completely with this statement of yours.

What I do have a problem with, is your suggestion that we lose around 80 HP from the flywheel to the tires. If these AWD drivetrains were really this inefficient, then I don't know how those old AWD cars with 80-90 HP (rated at the crank) 4 bangers were ever able to achieve highway speeds?

Kind regards,

Eric

 

I agree as far the highway speed statement overall. But I think the hp loss is a ratio/percentage rather than a raw number for all engines.

 

IMO it should all be measured at the flywheel . That way If one guy tunes his engine to 400 flywheel HP and another to 350 we at least know that the 400HP engine in a stock Evo will be fater than the 350. Engine dynos are much more accurate and comparible than actually trying to run the cars on rollers.

My 2 cents

Too bad it just ain't that practical to rip out the engine everytime we add a new mod and want to tune it.

 

Actually I have a fair amount of experience with 2WD Dyno Jets, and the smoothing correction was exactly my point. You suggested that the (low frequency) jaggedness was caused by the center diff loading. I am suggesting that the dyno plots you are using for reference may have just had the smoothing set low, or the boost was unstable, etc. This seems like a reasonable assumption since we have EVO AWD Dyno Jet plots that also look very smooth, I can also show you 2WD plots from other 2WD cars that look as bad (or worse) as any EVO plots I have seen.

Some of the AWD Dyno Jets have different size front/rear rollers, others do not. I agree that the units employing different size rollers do stress the center diff more, and that tunning/excessive repeated runs on such a system is probably not a good idea. But there is no concrete evidence to say that it signigicantly affects accuracy, or that doing 2-3 pulls every once and awhile for reference will cause any significant wear (compared to average use) or damage to the center diff.

I misunderstod your explination, I thought you were rotating a fixed mass to check accuracy (like a dynojet). After reading your latest reponse and hopefully getting a better understanding, I have more questions though:

How does the load cell actually measure torque?

What percentage of error is there in calibrating a device with a fixed weight, that actually measures torque rotating at high speed?

And how do you check that the Eddy current retarders are loading accurately?


I do agree that your dyno is probably better for tunning/trouble shooting, especially for part throttle applications. Looking at the manufacturers site, it seems that this is what it was designed for, more so than comparing absolute values from WOT pulls from one car to another. I just think that the absolute values could be inaccurate due to the complexity of the design and operation.

Wether your dyno is more accurate or not, I do not think you can apply any fixed correction factor to compare your HP/TQ numbers with other dyno's (raw/absolute or percentages of change).

It would be interesting to see a dyno dash type comparison with your system and a 248/248 AWD Dynojet, with a variety of cars in various states of tune.

Kind regards,

Eric

 

So you believe that for a given engine and drivetrain, the drivetrain loss would increase in a linear amount relative to engine output?
While I have not seen any one even attempt to accurately explain/calculate drivetrain loss, most seem to agree that the majority of it comes from friction. While inertial and friction losses should increase with power output, most (including me) do not think it is even close to being linear. What do you think?


Are you listening magazine boys? Now this would be a test worth reading! Put an engine on a dyno, using intake/exhaust restrictors along with maybe adjusting boost as needed, making sure to document/simulate temperatures, pressures, and airflow, as acurately as possible. On an engine like the EVO's you should be able vary HP/TQ by at least 30-40% (maybe more). Then install it and run it on a few common chassis dynos, in the same states of tune.

I promise to buy a five year subscription to which ever one that tries this and publishes the results

 

Yes and no. What I meant was, if anything, the loss of power accrued is more relative to the engine as well as all other differring factors rather than a simple 80hp loss for all awd cars. I think we agreed upon the theory just explaining it differently.

 

Deva Vu?????? Havnt we beat the dyno thing to death 8-9 times already?

Who is the unlucky sap to have 165 whp on his evo?

And also SHIV,,,,, what have you heard from Mits. on these low whp numbers, if anything?

 

While I have not seen any one even attempt to accurately explain/calculate drivetrain loss,


silver. it is very to meausre drive train friction by measuring the drag it places on a known load. The load inertia will be acted upon by the total drive train which functions as a brake. The time it takes to slow the mass will indicate with reasonable certainty, the load placed on it by the drive train. I believe this is called coast down.

 

You know, your right, in fact I remember running this very test on a Dyno Jet several years ago with my FD RX7. At the time my car was making almost 360 WHP (according to the Dyno Jet), I even seem to remember the coast down test showed an unusually low value for this car, like around 8%.
But if you consider that 8% of 360 is almost 30 HP, and that a stock FD (rated at 255 flywheel) usually dynos around 215 WHP, these number jive quite nicely.

Hey Shiv, can your dyno do a coast down measurement like this?

I will be getting on an AWD Dyno Jet and the Mustang system in the next few weeks, I will be sure to run this test. I had almost forgotten about it, thanks for bringing it up. In the meantime, if any one has done this test, or gets a chance to do it, please share your results.

Kind regards,

Eric

 

I raced my evo against my sons 03 clk55AMG... I stayed right with him... I do not care if i have 20 or 200 hp at the wheels.. this is one fast ride. Real world is better than academic , I know, I live in both worlds.

The question remains thought... what is the easiest way to make 300whp?... or in other terms, what is the "easiest" way to gain a 15%-70% in power (depending on what his assumptions are as to the stock WHP).

This thread is a very amusing pi--ing contest about who or what metods generate the most credible data. Personally I like my stop watch and a place where I can pull 60Mph to 90Mph as a means to separate fact from conjecture. It may not be all that inexpensive or prudent though given the speed limit.

Shiv... I hope you will continue your significant contributions to the import tunning arena. I look forward to putting water injection on my ride for performance, saftey, and long term economy. My conjecture is that this is the "best" way to 300WHP given current fuel limitations.

Steve, your company touts promises on water injection.. I eargerly await the fruits of your endeavors here.

Erik, in my academic mind, to question is to learn . Please post your findings.

I am off to HWY 1 to practice a few drifts while the air is cool and wet. Who knows I might see one of you there.. The Rally red number with a niner plate frame and an old dude at the helm would be me.