If you want to see how exaggerated dynojet (or equivalently reading dyno) is, it doesn't take more than some simple injector math or compressor map calculations to prove the point. 400 wheel hp (typical aftermarket translation: 465bhp) out of 4 injectors flowing 52lbs/min at 3bar rail pressure? Do the BSFC math on that and see if that engine output estimate is reasonable for a turbo engine.

Just my 2c,
Shiv

 

After a nice blast through the canyon today, my car is DEFINITELY making serious power. Anyone who says the RMR turbo-back and intake don't help need to stop and put the crack pipe down...

The whole dyno issue is moot anyhow - regardless of which dyno I used, the net gain was 80 whp - whether my base was 220 or 190...

 

Not true. 80hp on one dyno isn't 80hp on another. Best to look at gains as %improvement over stock. An 80whp gain on a car that baselines at 220whp the same thing as an 80whp gain on a car that baselines at 190. The latter suggests a much greater improvement.

shiv

 

Shiv - the point is that the mods are making power - I am not going to try to win a dyno contest, nor do I care...

The car is making a serious jump in performance over stock which I demonstrated all day today running with stock and lightly-modded cars.

BTW you're talking 36% to 42% - not a big deal in my book.

 

You're more than welcome to test your car on our dyno if you're ever in the area. I would love to see any mildly modified EVO with stock turbo making ~40% over stock.

 

Where is this magic 400 Hp engine example comming from? I am not seeing the relevence here, not to mention your assuming fixed precision values for a number of engine variables. It would not be the first time that calculations on paper strayed far from reality.

If you believe I am mistaken about your dyno and our trying to convince me of this, then why haven't answered a single question (like the STI/EVO TQ curves on your site). I am afraid some theoretical fuel calculations for some engine we no nothing about, tested in circumstances we also know nothing about, is not going to do it for me.

Kind regards,

Eric

 

All-wheel hp, as measured on an AWD dynojet is anything but controlled. Measured power will depend on roller configuration (224 front/248 rear, 248 front/224 rear, etc.,) and center differential type and wear. It's not a good tool for accurately measuring all-whp by any stretch of the imagination. A varying portion (depending on the above conditions) of the power will go up in smoke (or, more appropriately, heat from the center diff) since the dyno has no ability to make the drivetrain work _as it is intended to_ with the front and rear wheels rolling at the same pace.

shiv

PS. The 400whp numbers comes from a turbo Miata, a turbo MR2 and a couple 2wd DSMs I've tuned back in the day on Dynojets. I would question dyno calibration (and typical wheel-to-engine correction factors) before I would draw a shadow of doubt on SAE-derived injector math and compressor airflow data from Garret, KKK, IHI, etc,. There's a reason not one OEM uses Dynojets for in-house dyno testing. Typically, they rely on dynos from Bosch, Maha, Dyno Dynamics and Superflow-- all which read withing spitting distance from each other. No surprise considering that a HP, like an inch, is a quantifiable unit of measurement. It is NOT subject to opinion or interpretation

 

That's basically what my tired, old *** has been trying to say... probably in too many words.

 

So just how inaccurate is the all-wheel Dyno Jet because of this supposed design flaw? 1%, 5%, 10%, 20%? This sounds again like assumptions based on theories that on the surface sound reasonable, but may not be taking all necessary factors into consideration. This is why only hard verifiable and easily reproducible test data can tell us if this proposed design flaw causes any significant accuracy problems (as you suggest it does). Proper laboratory testing for something like this would require significant time and access to equipment that I do not think any of us have.
This is why I look for dyno results for a given car that translate into real world acceleration, if they don't then they truly are meaningless.

It's not the math itself that I have doubts about, it's that the values you’re using may be incorrect, and not just air flow. If all you can sight as proof is one obscure test you did in the past, I am afraid it's not very convincing. Proof would have to be a test that can be verified/reproduced by other people in other parts of the country.
I thought the OEM's used engine dyno's, not chassis dyno's? Some one could jump on here and tell us that the OEM's do use Dyno Jet's, how would we know if this is true or not?
Not to mention that the Dyno Jet (and other similar) seem to agree with the OEM's, but yours does not, what does that tell us?

The only real proof would be a test that can be reproduced and verified by everyone else. This is why I am sighting test's done with the EVO/STI, they are new and should be pretty much the same through out the country. While I know it's not perfect, after a large enough testing sample of these cars, dyno/real world acceleration test's. We should be able to have a pretty good idea on what real or not, so far your numbers don't seem to add up.

Still have not seen an explanation for the apparent discrepancy within your own test data (EVO/STI TQ CURVES)?

 

Two problems with this;

IMO percentages can be even more mis-leading, mainly because as you see here at some point you have to convert those percentages back into actual units of measurement, and that's when the fun starts.

The same people who are telling us that we cannot compare from one system to another, turn around and do just that all the time. Even if they are just looking at percentage improvement's from a given mod, they argue about who's numbers are more accurate.

So this whole idea of let's all just compare within our own dyno tests and live happily is a load of crap.
There simply has to be a standard that gives an acceptable margin of error, at the moment that is the dynojet.
Shiv has come out numerous times saying things like well 180HP on my dyno is equivalent to XXX on a dynojet. When he does this he is trying to give you a point of reference to a well known standard.
They compare number's from other dyno's (even percentage of change) trying to evaluate the performance of a given car or modification. Then when discrepancies show up they turn around and say that you cannot really compare these number's because one of them is not accurate . Well then how is it that your able to apply some percentage correction value when you try to compare your dyno numbers?

I guess what it comes down to, is that without a reference, all of your dyno numbers have no meaningful value (other than tuning a given car). Even if it were true (I don't think it is), that your dyno is far more accurate than any other system out there, any and all comparisons (peak or % of change) are completely meaningless.

If you agree with this, then why are you constantly trying to support or invalidate other people’s dyno test results by comparing them with yours?
This is really what got me started with all of this.

Kind regards,

Eric


PS I understand that for marketing purposes you need to be able to compare your products performance to other's on the market. Unfortunately to have any sort of meaningfull reference, I think you need to use dyno jet plots for that.

 

There really does need to be a national standard for Dyno certification. Different Dyno's do provide different results, and hell even same MFG Dyno's are going to be somewhat off depending on calibration, logistics and environment, and inconsistencies between machines.

Developing some kind of agreed-upon baseline and testing methodology would be a start.The only way to get this under control is to develop and apply a standard method for load testing and calibration. If a group of tuners would get together and try to figure out how this could be done everyone would benefit.

This doesn't mean that any tuner has to give up the right to compete, its just an effort to develop a standard playing field, and play rules.

There are some excellent minds that could be working towards a solution, but instead this can of worms continues to provide a fertile ground for creating confusion and doubt in the minds of car owners of every kind.

Just my .02

N10S

 

This begs the question, are (repeated) dyno runs hazardous to an AWD transmission?

 

On dynos which can't control fore/aft wheelspeed, yes, the repeated dyno pulls can and will cause premature wear to the factory center differential. The task of the center differential is indeed to take care of fore/aft wheelspeed differential. But they typically only see reasonably small differentials for reasonably short durations. ~12 seconds of full load is an order of magnatude more stressful than anything one would see in the real world. Let along 10 back-to-back runs. I suspect much of the excessive jaggedness one sees on AWD dynojet graphs is protesting from the center differential.

Just my 2c,
shiv

 

The base lines are relative if they are valid. Scientifically meaningful data for supporting claims for specific gains attributable to any particular modification can then be drawn from the percentage increases present in the data extracted from the group as a whole. Individual cases contain variables that may be difficult to assess acurately and thus difficult to quantify in any meaningful way.

For a test method to produce valid measurments it must include a known standard, produce repeatable and consistant reslults, include a field that contains at least a repreentative sample of the test population as a whole. The degree of accuracy needs to be stated as do any particular conditions present during each test that might skew results.

Those of us who have had to stand an oral defense for our research are, of course, laughing at the reckless abandon that salesmen employ when promoting their wares. What is real can easily be differentiated from the purely academic. Why not just get together on the Freemont drag strip and go for it?

 

Driving casually on the street this may be true, but under race conditions (which these cars are made for) constantly accelerating/drifting through corners, I do not think you can honestly say that dyno jet puts more of a load on the center diff. I have also not seen any proof that an AWD dyno jet has damaged a single center diff, even though there have been countless runs made with them, not to mention the fact that not all AWD Dyno Jet use different size drums.

There are many 2WD dyno pulls that look every bit as jagged, so I don't think we can say for sure that this has anything to do with roller differential.

Other electro-mechanical load cell AWD dyno's (such as the Mustang system, that I have been using) that do minimize front/rear wheel speed, have generated dyno plots that are amazingly close to the AWD dyno jet. So I also do not think that you can accurately say that the center differential loading (what ever the actual amount is) is causing any significant measurement discrepancies.

boy 2c just doesn't buy what it used too, how about spending a little more and actually addressing some of the specific questions/points that I have made earlier in this thread? Unless of course you agree with all of them


This is kind of a vague statement, but if your suggesting that we can accurately compare power increases as percentages from one dyno to another, well then you obviously have not been reading this thread carefully enough. But just for you, here is a highly relevant example that will once again prove my point.

On Shiv's dyno he claims that STI's tested there are making 218 WHP (other's were even higher?) and that the average EVO makes 190 WHP, this is almost a 16% difference in outputs! With this test data, he then stated his conclusion that Mitsu must have over rated the HP rating of there engine and Subaru must have under rated there's!
Dyno testing of these two car's on Dyno Jet/Mustang show (with SAE correction) the average EVO making around 235 WHP and the STI makes 255-260 WHP that's 8-10% difference, and amazingly right in line with both manufacturer's claimed outputs! Imagine that .
So one dyno shows a 16% difference/increase in power from two similar cars, and the standard dyno shows 8-10%, why that's almost a 50% difference! Are we having fun yet? This is why my *** start's to twitch when people start talking to me about percentages.

Testing tuning/modifications on a given dyno to see if you actually get an increase is fine, but comparing numbers (raw or percentages of change) from dyno's that clearly read differently is just ridiculous. Every one seems to agree with this, but then they turn around and do it any way .

I do believe that an electro-magnetic load cell dynos are far better at simulating various loads, which is great for tuning part throttle fuel and ignition curves, because you can more accurately simulate real world load conditions. But for WOT maximum load testing (what’ll she do) there added complexity of design and operation simply create a larger margin for error. Buying one of these systems to just do WOT dyno runs IMO is a huge waste of money, they were designed to do so much more.
The thing about the Dyn jet type system is that it is so simple, other than making sure the drum mass is correct(very easy) there's just nothing to calibrate. The only possible variations that could be observed from one system to another is correction factors and gear ratios selected by the operator, or environmental conditions. I suppose the drum roller bearing could start to drag more over time, but that's easily checked and corrected.
With systems like the Dyno-Dynamics, every aspect of its operation has to be carefully calibrated, I am not even sure how these systems are initially calibrated? I am a Senior MRI Systems Field Engineer, so I have some experience with electromagnetic/electronic circuit design, maintenance and repair. Any system of this nature should be tested and recalibrated on a regular basis to ensure accuracy. Not to mention the additional parameters the operators have to set (ramp rates/load factors), its just more stuff that can go wrong. You can look at the different readings of the two US dyno dynamic systems on the EVO to see an example of this.
There is an old engineering rule of thumb that say's "The simplest method is usually the most accurate and reliable" I think this applies here.