I would to. But be warned across the 1/4, tenths of a second is a rather long time. And of course we're assuming the driver runs the 1/4 perfectly (or at least badly in the same way) each and every time.

Also keep in mind, the a 1/4 run doesn't have a lot of software/maths in it trying to turn a measured number into something that's comparable - the 1/4 uses a fixed method of making comparable numbers - TIME (no fudging/correction factors required to make time slips comparable).

LOL a 10th of a second is not 17 % error

so if u understand where im coming from

now stop throwing in other paramerters in the bathe room scales if there calabrated by the people that manufacture them and no one has played around with there calibration what would i expect to see if i weighed 110 kgs on one and then jump on another

u seem to miss the piont read the whole tread and im sure u will comprehend where the debate lies

cya

Details, Details, F**k Details.

Why make everything so hard???

What we have here esentially equates to a log with a sensor on the end, sending signals to a computer to calculate a power output as a number.

The software has internal corrections for temperature and atmospheric pressure. And once calculated this gives a power output. It even says so on their website.

All the operator needs to do is put his foot to the floor and hold it until the car redlines or the graph no longer shows its making power, but falling off. Its not that hard a principal. We all do the same thing on the street driving the car every day. Anyone can do it. (Not criticising any dyno operators here)

We are talking the same brand dynos here, so take that out of the equation. Im sure Dyno Dynamics had a service plan to ensure their dynos are calibrated to the highest precision. That question will have to be asked of the owners. So calibration shouldn't be an issue.

What other variables are there?

One version of software making a different calculations to previous version --> Question Dyno Dynamics why their philosophy has changed

Operator not setting everything up in their dyno --> Question Operator

Operator playing with variable nudge factors --> Question Operator.

Am I missing anything? Do you design dynos or their software for a living, so that you can talk in such an authority on them and they way they calculate the output?

And no, I'm not having a go at you, just wondering why a simple question has to be made so hard sometimes.

LOL


same can be asked about u " do u make dynos " and for that matter have u ever run a dyno "" ???

na i dont thinks it complicated but seems ur adding in all the complications lol mmmmmmmmmmmmmmmmmmm just quoting here " Operator not setting everything up in their dyno --> Question Operator

Operator playing with variable nudge factors --> Question Operator.

Am I missing anything? Do you design dynos or their software for a living, so that you can talk in such an authority on them and they way they calculate the output?

And no, I'm not having a go at you, just wondering why a simple question has to be made so hard sometimes.
"

LOLOLOLOLOLOLOL

mmmmmmm why dont u have a read im not against u or anyone else for that matter

its a simple debate

amazing LOL PMSFL

maybe we drop the debate seems ur the physics matts teachere here why dont u explain

I;m sure you work for a living

im sure u go to the grocery store to buy goods and yet u tell me scales a not a valid example

yet a dyno relies on a load cell and speed to make a calculations >

electronic scales use what to make a weight calculation mmmmmmmmm i must be missing something here

calibration weights must weigh differently in hallam or keilor dows sees like it

and again i repeat myself read what the arguement is before u sprout ur mouth off gessssssssssssssss maybe i call a friend LOL

cya

like u quoted " im not having a go at u " but yet u seem to be a authority on Dynos

have fun cya

ps
\ makes interesting reading " 10kgs weighs more in hallam by 17 % LOL"

It's not a simple question. It SHOULD be but isn't. I'm not a dyno maker/programmer, but do have a solid grounding in Physics.

You're log with sensor thing is great - now if only there wasn't all that software in between the sensor value and the printed value, then you'd have something workable.

The 'problems' come in when the dyno software tries to fiddle with the figure so it should then be 'compareable'. The problem lies in the fact that it now has a whole heap of variables and approximations to deal with. Variables used, ambient include air temp, barometric pressure (which of course assumes those sensors are calibrated properly). Then there's the approximations - tyre/log grip, wear factor on the log sensor, what difference does a HOT dyno make, remember it's gonna have X kilowatts of heat to get rid of from the load.

BTW: love the log thing, rollers is so boring. log is the way to go.

If we could get an 'uncorrected' power figure ie. stright from what the sensor says.

There are dyno calibraion and correction factor calculation standards, but the problem is there's more than one standard, which manufacturer uses which standard, if dyno X and dyno Y use different correction factor calculation stnadards, then their results won't be comparable even if they're calibrated the same.

Thanks... I thought I would take the opportunity to oversimplify things...

I cant see how heat on the dyno would be an enemy of the results. I understand what your saying, but we can assume 2 cold dynos. I would have thought, heat of all things would decrease the power output reading.

If various manufacturers are using different correction factor standards, then we have to ask the question of them "Are we both on the same planet. Why are is it necessary to use different standards?"

Considering we are using the same manufacturer, and potentially getting different results for the same input, then we have to ask "where and why do the calculations differ. Do you think you were originally wrong, or are you making up for something?"

I see where everyone is coming from here. And I think everyone is right!
I think, and FPV_GTp, you're not gonna like this
Chassis dynos are inherently going to have some inaccuracies.
It's almost impossible to avoid.
SO many variables can modify the end reading, that to control them all perfectly is nigh on impossible.
Maybe, instead of just a unified software mode (Dyno Dynamics, shootout), a unified set of operating procedures needs to be adhered to as well. i.e, allow dyno 'cool down' period(obviously this won't be possible in all circumstances), etc. etc.
It would go a small way to providing more repeatable results.
I call this method the 'Steady' method...

hi

no i have no problem with ur comment nor anyone else for that matter

we are trying to set a controlled test

its simple
same dynos ( brand wise , same software ) ( same car ) ( same operator )

we can simulate the heat factor in the dyno and the car that will be tested

steady state testing or graph mode im easy which ever people prefer

datamine has some interesting ideas and i dont disagree totally with his ideas


forum is open for discussion but some times yes we all think we know it all

im only stating my opinion and from my own experiences with drag cars and street cars and chassis and engine dynos what i have encounted in my work place

cya

Last edited by FPV_GTp on Wed Aug 24, 2005 2:19 pm, edited 1 time in total.

Of course, we're all forgetting thiese things are computer controlled, and any one whos been around computers long enough, will know they've got a mind of their own sometimes. :p

I really don't have any more to add, I'm at the limits of my scientific knowledge. But by all means, this is sure an interesting and provocative topic.

It would be really interesting to see a controlled test...
One car, one dyno, with controlled circumstances. Like seeing what difference dyno heat, malfunctioning sensors and such can make.
Would give us all an idea about whether these are the major contributing factor to Chassis Dyno inaccuracy.
This IS a really thought provoking thread. I sat here for like an hour re-reading peoples posts and really thinking about it all.

Yeah it is a good value thread !!

So many experts,not enough dynos !

I think Arm 79 is currently doing a lot of research atm, and He is learning the REAL workings of a chassis dyno.
He will be a good person to talk to in 6 or so months once he has finished his research.

Apologies for buying into the thread late in the arguement, but just my opinion:

Remember what you are actually arguing. If you are trying to get a reading between two different cars, to see which produces a higher RWKW figure in a given dyno mode, it doesn't matter if some of the uncontrollable variables are not correct (air pressure, temp, 'nudge' settings, etc) as long as they are the SAME for both runs.

If you are trying to compare results for the SAME car, between different dyno's, then the outside variables become VERY important. As mentioned above, the software used takes into account these outside variables to calculate a produced figure. Even with the exact same input figure from the rollers, simple logic will tell you that if the mutating factors are different, a different end figure will be produced. That is, with ALL other factors being controlled (car, throttle application, dyno software) a different end figure will be produced if the outside uncontrollable figures are different. Even with the standard SI 1kg weight method, if outside variables differ, and these are taken into account when producing the end figure, the displayed result will differ between dynos, even though you are using the same scientifically calulated weight (or car) and dyno software. Out side figures may just seem like details, but they are details that will affect the displayed result (perhaps the 'real' result to a much lesser extent), and for that fact they cannot be ignored.

The state that chassis dyno technology is at now, results are only ever useful for comparing with other results obtained where all other figures besides the car in question are EXACTLY the same. Same temperature, same throttle application, same dyno software and settings, same state of car (warm or cold), same air pressure, same dyno temperature, same wear on dyno and car. Any of these figures that differ, will change the outcome, and reduce the accuracy of the figure produced. Aruging that becuase your car made xxx kilowatts on Jims dyno, and is therefore better than your mate's car that made yyy kilowatts on Sams dyno is useless, unless ALL other figures than the car are exactly the same. In the same way that saying because your car made x kilowatts more today than last week on the same dyno (but with different temperature, different state of car and dyno, and different dyno software) is rubbish.

In summary, use dyno figures for whatever you want, but never forget that unless you exactly replicated the circumstances of which you are comparing them too, with the car being the only differing variable, then you should take them with a grain of salt as, the rate at which the car can turn the rollers is not the only input factor to the result produced. It is amazing that by trying to make dynos more accurate by adding in compensation variables, we have actually achieved the exact opposite

Tom.... Nice of you to join in...

Have a read of this: {DESCRIPTION} and read the bit about being able to accurately reproduce figures, regardless of the variables.

I had a little chat to Dyno Dynamics this afternoon. They tell me that their Shootout software takes into account many ambient, atmospheric and operator variables and will accurately reproduce the same figures regardless. Their words were (not verbatim I'm afraid) "You do not need to reproduce the exact conditions to get the same results, our software will account for this"... Now, how does that kill every little opinion so far in this thread?

Now, everyone, don't go jumping up and down complaining and providing opinions. This is from the "horses mouth", not an assumption or opinion. It is a definate response from the manufacturer.

Plus, we are not looking for a little pissy 1 or 2% or 2 or 3kw difference here. These are explainable. We are looking to explain excessive differences... 10% or 15% or 17% or 20%...

Nice!!!

But Adrian your still forgetting the variables!!!!!!!

LOL