What makes a good Dyno test??

[Geronimo '74]

Just wondering.
Since I'm gonna try having my car dynoed before and after the installation of a Steve Wong chip, what makes a good dyno test?

What are the do's and don'ts of an accurate test?
Things you cannot do without?
Should I prep the car?
Tell me, I'm a dyno virgin.

[Wil Ferch]

There has been a lot already discussed on this topic..do a search.

As a minimum, I would suggest:
1.) same machine...not just same "model" of machine...before vs. after.
2.) do each run under "corrected SAE conditions". This will flush out differences of air temp , relative humidity, local air pressure and density etc..in the comparison runs, and it will "normalize" the actuals to the same set of SAE conditions. Usually when this is done you will see a "CF" factor at the very bottom of the dyno run chart ( "correction factor"). I was able to compare high altitude dyno runs on a particular machine with someone elses' identical car at sea level...and the graphs almost superimposed over itself, since both were SAE corrected...impressive.

Wil

[randywebb]

& same operator

Road dyno will never be as repeateable as engine dyno

Can you afford 3 runs before and 3 after??

[kucharskimb]

it doesn't really matter, because when you post your results on this board, everyone will say that they were flawed and that you can't possibly make that much Hp...

2.2 E stroker HP output

awww... what do i know, anyway?...

[Geronimo '74]

Originally posted by randywebb
[B

Can you afford 3 runs before and 3 after?? [/B]

3 runs before and after!! I'd rather spend that money on something else man. Do I really need to make three runs before and after??

I don't know whether the dyno will be a road or engine dyno.
To be honest I don't even know what the difference is between those two. (Dyno virgin remember)

[randywebb]

Do you really need statistically valid results? And, actually it's nice to have 5 per cell for any sort of ANOVA design.

So, if you want a good dyno test...

[Geronimo '74]

Originally posted by randywebb
And, actually it's nice to have 5 per cell for any sort of ANOVA design.

Two words: DYNO VIRGIN!!

you might as well be talking chinese now

[randywebb]

I only know a few words in Chinese...

The problem is that you could get a difference b/c of something else besides the change you made - it could be random or result from differences in air temp., pressure, moisture vapor pressure, different altitudes, tire squirm, squirrels caught in the gears, etc.

Or two effects could cancel out a real difference -- so you'd think the changes made do nothing, even when they really did help (or hurt).

Repeated measurements and statistical analysis are designed to overcome those problems. But don't worry - few car owners do this sort of thing -- only engineers at the factory or scientists.

[Geronimo '74]

What I was planning is to do two dyno runs in one day.
One in the morning, then install the chip and then do the second run in the afternoon.
This should eliminate alot of possible influences as well I guess

[Mark McClure]

Fantastic...how long does it take to change a chip? Why not open the box and have everything ready so that you can change the chip there and then and then run straight away. The engine would still be red hot and ready!

I may be missing something about the Steve Wong chip but it is only a chip change after all.

Cheers

Mark......

[Geronimo '74]

Marc,

Any relation to Troy McCLure? just kidding

Doing the two runs immediatelly after each other will be a bit too much.
I'd like the time to install Steve's chip properly without any rush.
But I think same day dynos will give me an accurate difference too

[KFC911]

Originally posted by Geronimo '74
What I was planning is to do two dyno runs in one day.
One in the morning, then install the chip and then do the second run in the afternoon.
This should eliminate alot of possible influences as well I guess

If you have your DME already opened up, it only takes a couple of minutes to swap the chips. I did six runs 'back-to-back', swapping between the stock chip and a SW chip between each run (I could change the chip in the amount of time it took the dyno operator to process the previous run). Doing more than one run on each chip adds a little more validity than a single run. Have fun!

[Grady Clay]

Dear Dyno Virgin,

First some basics: An engine dyno is used when the engine is out of the car. It is mounted on a support stand and attached to a brake via the clutch. It is possible to keep very close control of the engine temperatures and induction air. A good dyno setup and operator can get better than ±1% of mid-scale reading accuracy and better than that in repeatability.

A chassis dyno has the complete car strapped to rollers for the drive wheels. The engine power goes through the transmission, axles, tires, rollers and finally to a brake (usually). Some simply use a rotating mass and measure the angular acceleration. There are far more uncontrolled variables compared to a good engine dyno setup. This makes the measurements less accurate and sensitive to things like transmission temperature, tire pressures and temperature and more.

With any dyno keeping the temperatures under control is critical to good measurements (and not damaging anything. ) This is where I first used the “Rubbermaid Solution.” Another critical issue is exhaust and hot air reversion into the engine fan and intake. You can’t make good measurements if the exhaust is getting back into the air intake or the air temperatures are varying.

Most of the cost of a dyno run is the set-up for the first run. The subsequent runs should be relatively little cost. The advice to do six or more runs with the chip switched in and out each time is excellent procedure. The key is keeping everything consistent among all the runs. If you are unfamiliar with statistical analysis, post all the raw results and conditions data and let the Pelican Scientists have a go.

You should go home with both a printed hard copy and a disc with all the data in spread sheet format (Excel.)

Depending on the type dyno, each run should include acceleration to top RPM and then the dyno dragging the RPM back down. Each run can be multiples of these. The different runs are with the new chip installed and with the old chip installed – back and forth 6+ times.

Remember to show up with a full tank of the highest octane premium fuel and perhaps an extra 20L can or two. Dyno runs are fuel hungry, you don’t want to run out of fuel. Set the tire pressures high so the operator can simply bleed them to his choice of pressures. Check & set the pressures between runs. It would be good to do a pre-race tech checking engine & trans mount bolts, CV bolts, clean oil & trans lube, oil level, etc. You don’t want to spend dyno time servicing the car.

Best,
Grady

[randywebb]

I ahould add that it is entierely up to you as to how accurate you want to be... I assume you aren't planning to publish in a refereed scientific journal....

Let me paraphrase an old racing or hotrod aphorism:
Valid experimental design costs money -- how accurate do you want to be?

[Mark McClure]

Geronimo....you may remember me from such films as "dyno my lawnmower" and "sure my fiat has 600hp look at the dyno results"

[Highfield]

Geronimo

You have been following my thread re: my 256bhp on a SW chip.

Good advice given above. Find a recommended and friendly dyno operator and it is a breeze.

Arrive, hand the car to him, change chip, hand car to him, leave.

My dyno operator at Weltmeister had done 100s of these so no big deal to him (but it was to me as it is everyones first run sometime).

Changing the chip in the ECU is a 10 minute job so it gives the car a little time to cool down.

My operator checked the rear tyres had equal pressures and also checked the oil level !

He then gradually eased the car up through the revs and back down again, checking for the sound of the engine and the readings from the machine.

When he is ready, he will use wide open throttle in say third gear and you watch the screen on the wall - the graphs will plot in real time. I took a photo of my 4 runs as the horsepower built from 251 right up to 256. Weltmeister give as many runs as needed to give a reliable figure at the end.

It is key to have lots of cold air blowing through the cell so see photo below for likely setup.

Hope you enjoy it (treat it as a day out) and post the results when done.

[ssetek]

the post above is a a dyno dynamics dyno which is awsome, we have one and it was run in shootout mode which means if he were to bring the car to us and run it on shoot out mode he would read the same numbers, a dynojet will tend to read about 13-15% higher than the dd dyno,
as far as the sae corrected numbers statement every manufacturer will use a sae cor. number but the sae standards that some companys use are old and will yeild a higher #.

[KFC911]

I've only been to a dyno one time (5-6 of us Pelican's rented the dyno for 1/2 day and split the costs), and all of us drove our own cars for the dyno runs. Is this typical, or do the dyno operators usually drive them? In my case (even though I could have performed as many runs as I desired), the dyno results were so consistent (Mustang Dyno) that I could have stopped with a couple of runs on each chip and been happy. It was a learning experience for me (plus I got to see some NICE early hotrods on the dyno), so have fun!

[jluetjen]

Originally posted by Grady Clay

Remember to show up with a full tank of the highest octane premium fuel ...

Grady, normally I agree with you -- but in this case I'm not sure that I do. To be meaningful, shouldn't you have a full tank of the fuel that you will normally be using. In the case of a street car, this will allow you to ID any speed ranges where the engine may be pinging in the relative safety of the dyno, and then adjust the tuning. Doing your dyno runs with a tank of 110 octane with a aftermarket chip and some ignition advance tuning may result in some great HP numbers, and a blown engine once you fill up with a tank of 91 octane at your local service station.

In the case of a race car, race fuels are different from brand to brand and I think it would be very important to use the brand that you intend to race if you intend to get he maximum performance from the engine. If you get great performance from Phillips 66 114 octane (leaded), and then race Sunoco's 110 octane (unleaded), -- or vice versa! -- you may find that the engine runs different due to differences in the vapor pressure, RON, MON, density, burn speed and such. Which brings me to another thread...

[Grady Clay]

John,

The dyno can put sustained loads on the engine akin to speed record attempts. Perhaps the dyno operator doesn’t pay sufficient attention to my admonitions about temperatures. My suggestion to use high octane fuel is to protect the engine. Since this is a cat engine, he is unlikely to find even 100 octane unleaded. I suspect he can buy 95 or 98 octane at the best. I wasn’t suggesting he use leaded race fuel.

A knowledgeable dyno operator can spot the signs of detonation from the exhaust gas composition (pre cat) even without knock sensors. I think it prudent to use premium fuel and not rely on his ability to spot detonation before it becomes an issue.

I agree; it would be useful to know the octane requirements for this chipped engine. I think that should be the subject of another investigation where you have a baseline with say 98 octane and then carefully test using 95 or 91 octane.

With a well instrumented engine dyno, I would be willing to investigate the knock limit. I have done that but with great caution and still was nervous about it. To simply test for the effectiveness of a chip on a chassis dyno, I think protecting the engine is most important.


Let’s start a thread about fuels, dyno tests, knock sensing, etc. I have a lot to learn. I intend to get a well instrumented engine dyno up and working this winter.

Best,
Grady