jahwerx

Prior to making any mods to the engine, I wanted to run a baseline dyno:
MY 2000, 19,000 miles, 93 Octane gas with only the following modifications that would impact the dyno run from bone stock:
- Spoon lightened flywheel (and clutch)
- Regamaster wheels with Pilot Sport Cups

I did three pulls - the first was bunk because the RPM signal went haywire.
We found a reliable clipping point and performed the first "good pull". Keep in mind that the engine was as "heat soaked as it could be" for this run (radiator fan was cycling, etc.) since it took a while to get the signal.
We let the car cool for about 20 minutes in front of a fan, and did a second pull . . . (the temperature gauge indicated 3 bars in both cases)

Here are the results



This is just the second, better run with temp and pressure stats as well (we are at about 800 ft elevation)



Note the fairly dramatic difference in hp/torque curves (~5 hp and 3 ft-lb) and A/F (up to 1 full point across the operating range).

1st thing - I'm pretty happy with the numbers
Could these differences be attributed purely to engine temp? The air temp didn't change significantly over our test period.

brushman

Your numbers are good.

For reference: Stock with 625 miles

jahwerx

Brushman,

What do your two runs represent? It looks like you are seeing the same sort of differences in HP and torque that I am...

By the weigh (pun intended), shaving 10 lbs at each corner could attribute to the *absolute* increase over your peak numbers.

I'm more interested in figuring out what the computer is doing to change the A/F so much...

- Josh

brushman

Runs represent: (stock with 625 miles on car w/fan in front) No change in air temp or humidity from one run to the next)

Dynorun.001 just warmed up with 3 bars
Dynorun.002 15 minutes later

I attribute the increases from run to run in ECU programming the fuel to better suit the conditions at that time. (this is my guess, UL would be able to better answer this question) On your next runs try pulling the ECU fuse or disconnect the battery for a few minutes and watch the difference. I believe that ECU has 2 settings WOT and % of full throttle. The car tries to use it's sensors to gather information, at that time, to maximize power at WOT.
(Note: I am putting my flame suit on as I am by no means an expert)

Remember that dyno's are not the end all be all in how the car performs debate. Unless you can reproduce EXACT conditions from one run to the next, your results will change. Different days = different results; Different dynos = different results.

Every car I have ever dyno'd has shown some increase form run to run. At a certain point they seem to hit a wall and show a small decrease (my guess in heatsoak).

Gernby

There are lots of other temperatures that could cause the numbers to change. Hotter engine, transmission, and differential oil will reduce loss, and if your tires heat up, the air pressure will rise, which will reduce rolling resistance.

jahwerx

Bottom line it sounds like I can chalk up a 1-2% difference in numbers to "noise".

Thanks folks - I'm good with that. I will now be forever super skeptical for mods which claim a "5 hp gain"

brushman

Look for what the mods do across the range of hp an tq rather that total top end.

ultimate lurker

If you want to get accurate resolution on performance changes you _must_ monitor conditions. Engine coolant temp and intake air temp can dramatically affect how the engine behaves, causing large variations in timing and mixture. That's why on all OBDII cars we always plug in an OBDII scantool to monitor these things and make sure runs are made at the same temps.

We still throw out the first run because usually things haven't stabilized, tranny fluid has cooled (more losses), etc.

The variations you saw jahwerx are almost certainly due to coolant and intake temp variations. In fact, if you had just told me your intake temps during both runs I could have predicted the results for you (actually, just your test procedure would have revealed this).

Which number actually represents your real world performance depends on what the actual temps were vs. the temps on the road.

UL

jahwerx

Brushman - good point - peak numbers are meaningless without taking the "area under the curve" into consideration.

UL - well summarized (as always). Thanks for the feedback - In subsequent testing/tuning I'll be sure to capture more data to help correct for these variables.

One additional question - Under controlled testing what is the accuracy of the dynojet? - I know they report down to the tenth of torque and HP, but is that a bit optimistic?

ultimate lurker

I don't know that any dyno has accuracy numbers. What is more important is repeatability. I believe the Dynojet is repeatable to within at _least_ 1%. I know my Dynapack guarantees repeatability to within 0.3% and the Dynojet can't be much, if any, worse.

UL