IMAPCalcs is affecting WOT load calc.

[richardh]

Originally Posted by black e

Damn, I got left out

I take it you want more ROM IDs done?

Will flesh out goldenevo.com more tonight.

Rich

[UT_EvoX]

Originally Posted by richardh

I take it you want more ROM IDs done?

Will flesh out goldenevo.com more tonight.

Rich

Thanks Rich!

Do you suspect any of the other MAP paramaters will affect my load calculation as I described it? If you can get the rest of those ones for 5268xxx I would be more prepared to get my car on the dyno tonight and I'll let everybody know which fix solved my WOT MAP calculation issue.

[black e]

Originally Posted by richardh

I take it you want more ROM IDs done?

Will flesh out goldenevo.com more tonight.

Rich

Thanks. I did figure out the address for my ROM.

[UT_EvoX]

It worked wonderfully. No more jagged load calculation, no more jagged IDC. I seriously was seeing like 10 IMAPCalcs spikes in one 3rd gear pull.

[black e]

So you just upped the load multiplier table that Rich posted?

[UT_EvoX]

Originally Posted by black e

So you just upped the load multiplier table that Rich posted?

Well, it's a hysteresis table, but yes. I increased it from 0.4 to 0.8% and it worked spot on.

[richardh]

Originally Posted by UT_EvoX

It worked wonderfully. No more jagged load calculation, no more jagged IDC. I seriously was seeing like 10 IMAPCalcs spikes in one 3rd gear pull.

Awesome.

I didn't work this stuff out, by the way. Tephra was the one who pointed me at the "hysteresis" value, and suggested the tweak as a way to eliminate the dodgy load wobbles. Worked a treat.


Now, let's back up a bit and answer a more basic question... asked by Kooldino a short while back...

WTF is IMAPCalcs?

As noted in older research into load calcs (not mine), the ECU tracks three inputs...

1. MAFCalcs - derived from MAF (MAF Scaling tables, etc.)
2. MAPCalcs - derived from MAP (3xMAP Tables, etc.)
3. IMAPCalcs - "interpolated" variant, derived from MAP.

Fun fact A: The ECU chooses the "middle" of the three.

Fun fact B: Item 3 should always be lower than 2.

Now, items 2 and 3 are kind of joined by the same rubber band... there is an interpolation/ratio mechanism that acts on both. That ratio mechanism gets buggered around with by a TPS-based input.

If throttle is nailed to the floor, MAPCalcs should be nice and steady, and IMAPCalcs should be zero.

But... the ECU does weird stuff behind the scenes with its TPS, including some odd "decay" processing over time. TPS can jitter slightly, and the default "hysteresis" value makes the load calc extremely sensitive to TPS jitter.

So...

IMAPCalcs is the canary in the coal mine. If you get any TPS jitter, the MAPCalcs may step up and down a little bit... but IMAPCalcs will look like a horrible sawtooth nightmare. It's a huge red flag.

If you see an IMAPCalcs sawtooth pattern, you have a TPS->load-calculation problem.


And how do you log IMAPCalcs? Why, RAX Fast Logging, of course!

Rich

PS. Here are some before-and-after images. Ignore ( A ) and look at ( B )

BEFORE:



AFTER:

[black e]

Aahhh, that explains a lot , thanks!

[richardh]

Guys, if you change that hysteresis value from 0.4 to 0.8 and still get IMAPCalcs spikes at WOT, feel free to try 1.2.

Rich

[dj_xmx]

So if it looks like this its good then lol




Sent from my iPhone using AutoGuide.com App