scheides

iTrader: (10)

Looking for some feedback:

So my initial thought is to go in and change the scaling of the MAP sensor, swap it in, and be done with it, but looks like with 9653xxx6 I'm SOL and will basically have to re-tune my SD VE tables, then rescale the direct boost scalings (and then re-do the tables) for proper boost control?

Step one: getting VE correct. JB said to just use 1:1 on both VE tables, which jives with what he has posted in the past. This should get the car driving, disable error correction and get boost/afr back to where I want it.

Step two: change my scalings to match mrfred's 4-bar settings in this thread:
https://www.evolutionm.net/forums/ec...t-control.html

<scaling name="BoostErrorPsi" units="psi" toexpr="(x-128)/(8*4.23)" frexpr="x*8*4.23+128" format="%.1f" min="-3.8" max="3.8" inc="0.1" storagetype="uint16" endian="big"/>
<scaling name="OffsetWGDC" units="WGDC Change" toexpr="(x-128)/2" frexpr="x*2+128" format="%.1f" min="-64" max="63.5" inc="0.5" storagetype="uint8" endian="big"/>
<scaling name="psia16" units="psia" toexpr="x/(4.23*2)" frexpr="x*(2*4.23)" format="%.1f" min="0" max="32" inc="0.1" storagetype="uint16" endian="big"/>
<scaling name="psia8" units="psia" toexpr="x/(4.23*2)" frexpr="x*(2*4.23)" format="%.1f" min="0" max="30.1" inc="0.1" storagetype="uint8" endian="big"/>

Then update the values in evoscan:

0.2369*x-b

per this thread:
https://www.evolutionm.net/forums/ec...evo-8-9-a.html

log, turn on error correction, and should be back in business.



Am I on the right track here?

***edit***

I was on the right track, but the scalings are already included in phenem's default XML that comes with 96531706, but a key component I missed that ended up being a big deal (and would have made this painless) is the kPa MAP 16-bit scaling:

Once I changed this and put my original values in there, it was absolutely painless.

RoadSpike

iTrader: (5)

I was under the impression the scalings included with phenem's and bank's effort for the xml was good enough for boost control.

All you should have to do is right click the table and change the scaling to one of the canned omni 4 bar ones. Then you just adjust your table back to the corrected values and call it a day. Shouldn't take that long to swap out the default scaling to the correct omni 4 bar.

Beats messing with the default scaling if you want to go back as well.

scheides

iTrader: (10)

*hug* I didn't even notice that there were alternate scalings included! Default is PSIa8 JDM3barMAP and yea, I'll just switch to PSI8 OMNI4barMAP, update EvoScan, and go nuts.

Just to confirm: all of my VE tables will need to change/be re-tuned, right?

RoadSpike

iTrader: (5)

I'm assuming you had a 3 bar in there yes?

If so yes you will need to update your map VE table with new settings the values will change when you update the scaling.

The RPM VE table should be fine on its own likely but you never know some map sensors have some different quirks then others which need to be tuned for.

scheides

iTrader: (10)

Ya, 3bar.

Yes I'll need to update my map VE table, but those values are not directly related to the scaling changes mentioned above, so it will simply be re-tailoring the VE based on the different scaling/voltage values of the new sensor.

This seems kind of odd to me; why is there no MAP sensor scaling as part of the SD implementation? You'd think you could swap out any sensor for any sensor, and as long as they were each scaled properly, you wouldn't have to re-tune.

RoadSpike

iTrader: (5)

That would be nice but would also slow down the calculations for SD considerably as the ECU would then have to look up many tables in order to translate the output. With this method the scaling just allows you to set a prettier looking target for the ecu to match so it doesn't have to worry about what 1.5 volts means for any sensor

JohnBradley

iTrader: (30)

I no longer use 1:1, it has drifted somewhat. I think I have it posted somewhere else around here but the main changes are a little less at idle and cruise (i.e. 61-->57) and out the top it seems that all turbos have worked well with 340-->430.

scheides

iTrader: (10)

How about RPM VE? Here's where I'm currently at with the JDM MAP sensor:

l2r99gst

iTrader: (2)

You guys have a little different tuning style of SD than I do. My style matches more of what John had when he introduced this.

You can see the trends if you multiply your maps out into a 3D chart in Excel or something.

VE is going to be a lot less for map under vacuum (part throttle) than during 0 vac/boost (WOT). So, for my map ve tables. It tapers way down at lower vacuum ranges, then to 1:1 ish at WOT. It makes for a much better correlation to what the VE was when on MAF.

Of course 2 2-D tables aren't perfect for trying to capture a 3-D chart, but if you multiply them out, you will get what I am trying to say.

Edit: Here is what my tables looked like, so you see what I am saying. This is of couse on a stock turbo, with stock exhaust, so VE falls at the higher RPM:

Slo_crx1

iTrader: (9)

Here's what I have so far for my Omni 4-bar. I never really bothered taking the load matching out past 300%/kPa since for now I don't peak over 285-290 right now.



Also, I don't know if it was just me or anyone else had the same issue, but I could never seem to get the PSI-based boost control to work at all. All it ever gave me with any setting was 12psi...plain wastegate pressure. Once I swapped the hex values back to Load-based it works fine, so I left it like that and tuned the load-based tables instead.