RomRaider

NSFW,

I couldn't patch my current rom with these files, i assume it's because i patched with your LC/FFS prior to this.

i did however patch a completely stock rom, and got the SD tables, but it's missing the MAP psi values on the top line.

i attached a pic of the stock rom, and the failed attempt on my current map.

Is there a way to get these definitions after patching to the SD 540C Rom for RomRaider?

Stock Rom Missing psi info:


Failed patch of LC/FFS current rom:

I was able to get it to patch, after removing the old LC/FFS patch.
I still have the xxx in psi axis problem.

any ideas?

I tried Scoobyrom for the first time, and managed to find the SD VE Table, and created a DEF for RR.. but i think i screwed it up... not showing psi, and the values are different that in EcuFlash.

Pic Attached:

Scoobyrom/RR is reading the raw floating point value of the column used by the ecu, units are Torr. Close to mmHg..

Something is wrong with the ecuflash definition, probably a scaling issue. Post what you're using.

i'm using the def from NSFW's last post.

i wish i could help you guys with this programming stuff. but all i know is little vb.net.

Merp, how hard would it be to make the CEL flash on knock?
thats the main thing i miss about my UTEC. That and spark-cut launch control.

Looks like you're using the standard ECUflash metric definitions which do not include a 'psirelativesealevel' scaling in the 32bitbase.

Add this line to the scalings in your rom's xml file.



I've identified a straightforward porting process for the cel flash stuff. In order to simplify things, I'm planning on condensing SD, CEl flash, and programming mode into a single patch, then port a single patch instead of porting 3 different patches.

I hope to have the programming mode stuff testing on my car before the end of the month, or sometime in February if I switch over to HEW and C coding. Ports shouldn't take much longer. NSFW is working on a RAM dump utility, when it's ready for release it will simplify the porting process even further. No more preliminary test roms... just check the RAM hole with the utility, set the addresses and flash a true test rom.

that didn't work, but i took the values and added a new scaling with the scaling manager.


here's my VE map that will idle and cruise. idle at about 2% correction
it's very rich under boost.. 9.0:1 or so.



im waiting for NSFW to check his logger.xml, because almost everything is screwy when i use it.

how should i start to build a VE map?

I suppose my next question is, why is the VE map only 13 columns, when the fueling and ignition are 16?
I'd like to get as much resolution as possible, since i plan on going to 27-30psi.

Freon set them up that way to minimize the number of cycles the pull3d routine performs when pulling the values. Basically, bigger tables are slower tables, and 'just enough' is optimal. I know standalones run HUGE tables, but don't forget that the stock ecu is bogged down with a bunch of other stuff (OBD, Emissions equipment, etc), and was designed/coded with the stock motor in mind.

I'm going up to 34psi and resolution hasn't been an issue, if it does present a problem, adding a few rows/cols is a snap.

The column values shown in your screenshot are correct, so your scaling is good. I'm still curious why it didn't work the first time though. I like Merp's theory that you were using the metric definitions for EcuFlash, was he correct? If your target boost table shows boost in bars, then that was probably it.

Are you using the logger.xml from the last post on page 5 of this thread? I'm pretty certain that the logger.xml that I posted is the same one I copied to my CarPC when I flashed my SD-patched ROM, and I didn't see any problems. I will copy the one from my car tomorrow and compare it to what I posted, just to be sure.

What do you mean by "almost everything is screwy" ?

Try creating a new profile and adding the SD parameters again, if you haven't already. The SD parameters all have names that start with "SD," like "SD Calculated MAF."

I was planning to build the VE map by leaving it in regular-MAF mode, and logging regular-MAF and SD-MAF, MAP (psi relative sea level), and RPM. The challenge is how to collect logs in every MAP/RPM cell. Once you have data from a bunch of cells, tweak the VE table by an amount equal to the ratio of regular-MAF to SD-MAF. Or maybe 75% of that ratio, to avoid overshooting.

The cells top-left corner of the VE table (idle and cruise), should be easy. The lower-left is basically high-RPM cruise, so downshift and drive at a constant speed for a few seconds at 4000 RPM to 7000 RPM. With a manual boost controller you can get most of the left half of the table just by progressively reducing boost with the MBC until you reach wastegate-boost. I'm not really sure about the area in between WG boost and cruise, though. You could use gradual acceleration to cover some of it, but try to avoid having RPM or load or boost changing rapidly. The faster things are changing, the less accurate things will be.

And for the cells that are really hard / impossible to reach, like 20psi @ 2000rpm, just guesstimate based on the shape of the table in the cells that you actually can reach.

When I actually switch to SD, I was planning to cut all of the values in my dynamic advance table in half, or maybe even zero it out. That creates some room for error, just in case. Then verify that AFRs are reasonable for wastegate boost, and gradually turn up the boost until full boost is reached. If AFRs are good, then put the DA table back where it was (maybe in 3-4 stages, just in case). I follow that basic process a lot when I make big changes. I'm told that reducing timing that much causes high EGTs, but I don't have an EGT gauge to see how much higher they get. I don't think it's done any harm, but let it cool down between pulls just to be sure. I always did that anyway, just as a side-effect of checking my logs between pulls.

Also, I just noticed that there are a crapload of experimental parameters in that logger.xml, I had forgotten about those. They shouldn't cause any trouble, other than polluting your list of parameters. I posted the useful ones in that 'experimental stuff for 522N' thread. The rest of them are things that I couldn't figure out, or I figured them out but decided that they aren't useful (who needs "timer ticks?"). Or both - many of the compensations are always zero, so I don't know what they're for, but I don't think it matters.

I think resolution is over-rated. It's really only necessary if you want to create a lumpy table.

But as Merp said, it's not hard to change add more if it turns out that this isn't enough.

i'll get a screenshot at lunch, but like my final fueling base was 112184664688255.54 and my corrections were 100%, and injector duty cycle was like 9000%..

i hadn't even added the SD parameters yet.

Weird. That's enough info, don't worry about getting a screenshot.
Can you post your OLD logger.xml? I'll diff it against mine and see if anything interesting pops out.

Attached.