openecu.org

Has anyone here successfully modified these maps on any ECU to run higher (17+psi) boost levels?

I modified both maps this morning on my A4SG900C-based calibration, but the car continues to hit the fuel cut and get a P0244 at around 17.5 psig.

If you've successfully modified them, what map location did you edit and what did you change? What did you use for the axis of the map? Looking around the relevant locations on my ECU, I can't even find values for the axis.

Thanks!

/Andrew

Nobody?

Andrew, I haven't done it yet but I was planning on it once Enginuity is ready and I have time to help tune my buddy's VF22 (probably next weekend). I'll go take a look at the tables I was planning on changing and let you know.

Edit: I thought I remember more, but yeah, for fuel cut, it looks like BOOST LIMIT is the only table. I was probably thinking of BOOST SOLENOID CEL MAP THRESHOLD. Are you sure it's boost causing fuel cut? Do you have any logs? IAM and knock correction would be good things to watch..

Edit again: Any idea what THROTTLE DELTA FUEL COMP (BOOST ERROR) is?

qoncept wrote:Andrew, I haven't done it yet but I was planning on it once Enginuity is ready and I have time to help tune my buddy's VF22 (probably next weekend). I'll go take a look at the tables I was planning on changing and let you know.

Cool, thanks. I'd like to dial in the increased boost before the weekend, and I'm going to try cranking the maps up farther tonight to see if I can avoid the cut.

Edit: I thought I remember more, but yeah, for fuel cut, it looks like BOOST LIMIT is the only table. I was probably thinking of BOOST SOLENOID CEL MAP THRESHOLD. Are you sure it's boost causing fuel cut? Do you have any logs? IAM and knock correction would be good things to watch..

I'm positive it's the boost limit fuel cut... I've felt it before (looooong time ago when I was fiddling with manual boost controllers)... upon hitting the cut, pulling over, and checking the code, it was a confirmed P0244. At that point, I reflashed a less aggressive map and kept going, as I was already late for work.

I was logging at the time, but, unfortunately, ecuExplorer has a habit of dropping out during high load regions... and the time surrounding the boost cut got lost. IAM was at a nice, happy 16, and KC was in the region of 8 or so (for a total of 13 degrees BTDC at peak boost/load).

Edit again: Any idea what THROTTLE DELTA FUEL COMP (BOOST ERROR) is?

Never seen that variable.

I'm sort of at a loss, since I don't have any ROMs with an increased boost limit map to check against, and I can' wrap my head around the disassembly subroutine that seems to reference the table.

At this point, I'd just settle for knowing that I'm editing the right area of the map, and, perhaps, finding the axis for the map.

/Andrew

Are you using ecuEdit to alter the values or just a hex editor? If you send me your map and xml file I can take a look at it..

I know with the 2D maps I've looked at, the offsets start with the axis and then the table data (so half are axis values and the other half would be the boost values). So your axis should be right there.

qoncept wrote:Are you using ecuEdit to alter the values or just a hex editor? If you send me your map and xml file I can take a look at it..

I know with the 2D maps I've looked at, the offsets start with the axis and then the table data (so half are axis values and the other half would be the boost values). So your axis should be right there.

I'm using ecuEdit, but I've resorted to doing hex editing to actually make the changes in maps like this... ecuEdit has some endianness problems that I can't seem to sort out.

However, in the stock A4SG900C map, I've found the data at 0x29924-0x2992F... there's a total of six two-byte entries. There's no obvious axis data anywhere in the vicinity.

Attached is a screenshot of the presumed stock boost CEL map.

I could be way off here, but it looks like RPM right after the boost values -- the only descrepancy is the number of them (edit: and the gap between those values and the table values). The values at the right side of this table look reasonable, but they don't account for the extra RPM values I (think) I see in the image, and the left side is... possible.

You can put this in the Enginuity demo xml file and play around if you want to check it out.

Code: Select all

   <rom>
      <romid>
         <xmlid>A4SG900C</xmlid>
         <internalidaddress>200</internalidaddress>
         <internalidstring>A4SG900C</internalidstring>
      </romid>
      <table type="2D" name="Boost Limit" storagetype="uint16" endian="little" storageaddress="0x29924" sizey="6">
         <scaling expression="x/93.7" format="#.00" increment="1" units="PSI" />
         <table type="Y Axis" name="Engine RPM" storagetype="uint16" endian="little" sizey="6" storageaddress="0x29943">
            <scaling units="RPM" expression="x/256*50" format="#" />
         </table>
      </table>
   </rom>

I'm going to go ahead and assume what I just posted was wrong.. but it just occured to me that the ECU doesn't store relative pressure, at least in other tables, it stores absolute pressure.. So while those numbers look pretty reasonable, there's a good chance we're missing something. It also stores boost as 1 byte elsewhere instead of 2.

Yeah, those RPM values actually belong to the first of the two "mystery maps" that I've posted about in a thread elsewhere on these forums.

I increased the values in both of those maps by an additional 0x60 (for a total increase of 0xC7), and didn't suffer any boost cut tonight. I need to tweak my wastegate duties, but it seems that those are, indeed, the correct maps... still doesn't give me anything for the axis.

I've seen some screenshots of other tuning tools that show 2-byte axis and 2-byte data values for those tables... but the quality on them is poor (they're from an old Sport Compact Car article) and I can't see the details.

/Andrew

Kha0S wrote:I've seen some screenshots of other tuning tools that show 2-byte axis and 2-byte data values for those tables... but the quality on them is poor (they're from an old Sport Compact Car article) and I can't see the details.

Hmm, have a link? Was it Prodrive software?

qoncept wrote:

Kha0S wrote:I've seen some screenshots of other tuning tools that show 2-byte axis and 2-byte data values for those tables... but the quality on them is poor (they're from an old Sport Compact Car article) and I can't see the details.

Hmm, have a link? Was it Prodrive software?

No, it was the software from the Six Letter Company Starting With E.

http://www.sportcompactcarweb.com/proje ... c_projwrx/

The images are small, but you can figure out the general range of the data.

Hmm.. I wonder if we could find some better screenshots of those same tables from an app made by a 4 letter company starting with a C..

qoncept wrote:Hmm.. I wonder if we could find some better screenshots of those same tables from an app made by a 4 letter company starting with a C..

Quite possibly, and likely with less threat of litigation, as Trey seems to be okay with people "learning" from the off-the-shelf maps.

Certainly, anything published in documentation or articles (such as the above article, or the StreetTuner manual, etc.) can be used for our purposes with no risk.

I wonder if Ecuscratch's values are actually pulled from the ECU or static..

I really wish I had a copy of StreetTuner to tinker with, but that's the reason I'm here I guess. I'm surprised Trey is so willing to let people toy around -- Cobb seems to be pretty keen on IP as well.

I was logging at the time, but, unfortunately, ecuExplorer has a habit of dropping out during high load regions...

what comms settings are you currently using?

calvin.