What do I have going on here? (Chrome/ECUFlash/Evoscan)

[Greg E]

BTW I know I had the option to log the raw voltage from the fuel tank sensor (pin 93) and I'm pretty sure the raw MDP option is there too. Best to try logging them with the scaled and see if the raw voltage that you think you're sending to the ECU is actually what it's seeing.

[RealMcCoy]

Yes A1 is FC45.

I fed a known voltage to the ecu (by controlling the pressure of the map sensor, and monitoring it at the ecu connector ) precisely because it didn't work when I flashed the scaling. I was trying to diagnose why, and wanted to eliminate the possibility of a problem in the car, sensor or wiring. The voltage getting to the ecu does match the scaling at whatever pressure I have it.... The numbers coming out the other end aren't even close.

[Greg E]

Email me the file currently flashed to your ECU and let me dig into it to see if there's anything wrong on this end.

Nonstopmix@gmail.com

[RealMcCoy]

ROM sent...

Also trying to understand where AFRWideband comes from....

This confused the hell out of me as well.... I just figured it out this morning. I found out something about ECUFlash I didn't know.

If you go into an edit table screen, and select a different scaling, it stays that way until you select a different one... It doesn't matter how many times you close the ROM and open a different one, until you go back and put it right, it is looking at the wrong scaling.

No big deal until you open a ROM not realizing that something you did just screwing around with an unsaved file two days ago set you on a path to destruction....

So apparently at some point, I was screwing around with the scaling, and ended up with AFRWideband selected. I didn't know it, don't remember doing it. So every ROM I flashed afterward until I caught it had that scaling in it...

[Greg E]

Ah yes I know exactly what you're saying. That explains this.

Would you like me to send you another XML file? Replace it with your current one, then open ECUFlash, open your ROM, then the sensor scaling tables and the data will look strange. If it does, that was your problem.

I can verify the same thing when I get home after work.

[RealMcCoy]

Ah yes I know exactly what you're saying. That explains this.

Would you like me to send you another XML file? Replace it with your current one, then open ECUFlash, open your ROM, then the sensor scaling tables and the data will look strange. If it does, that was your problem.

I can verify the same thing when I get home after work.

I still have the files you emailed me originally that have been untouched (I still have the email with the files attached) that was how I caught what was going on, I started working on the files on my desktop, and the problem I was expecting to see with the wideband was not there...

I'm not convinced the problem I had with the wideband scaling is the same thing that's going on with the MAP, but just to be safe, I am going to replace all the files with versions I know have not been touched and see what happens...

[Greg E]

Well crap... In the file you sent me, the MAP Scaled numbers match so that means you didn't change that table scaling. I also looked at the ROM in the hex editor to make sure you didn't change the address of a table and screw the code but the segment that does the calculation for the MAP sensor also matches.

This also confirms you're dealing with a separate issue than the wideband.

So, here's what I think we should check next.

Open EVOScan
Right click*Add New item

Fill in this info:

Display: Manifold Differential Pressure
LogReference: MDP
Request ID: 38
Eval: 0.01952*x
Unit: Volts
Gauge Min: 0
Gauge Max: 5
Chart Min: 0
Chart Max: 5
ScalingFactor: .01
LogHide: False

Click okay, start your engine, log and go for a little drive.

Report back what the sensor is reading.

[RealMcCoy]

Well double crap... now it's working.... I have now idea what fixed it or when..

I have some work to do to to make the numbers match, it's a couple pounds off up around 15psi.

The part that screwed me was when it was having the issue, the numbers at atmospheric were exactly the same as they are now... (About 1.5psi low) I was expecting some movement, so I didn't test it through the full range every time I tried something. I would just hook it up, see the same numbers, assume no change, and move on to the next try... So I really don't know for sure what fixed it. It may have been the wideband issue screwing with it for all I know now...

So the next question, if I am going g to adjust the formula a little to make the pressures match, do I need to do so in both the scaling and Evoscan? I'm not really familiar with how the math is done when it's built into the scaling.

[Greg E]

Doesn't really matter. Offset the EVOScan formula or change each cell in the table to correspond. Don't change the scaling that the table uses.

[RealMcCoy]

Doesn't really matter. Offset the EVOScan formula or change each cell in the table to correspond. Don't change the scaling that the table uses.

I think I finally wrapped my brain around it... (I'm a little old and slow, but I just had coffee...)

At first glance without looking at the numbers, it appeared the math was being done twice, once in the scaling and again in evoscan... I was confused on what was going on. (mostly because I didn't look close enough...)

The scaling is taking a raw voltage value and converting it to PSI, represented in a hex 1byte value (0-255 decimal). The formula in evoscan is converting that to an actual psi value we want to see on the screen.... Correct?

It looks like I can just bump the formula offset in evoscan from -20 to -18.6 and it will put me right there. I do think I'd prefer to have the car output the correct numbers for someone who hooks up to it later, as this car lives in the Bay area, and I'm in Oregon. It's reasonable to assume I'm not going to be the only one that ever hooks Evoscan to this thing. So once I test my numbers in Evoscan, I'll probably go back and adjust the tables to match what I ended up with.