OSE12P VE Learn Issues

[RANND]

I am a newb running an 808 with USB Serial Card, NVRAM, TechEdge WB, TP RT and OSE12P on a 355 Holden.

I have made reasonable progress with my basic tune and idle. I am logging good WB data and using the VE Calculator Tool to calculate the required changes, which I am uploading into the ECU. However, I am chasing my tail with the VE Table. It appears that the VE learn feature is applying minor real time changes to the ECU VE table during my road testing.

When I check the ECU VE Table after a tuning run, I am seeing changes in the table that I did not make. The "Map A: set=Wideband VE Learn" flag is off, so VE LEarn should be "off". My Narrow-Band Oxygen sensor is not connected, so it should not be affecting the numbers. Can someone tell me what I am doing wrong? I want to disable VE learn so that I can manually fine tune my VE table before I move to my AE tables.

[Holden202T]

your not running on map b are you ?

i'd make sure map b ve learn stuff is off too just to be sure. failing that i can't see anything that would be doing it, if you don't have the narrow band connected then there shouldn't be any narrow band ve learn ..... but maybe just disable it all and see if the changes stop ?

[RANND]

Thanks for the rapid response. I am reasonably sure that the Map B VE Learn flag was already off. Unfortunately, my OSE Plug-in stopped working after I moved the ECU from the car to my test bench and dislodged the NVRAM Map A/B jumper. After a few hours of unsuccessful troubleshooting, I am in the process of reloading all of my TP and OSE12P software to see if that fixes the problem. I will report back once I am up and running again.

[antus]

If you have moved stuff and plugged unplugged stuff you have probably corrupted the contents of the NVRAM, in which case you'd need to flash a base OSE12P bin back on to it.

[Holden202T]

the jumper you are referring to is two different memory segments on the NVRAM, the a/b maps i refer to are only able to be used if you put a switch on B9 and earth it, without the switch you'll be running on all the tables in the tune that say map a before them.

[RANND]

My ECU is running again after reflashing my NVRAM. It appears that my "VE Learn Issue" is actually a VE Table Upload Problem. I have discovered that VE Table values that I upload into NVRAM are out by between -0.19 and +0.19 when I download the data depending upon which cell I look at (see attached). I am also getting a "Could not write to change log file" error when I click the "Commit Changes and Upload" button, but get "Upload Success" when I click the "OK" button in response the the error. I get no reported errors when I use the "Upload Current Bin to Emulator" button. This problem is probably finger-trouble on my part, but the incorrect valuea are a worry - Any advice would be appreciated. I am going to investigate the "Could not write to change log"error in the meantime.

[VL400]

The ECU is an 8bit system and the VE calc is 0.390625 * X. So the resolution of the table is 0.39%, that amount of error is really not that significant.

[RANND]

Thanks. I knew there would be a simple answer. I have also discovered why I was geting the "Could not write to change log file" error, so I will continue tinkering with my tune. I am am really pleased with the results so far.

[antus]

Yeah, thats when you have not loaded a file from disk, and you make a change. TP normally records a log of whats changed in filename.log (i think) and as there is no file name, it cant, and reports this error. As you have found it is possible to corrupt the nvram so its good practice to have a copy on disk and in the ecu, and save your changes to disk before you exit tp. that way if you loose the nvram, you can write a stock bin to the nvram and then send your current cal, and continue on.