Report bugs found in TunerPro. Please be sure to include as much info as possible, including system specs, OS, repro steps and TunerPro version number.
I've run into a problem where I can't trust the data from a selective export from a .xdl log file. If I export everything, the data seems OK in the .csv file. If I only export a few selected items, some of the data is bad in the .csv file. Attached are example files from a short benchtop test.
Here's what InjPW(ms) looks like when plotted from the full .csv file (top, good) and when plotted from the .csv file with only 4 exported items (bottom, corrupted):
good_vs_corrupted_export_of_InjPW.png (38.2 KiB) Viewed 172 times
Is this a known issue? Any way to get selective export to work more reliably?
Thanks for looking at this. Let me know, either here or direct email, if there's anything else I can get you to help with the debug, or if there's additional info/alternate config that might help.
I have a benchtop setup with knobs and switches to exercise the LH2.4 ECU. I've been trying to decipher the full LH2.4 fueling algorithms, which is why there's ~60 bytes being monitored in the .adx file. I know this is pushing the TunerPro capabilities -- I'll back off to 10 or 20 bytes, with a big delay between frames, for the basic on-the-road monitoring.
Are you sure, when doing a selective output, that you're selecting both the "Injector PW full (ms)" and its dependency, "Injector PW H"? If you don't and only select the former, you won't get valid data. When I first reproduced the issue, that was the mistake I made.
If you are still able to repro, so that it's easier for me to debug, could you narrow down your ADX to contain the minimum number of commands and values to repro the issue? I'm attaching a modified version of yours as an example. This still works with the log you sent.
Let me know, either here or direct email, if there's anything else I can get you to help with the debug, or if there's additional info/alternate config that might help.