Hi carlossus.

I'm potentially interested in what you're proposing. My motivation would be to understand my Revo remap better, because it's not practicable or reasonable to reverse engineer it. Problems for me are that, while I recognise the power and simplicity of setzi's logger, it looks a bit too oriented towards hardened IT professionals , although having grown up on early Windows and Unix command line inputs, that's not too much of an issue. Someone would have to help me in specifying variables to log.

Second issue is that until I get a new battery for my laptop, I can manage about 15 minutes away from a mains power source.

Third issue is that my code isn't stock. I have asked the question on the main thread for setzi's logger as to whether this would be a showstopper.

http://nefariousmotorsports.com/forum/index.php/topic,837.msg12182.html#msg12182

Good idea though and if I can support the effort, I will.

TTQS

 
I just noticed you used the word practicable. I had to smile because I have a feeling you have proabay had to deal with ALARP?

Please find attached my first log with setzi's ME7logger. This was done on a short drive moving through the gears and engine speeds so although there is some sustained high load operation, there is actually no clean WOT pull from just over idle to redline in there.

There were originally 127 variables as per the attached configuration file, but on review, I deleted those that were not useful (i.e. entire columns of the same number) and reordered some data so that the first 28 columns (indicated with pale yellow shading) contain some of the most commonly logged variables. I will probably prepare a new config file with only about 30 parameters in it for everyday logging.

I haven't done any analysis of the data yet but see what you make of it and what it can reveal about the Revo Stage 2 tune. I was surprised to see enrichment deployed in lamfa_w at high torque, especially in the light of recent discussions regarding using this for tuning and my general aversion to that. Perhaps I should start to prepare one of my hats for consumption...

TTQS

Hi Carl.

I haven't tried ECUx Plot yet. To be honest, everything has been done in a bit of a whirlwind over Christmas and New Year when I sort of suddenly realised that I had a small window of spare time to come on here again and start getting involved in interesting discussions, etc. Then I bit the bullet and started investigating setzi's ME7logger whereupon I twigged that with the GUI, it was simple enough for me to use.

After I had played about with it, I realised just how amazingly powerful it is. Then I ordered a cable which only came yesterday and I had about 20 minutes to get a log before going to a family get together!  

I'm 100% behind your idea of a statistical load-bearing dyno supported by real-world driving conditions and my motivation for that is to define/populate the common maps with Revo Stage 2 tuned data so that I can learn from a 'good' commercial tune and then feed that knowledge back in to my little tuning awareness document. As you comment, I agree that the part-throttle behaviour is just as interesting and important as WOT behaviour, particularly because I feel that the Revo Stage 2 on my car yields effective performance at part loads too, not just when mashing the throttle.

Anyway, the only thoughts I had last night were about filtering the raw data into a batch of three columns (for each x,y,z kennfelder obviously) then writing an Excel macro to sort it, populate an internal array, then interpolate it back to original axis sample points and write it to a table in a worksheet in the same format as in my tuning document. Then compare changes from stock to tuned parameters and attempt to understand why the particular changes have been made, etc.

I'm acutely aware though that it's not going to be straightforward to populate all zones of tables (e.g. low nmot, high load and high nmot, low load). Once we've dived in and had a play, we can kick around ideas on the way forward on this thread. I'm up for it! I had a quick look at ggobi. It certainly looks glossy but I reckon I can achieve my goal with Excel and a tidy macro.

TTQS