yep, you jsut lost me there.

i can get the dynorun by doing a high speed log and importing the run to where? will this show me the dyno info?

 

Yup, it's the same info. Actually it's even the same code underneath, the Roaddyno option is just trying to do more things at the same time.

Alls ya got to do is collect a highspeed log, then trim off all the extra data (everything before you started the run and everything after you let off the accelerator). The import dialog will prompt for which column of the log has rpm, afr, knock, timing, and timestamps. There are actually several ways to specify time, but for pyxede logs just select "from data" and it's always column 0.

cheers

d

 

do a regular highspeed log to a csv file.

then on the Duno Runs screen, go to Data, Import. Pick the log file you created.

then you have to tell it what colums are what.

RPM: 1
A/F: 3
Knock: 2
Timing:4

Powercalculation: timestamp: 0


that should be the most common.


Donour, if the dyno run and highspeed logging use the same format, why cant you just use the 'Load' to open a highspeed log? gives an invaild file format error. i havent been able to do a dyno run, so i havent been able to look at the files it makes.

 

subscribed

 

Because they're not the same. They are generated from the same internal data format, but a dyno run has torque/power along with a given set of other data. Currently this is AFR, knock, and timing, but later it will have things like the car model, date, location, etc.

You can only 'load' files that you 'save' from the dyno run dialog. Once you have the run imported, goto file->save. The resulting CSV will have torque/power already calculated as well as any smoothing/trimming that you did. You can then open it directly or give it to somebody else if they want to open it in excel or something.

d

 

did a high speed log this morning....

got the excel spreadsheet but the entire timing table column was 0 all the way down. any idea why?

and also, did you say you could somehow get the road dyno out of the high speed log?

 

Subscribed - this is nice work donour! When I get ahold of an XEDE I'll have a lot of fun with this! I haven't ran anything yet, but it looks like you've added quite a bit since v1.0.

Just a couple of questions -- and BTW sorry if I've missed something obvious

*You're using gnuplot to plot the graphs right? What's the crpmplot module in the C folder? Is that something from CVS, or is that something you're working on?

*In dyno.py (class car) it looks like your hardcoding some of the car attributes -- such as weight, gear ratios, wheel size etc. Do you think it would cause any problems if I tried to put these values in a separate config file (maybe an xml file or another comma delimeted file) -- or are these values able to be changed from within pyXede already. I have very little experience with python, but I'm a C and perl guy so I'm sure I can figure it out.

Thanks again!

 

Were you just idling? The timing value is offset from top dead center. If your engine has basically zero load, then there's not any advance/retard applied.

Yes, use the import feature. Read back a few posts in this thread.

d

 

no, i did like a 3k-6500 pull.

 

One other thing -- could someone post their road dyno file?? It doesn't have to be anything special, I just want to mess around with a few things. Thanks!

 

Just multiplots. The other plotting is handling internally by the native windowing kit. In other words, the plots are actually part of the pyXede windows.

Oops. You weren't supposed to see that. When you're trying to do stuff realtime, performance becomes really important. I've identified some parts of the plotting code that would really benefit from the features of an agressively optimizing C compiler. It doesn't really work yet though.

Ack. More stuff that you're not supposed to see. The code you see is for calibrating the dyno runs to get SI units (N*m, kg*m/s), but the code in dyno.py is really old. I have much, much better code sitting on a backup CD from my old workstation. It has an XML schema for a 'car', code to read/write these files, and code to take a loaded 'car' object and apply the conversion from pyxede-style power data to SI units. I'm fairly convinced that the code is correct. Everything is derived from first principles. I'm reluctant to release it as people are probably going to be unhappy at the numbers it gives them.

EDIT: here's a datalog that you can import. it's not even generated by pyxede even. it's off dustin's car. notice that it was done on a loading dyno, not the road, to the power curve is really funny looking. change the filename to .csv


d

 

You have any idea why i am getting a 0 reading on my timing column? you said it just shows offsets of stock, not actual degrees advanced or retarded?

 

Ahh, gotcha -- that makes sense, since the weight units were in kg

So now my question is, which module handles the calculations (of max power and max torque for example) after I import the data file you posted (smart2.csv)? Looks like both pyxede.py and wxdyno.py.

Also could someone explain the 'Power Calculation' frame when I'm importing smart2.csv -- Like a previous message said, I chose RPM=column1, A/F=column3, Knock=2, TimingShift=column4, and Timestamp=column0. This gave a Max Power ~ 1571 and MTorque ~ 1088. Is this just funny data, or have I done something wrong in the import process?? Funny data is fine for my purposes, but I wanted to make sure I'm using the tool right.

Thanks again!

 

The value you see under "timing" is "Tuned timing out". It's value ranges from -128 to 128. In other words, it's the ouput crank angle sensor.

EDIT: I'm not sure how you could get all zeros unless your timing table is empty. Are you running a map with a zero'd timing table?

d

 

This is all handled in DynoList->UpdateListandPlot in wxdyno.py. Max power/torque is determined by a simple linear search.

Ok, that import option doesn't do what you think it does. I was forunate enough, when I started coding that up, to realize that you could get your power/torque numbers a lot of ways. Some logs have timestamp. Some just rely on a given fixed sampling rate. Some already have torque in power in them. The options that you see just allow you to specify what data you have available. Torque/power is always calculated the same way.

torque = time2-time1 / rpm2-rpm1 = dTime / dRPM

Those numbers are fine, except they aren't in any unit that most people would recognize. If you didn't notice, all the torque/power calibration code is completely disabled. The torque values you see are in the unit RPM/second or more correctly 1/seconds -- useful for tuning, useless for bragging.

d