[pheno]

My intention would be to get creating/editing custom jobs easier (all done in the GUI) so that no manual editing is needed. I added 'start job' button added thinking that it would be nice to quickly test new parameters but currently it's missing the important thing - possibility to edit the parameters..

It seems to take time to build the logic that forms the final job argument. Column used for argument varies a little between DDE's. There's also MW_SELECT_LESEN_NORM that uses different separator compared to STATUS_MESSWERT_LESEN, etc.

If you use Megalogviewer to visualize graphs .csv files please select 'Time from start' or 'actual time' from Time format in the logger setup windows. Then it'll show the time correctly.

[pheno]

Originally Posted by TDIwyse

Yup. Tried it multiple times, same result. Windows 8 OS comes up and says the program is not responding ... and then the program times out and has to close.

Interesting that all the available "jobs" from the Custom #1-3 work fine, and every other jobs I've looked at also work.

This is the first issue I've had with the program. Otherwise it's been FANTASTIC! And I'm inclined to think it's something odd with my setup ...

For taking traces from the program itself: Unzip files into testo folder and run 'testo-trace.bat'. It will create testo.log - reproduce the problem and send the log to me. Thanks.

[DWR]

BTW, Pheno, you have one of my favorite automobiles. We like torque here and that was the first BMW I ever drove that reallly pulled from low rpms. Don't get me wrong, an M3 is nice, but for different reasons. And the older 535 had a better engine for turbocharging.

[pheno]

Originally Posted by DWR

BTW, Pheno, you have one of my favorite automobiles. We like torque here and that was the first BMW I ever drove that reallly pulled from low rpms. Don't get me wrong, an M3 is nice, but for different reasons. And the older 535 had a better engine for turbocharging.

Yeah, I love the car. I've had it almost for 9 years! My daily driver. Still gives me nice wide grin when pressing pedal to the metal :-) Well, I haven't actually driven faster car than mine so my opinion might be a little subjective...

Pretty much stock, except for Morimoto Mini H1 projectors and E34 M5 3.6 exhaust (includes middle silencer). It's nice and quiet on highways on low rpm but you can easily hear it's V8 and starting from 3500-4000k sounds are awesome. Odometer shows "only" 346k km at the moment

[TDIwyse]

Originally Posted by DWR

TDIwsye,
Try putting the attached file in your Jobs > Custom Jobs folder. You should get an "RPM, Air & Fuel" along with the other custom values. Remember to remove the .txt from the file name. Let me know if that works.

Got a good pull in after work using the boost and "drive pressure" you helped me get going. This is with the external wastegate again ... Colder temps and a big high pressure system moving through (lots of North wind ... did the pull into the wind).

Interesting to me that the baseline turbo pressures are a lot higher today than I've seen in the previous logs. The analog gauge showed the same peak boost behavior (<32 psi peak) but the software was logging significantly higher numbers.

You can definitely see the transition between the small/large turbo across the rpm range. A HUGE shift in drive pressure to boost ratio hand-off happens. These ratio's are so good I'm thinking of opening up the boost controller for the EWG a little...

[DWR]

Well, the first noticeable thing is the sampling rate is much improved! As you note, the transitions are better defined. What that graph says to me is the HP turbo should be coming offline sooner. No sense in blowing out drive pressure with the EWG, when it can be used to bring the LP turbo up to speed sooner.

Also, you should know I have found more "turbine pressure" measurements. Many have different scaling factors. Not absolutely convinced we are getting what we think we are getting with regard to psi accuracy. As usual, more answers create more questions ...

[pheno]

Looks good but maybe I was expecting a little bit better resolution/sample rate. If you look at the graph window bottom right what does it show for job's speed, should be something around 50-60ms I guess ? Maybe ecu does not refresh sensor data for each job call ? Need to take a look at this...

Here's an old log from my M60B40 DME 3.3.8 to compare




Here's log from MS44, DME 5.2, Uses LC-1 and LMA-3 which TestO can also record in the same .csv file... external sources are easy to add in TestO if you know the communication protocol and there serial interface on the external logger like 14.7 or Innovate LC-1




Log from DDE 5.0 E46 330D, 4th gear:



Log from DDE 6.0. E60 530D D60M57A0

[pheno]

It would be interesting to see the real sample rate that ecu gives us.. i.e. how 'smooth' is the graph if you just run one job, for example boost pressure ?

[TDIwyse]

Originally Posted by pheno

Looks good but maybe I was expecting a little bit better resolution/sample rate. If you look at the graph window bottom right what does it show for job's speed, should be something around 50-60ms I guess ? Maybe ecu does not refresh sensor data for each job call ? Need to take a look at this...

I had exported the data and brought it into excel to plot. Unfortunately that laptop is at home and I'm at work ... I'll look at the sampling rate when I get home and report back. I do remember it showing more time step resolution than the data was able to keep up with (there would be multiple time steps with the exact same "value" for the data).

I'm actually very pleased with the data sampling resolution for logging ~5-7 parameters at a time. Much better than BMWhat and I'm able to see more things than the Bav Tech or Torque tools can ...

Hopefully this weekend I'll try to follow your suggestion on the crash issue with the custom jobs and see if i can find anything useful.

Thanks again.

[TDIwyse]

Originally Posted by pheno

It would be interesting to see the real sample rate that ecu gives us.. i.e. how 'smooth' is the graph if you just run one job, for example boost pressure ?

Post 42 and 93 were example resolutions of logging 3 parameters at a time. Post 42 didn't use a Custom Job, post 93 did. Post 37 had ~16 things being sampled and a mix of Custom Jobs and not ... It was pretty coarse in the time steps. I'll look at what the sampling time was for these examples when I get home.

[DWR]

Originally Posted by pheno

It would be interesting to see the real sample rate that ecu gives us.. i.e. how 'smooth' is the graph if you just run one job, for example boost pressure ?

With 4 parameters in a custom job, I was getting 70ms per sample set. I know you shared ideas on maximizing the sample rate with me and I need to follow up. If we run multiple custom jobs and don't sync them, do they tend to run just like a collection of standard jobs, i.e. slower?

[DWR]

Originally Posted by pheno

Looks good but maybe I was expecting a little bit better resolution/sample rate.

You should also be aware TDIwyse has over 400Hp. That means he goes through 4th gear faster than you might expect!

[pheno]

Originally Posted by DWR

With 4 parameters in a custom job, I was getting 70ms per sample set. I know you shared ideas on maximizing the sample rate with me and I need to follow up. If we run multiple custom jobs and don't sync them, do they tend to run just like a collection of standard jobs, i.e. slower?

Correct. To maximize sampling rate start only one custom job with all needed parameters. Means that each additional running job will waste time.

[pheno]

Originally Posted by TDIwyse

Post 42 and 93 were example resolutions of logging 3 parameters at a time. Post 42 didn't use a Custom Job, post 93 did. Post 37 had ~16 things being sampled and a mix of Custom Jobs and not ... It was pretty coarse in the time steps. I'll look at what the sampling time was for these examples when I get home.

Ok, thanks for info.

Well, technically speaking custom job is a job (standard job if we call it that ) but it includes custom parameters. And actually it looks like standard jobs like 'boost pressure - STATUS_LADEDRUCK_IST' utilizes parameters like you do now but contains only one parameter per job.

[TDIwyse]

Originally Posted by TDIwyse

Post 42 and 93 were example resolutions of logging 3 parameters at a time. Post 42 didn't use a Custom Job, post 93 did. Post 37 had ~16 things being sampled and a mix of Custom Jobs and not ... It was pretty coarse in the time steps. I'll look at what the sampling time was for these examples when I get home.

My memory was off a bit. Instead of 3, there were 4 parameters being logged. And instead of 16, there were 14...

Attached some of the first lines of the 4 an 14 parameter logs to show the time steps (msec's) and the column names and values...

The DDE is out of the vehicle now so I won't be able to do anything else for awhile.

[DWR]

Originally Posted by TDIwyse

The DDE is out of the vehicle now so I won't be able to do anything else for awhile.

Hey, you can't just leave us wondering! What's happening to the DDE?

[TDIwyse]

Might be related to pursuing an Arachnid Galosh ...

[pheno]

Originally Posted by TDIwyse

My memory was off a bit. Instead of 3, there were 4 parameters being logged. And instead of 16, there were 14...

Attached some of the first lines of the 4 an 14 parameter logs to show the time steps (msec's) and the column names and values...

The DDE is out of the vehicle now so I won't be able to do anything else for awhile.

Thanks. You can see in the log that one job takes about 50ms and other 250ms. Latter must be the one with 14 parameters. There's also job mw_select_lesen_norm2 that works the same way taking a list parameters. It wants to have parameters from LABEL column separated with space char. Could be worth to try to see difference with status_messwertblock_lesen..

[GreekboyD]

DDE out of car? Looks like TDIwyse is up to something.

[2deerwhistlers]

Originally Posted by TDIwyse

Might be related to pursuing an Arachnid Galosh ...

Very tricky

[DWR]

Originally Posted by GreekboyD

DDE out of car? Looks like TDIwyse is up to something.

He's caught in Jarek's web!

[iaknown]

Maybe he's sending it back to Redline?