MegaLogViewer HD has a new feature

[hrdtail78]

Gotta ask. Since the AFR trace shows a lean condition, why not just look at that and richen up the area the AFR graph shows as lean?

What area in the AFR graph would you make adjustment?

Why do we need all the other graphs? AFR is all I need to see lean spots that need correcting.
Same with knock. Start a bit of retard right before the event and retest.
Those basic traces are all that's needed IMO to identify and correct tuning issues.
What am I missing in all the other graphs?

You need a bit more than that.  With the last scatter graphs Andy posted.  The area being pointed to in the right graph is actually the rev limiter and decel fuel cutoff.   MLV does allow you to filter out this data.

A 3D scatter graph is a good way of looking at a lot of recorded data at once.  With repeated use with different tuning equipment on different combinations.  Trend will start to show, or things will start to look out of place.

   

[whittlebeast]

Hrdtail

You can see the same spot in the tune right here.

http://www.nbs-stl.com/HarleyTuning/MAPxRPM%20Issue%20Trace.png

The dyno operator may have been very close to the rev limiter when he lifted, but both injectors were still running.  I think we are just finding a simple bust in the tune.  And after all that is exactly what we would be looking for at his point in the tuning process.

Adding a rev limiter filter could be

RevLimiter_FuelCut and define it as

[INJ PW F] = 0 or [INJ PW R] = 0

Note that doing it this way, it would also throw out any data when the trailing throttle fuel cut was active.  The names for Injector Pulse Width chances depending on the logger you are using.  TTS, TwinScan, Dynojet....

Have fun tuning

Andy

[whittlebeast]

Check out this trace now.  I put the VE Front and Rear to the same scale.  Notice the difference in the VE Front and Rear but the Duty Cycle Front and Rear are the same.  That sure looks like a code bust.  Or at least an undocumented quirk.

http://www.nbs-stl.com/HarleyTuning/MAPxRPM%20Issue%20Trace.png

I also added the AFR 13.2 line to show the target AFR that the tuner was shooting for in this trace.  Things were going good until he hit this spot in the tune.

Andy

[whittlebeast]

This is interesting.  Look how well the AFR was holding close to the target AFR except at idle.

http://www.nbs-stl.com/HarleyTuning/AFR%20vs%20Target%20AFR.png

Andy

[FLTRI]

What area in the AFR graph would you make adjustment?

Not sure but I would be looking at where the AFR goes lean for RPMs and 100 KPA MAP then follow the change up in rpms to where the AFR comes back to target.
Disclaimer: This is assuming the sensors are seeing good sampling. But the drop in power looks to be with the lean area.
Bob

[whittlebeast]

Bob, this will all get way more interesting once we get past the basic concepts and move on to looking at data logs with a 1/4 of a million records. This log is only 6100 records.

Plots start looking more like this one with 135,000 records.

http://www.nbs-stl.com/RxpxTuning/Race%20AFR%20Plot.png

Andy

[glens]

The dyno operator may have been very close to the rev limiter when he lifted, but both injectors were still running.  I think we are just finding a simple bust in the tune.  And after all that is exactly what we would be looking for at his point in the tuning process.

Or you may be missing some data that would perfectly explain what went on.  Messing with the calibration just because you saw that might not be prudent.  Best would be replicating it taking a chance you'll get a log showing you what went on.

[whittlebeast]

Or simply hang on a pc5, autotune and lcd 200 and be willing to look.  Rev limiters almost always show up at the 02s.

[whittlebeast]

As a little reminder, this is what banging hard off the rev limiter looks like.

http://www.nbs-stl.com/HarleyTuning/TTS%20Tune%2002.png

Andy

[lt401vette]

Since the update i have had nothing but trouble with the software

Efianalytics have stopped communicating with me in regards to the issue, they keep telling me i have some other software on my computer which is doing something to log file

Worked fine prior to update, since update problems

I use an apple computer, bought a cheap windows XP for winPV and MLVHD, they the only two softwares on pc apart from windows XP operating system

We haven't given up on you..

You can try an old version, it should install fine. It will prompt you to update, tell it no.
I have tried to open your files with 3.4.11 and 3.4 12 and they open the same way there.

We can clearly see why the logs are not opening properly, it seems the Carriage Return and Line feed in the logs you sent us that don't work are quite odd.
Each line is terminated by 2 Carriage Returns and 1 line feed. The normal line ending for PowerVision log files is 1 carriage return and 1 line feed.

We can add special handling to ignore the 2nd Carriage return and will do so if we can not come up with why it is there on your computer, but 1st as we have not seen this with any other users, so it seems to be something specific to your PC or some other software involved, I want to understand how the extra Carriage Return is getting in there.

Brian will stay in communication with you and we will make sure it does work for you even if we do need to simply filter the extra carriage returns.

[glens]

sed will handily fix that problem...

Sounds like a Mac is getting its hands on the files.

[Mirrmu]

thanks, okay re-installed MLVHD things still not right

I will disconnect PV cable from bike and just ride without it for a while, don't even know if there are datalog settings within the Powervision which is altering the .csv log files

will go through a few key on/off cycles, plug everything back in, ride and see what happens

[whittlebeast]

Is the PV running the latest code?  I am guessing this is the PV generating this issue.

Andy

[Mirrmu]

yes, all updated, only the tune database was an older version, all firmware etc is current

[hrdtail78]

https://app.box.com/s/2q763ixqm3a3c2v6gapsyp97nd68f6wa

Can you open and view the above?

[Mirrmu]

file open but takes quite a while to load and be fully operational, much the same as what my logs are

so would be good if MLVHD could have a look at that file of yours hardtail to see if has same faults as mine

thanks for posting it

[whittlebeast]

I am running an older 1.8 mhz, 3 gb windows 7 box.  If I load up MLVHD first, it takes about 5 sec to get the program to load.  Once loaded, I can open Hrdtails78's file in under a second.

Andy

[hrdtail78]

Hows this one.  it's a twin scan log.

https://app.box.com/s/ugyrzhy8u7s7xjw4ce7boc9cio4s1szd

[glens]

At this stage i give the MLVHD a big thumbs down, since the update i have had nothing but trouble with the software

Efianalytics have stopped communicating with me in regards to the issue, they keep telling me i have some other software on my computer which is doing something to log file

Worked fine prior to update, since update problems

I use an apple computer, bought a cheap windows XP for winPV and MLVHD, they the only two softwares on pc apart from windows XP operating system

I'd missed you saying you use an apple computer when I later said:

sed will handily fix that problem...

Sounds like a Mac is getting its hands on the files.

as a reply to:

We haven't given up on you..

You can try an old version, it should install fine. It will prompt you to update, tell it no.
I have tried to open your files with 3.4.11 and 3.4 12 and they open the same way there.

We can clearly see why the logs are not opening properly, it seems the Carriage Return and Line feed in the logs you sent us that don't work are quite odd.
Each line is terminated by 2 Carriage Returns and 1 line feed. The normal line ending for PowerVision log files is 1 carriage return and 1 line feed.

We can add special handling to ignore the 2nd Carriage return and will do so if we can not come up with why it is there on your computer, but 1st as we have not seen this with any other users, so it seems to be something specific to your PC or some other software involved, I want to understand how the extra Carriage Return is getting in there.

Brian will stay in communication with you and we will make sure it does work for you even if we do need to simply filter the extra carriage returns.

For a moment reflect back on the old dot-matrix printers we all used to have.  There are several "unprintable" characters used for control.  A LineFeed character sent to the printer will cause the platen to index the paper to the next line.  A CarriageReturn character sent to the printer will cause the print head to return to the start of the line.  In combination they prepare the print head to continue printing at the beginning of the next line.  Just like with that old mechanical typewriter where you hit the lever which then both fed the paper upwards one line and  back to the beginning of the line.

The same "mechanism" is used to display text on the (console) screen of a computer.  The cursor being directed down one line and/or to the beginning of the line.

Microsoft has always used the CR/LF combination as new-line control in their text files.  Unix uses just the LF character as line separators and Apple uses (or at least historically has used) just the CR character.

The different systems handle things differently in conjunction with the way they were designed.  In a strictly literal sense, the three methods would "print" like:

"This is a string of textCRLF-----------------------"

Code Select Expand

This is a string of text
------------------------

"This is a string of textLF-----------------------"

Code Select Expand

This is a string of text
                        -----------------------

and "This is a string of textCR-----------------------"
This is a string of text

I see mismatches of the different mechanisms all the time.  It often appears in the source of web pages where more than one system has been used to set up the way the pages are created.  Web browsers have been designed to "just do the right thing" in the face of various end-of-line conventions.

I suspect your Apple computer is somehow modifying the files by adding the extra end-of-line control characters.  Try collecting a log and not letting anything Apple touch the file and see what happens for you.

EFI Analytics should probably take a cue from web browsers, but different operating systems should also not assume anything about and make adjustments to the data they're handling.

I've played with MLV in days gone by, but to me it's not worth any cash outlay and once they went to nagging and limited functionality I said "goodbye".

[Mirrmu]

i use a windows XP pc for winPV (Powervision software) and MLVHD only, I bought this PC as I was using a mate's for the winPV, they the only 2 software on computer

it only gets switched on when I connect Powervision for tune and logs, I use Apple for all other computer needs

for a DIY'er the MLVHD is a great program, I had no need to buy Excel to view the logs in and for $40US its not expensive, there just something not right with my logs/software things at the moment

[whittlebeast]

Hows this one.  it's a twin scan log.

https://app.box.com/s/ugyrzhy8u7s7xjw4ce7boc9cio4s1szd

Here is the VE correction table if you are doing a MAP based tune.  Note that the MAP has been converted to KPA.

http://www.nbs-stl.com/HarleyTuning/Twin%20Scan%20S_D%20Correction%20Multipliers.PNG

If you were doing a TPS based tune, this would be the correction table.

http://www.nbs-stl.com/HarleyTuning/Twin%20Scan%20A_N%20Correction%20Multipliers.PNG

Have fun tuning

Andy

[whittlebeast]

Here is the AFRs trying to hit the AFR Target.

http://www.nbs-stl.com/HarleyTuning/Twin%20Scan%20Target%20AFR.PNG

[FLTRI]

Here is the AFRs trying to hit the AFR Target.

http://www.nbs-stl.com/HarleyTuning/Twin%20Scan%20Target%20AFR.PNG

Andy,
Got any idea why the t/p and the data looks nervous...except the WOT but there seems to be more than one WOT t/p. WTFS?
Since there is no scaling to refer to it's hard to quantify but that's some crappy looking data...unless the sampling is so slow it makes the data look like stair steps?
Bob

[whittlebeast]

In that log, it appears that the the data was collected at 4 per sec.  Slow but workable.  I have no idea what brand dyno or the capabilities of that dyno are.

I find data from the street or track far more telling as to what the motor really does.

Did you guys know that MLV can open several logs at a time?  This is great for combining several rides into one view.

Andy

[whittlebeast]

Bob

Add a few filters and it is fairly apparent that you have hit the capabilities limits of the dyno.

http://www.nbs-stl.com/HarleyTuning/Twin%20Scan%20RPM%20MAP%20Throttling.PNG

Andy