Tuning VVT

[JAM]

Special thanks to James for working on this trigger wheel application (currently using a test code)
I am new to the VVT tuning so I have many questions hoping some of the experts on here can help me out...

After determining range of motion (with some input from James) I set min/max at 77/127 and then went into closed loop.

my initial thought was to set the cam table with all values at 10 deg and see how stable and close to target i could get, and then step it up gradually

Started with P at 50 and I/D zeroed out... guess that was not right the right thing to do- cam angle it stayed at 0 until it had values in I/D, this seems contrary to what the manual suggests? so i had some low values to start

on the first log i attached you can tell the angle never really quite got to the target, on the second log i changed from 48hz to 99Hz and then started moving P until it was closer to target but my P value was high ( close to 200) . looks like the vvt duty is more erratic, is this because of the increase of frequency?

Here is my list of questions:
1.)how stable i can expect the cam angle to be compared to what it is in the log?
2.)is there any way to see the PID values in a datalog?
3.) when moving P, it seemed the higher the number the closer i got to the fixed target angle (actual was hanging lower than target). does this make sense-is that the correct direction?
4.) How do i determine the best frequency to use, what behavior am i looking for to optimize?
5.) any general VVT tuning tips would be great to hear

2014-08-14_19_1modified.msl

2014-08-14_20_modified(2).msl

test GM7X 3 CurrentTune.msq

[Matt Cramer]

This is showing less than 1 degree of error throughout the log for the most part - you may be able to get a little less with some fine tuning, but it's overall working quite well.

Not sure if there's a good way to see current PID values in the data log.

Yes, adding extra P terms can help reach a target faster but may give more oscillation.

Frequency can be a bit of a matter of trial and error.

[JAM]

This is showing less than 1 degree of error throughout the log for the most part - you may be able to get a little less with some fine tuning, but it's overall working quite well.

great, thats what i wanted to know

Yes, adding extra P terms can help reach a target faster but may give more oscillation.

Frequency can be a bit of a matter of trial and error.

Thanks, what is the process you use/what will i be looking for?

I understand the factory was using 128Hz, the first log was using 48Hz because this is where my idle valve valve was tuned, then bumped them up to 99HZ, (second log of this attached) seemed like more oscillation but may have been due to increased P

one more question; should i have expected response with I & D terms zeroed? its like it was turned off no matter what P value was in it.