|
nyet
|
 |
« Reply #15 on: June 10, 2015, 11:41:09 AM »
|
|
|
its med9, I don't have a way to log those other variables.
that's why I'm just trying here and there.
could you tell me under what circumstances would one adjust Q0 or Q1??
There are no shortcuts here. You have to understand PIDs. |
|
|
|
|
Logged
|
|
|
|
|
majorahole
|
|
« Reply #16 on: June 10, 2015, 11:48:03 AM »
|
|
|
maybe that's a better way for me to understand, u get it? like working backwards in math when having the answer to begin with, I find it easier to figure out how to get there
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #17 on: June 10, 2015, 11:49:43 AM »
|
|
|
Again, without logs, there is no context.
We don't need all the pid variables, just as much as you have for now.
|
|
|
|
|
Logged
|
|
|
|
|
majorahole
|
|
« Reply #18 on: June 11, 2015, 07:24:46 AM »
|
|
|
here's the last log i took with Q2 lowered
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #19 on: June 11, 2015, 10:03:01 AM »
|
|
|
That looks fine.
|
|
|
|
|
Logged
|
|
|
|
|
majorahole
|
|
« Reply #20 on: June 11, 2015, 10:09:26 AM »
|
|
|
I never hit requested till later on? I know its not bad, just wanted to get to my requested. I suppose it might not be possible
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #21 on: June 11, 2015, 10:19:32 AM »
|
|
|
more IMX then, but eventually you'll overshoot.
I doubt Q2 is the issue; your IMX could still be too low.
|
|
|
|
|
Logged
|
|
|
|
|
majorahole
|
|
« Reply #22 on: June 11, 2015, 11:25:51 AM »
|
|
|
i've gone all the way to 95, but then in higher gears in that range, if i go WOT i overshoot like crazy
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #23 on: June 11, 2015, 11:28:15 AM »
|
|
|
Then you'll have to be careful and try to balance Q2 and IMX.
It isn't easy to do.
IMX should be more or less the steady state dc needed for the boost you want.
Q2 is there to prevent overshoot.
|
|
|
|
|
Logged
|
|
|
|
|
majorahole
|
|
« Reply #24 on: June 11, 2015, 11:33:14 AM »
|
|
|
but when I go up in IMX it still doesn't add any WG DC... I'm not sure if I've gone to 95 all the way up to 3000rpm tho, now that I think of it.
so you think I should try IMX at 95 up to 3000rpm just to see if that helps, and then tame the high gear overshoot with Q2? just to start? then maybe come down with IMX to have less Q2?
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #25 on: June 11, 2015, 11:38:08 AM »
|
|
|
but when I go up in IMX it still doesn't add any WG DC... I'm not sure if I've gone to 95 all the way up to 3000rpm tho, now that I think of it.
so you think I should try IMX at 95 up to 3000rpm just to see if that helps, and then tame the high gear overshoot with Q2? just to start? then maybe come down with IMX to have less Q2?
Absolutely not. IMX has nothing to do with the "95" thing. IMX has to do with where the DC settles after Q2 approaches 0. You are really not understanding how the PID works  Again, it is a complex topic and unless you have at least a little bit of theoretical background, the idea that you are going to figure out PIDs empirically is a pipe dream. |
|
|
|
|
Logged
|
|
|
|
Beaviz
Full Member
Karma: +8/-4
 Offline
Posts: 190
|
|
« Reply #26 on: June 11, 2015, 01:09:27 PM »
|
|
|
The issue might also be that the stock axis on KFLDIMX on K03 equipped MED9 cars only goes to 1000mbar. Change the axis and the values accordingly and you will have much better control when you reach the requested boost.
|
|
|
|
|
Logged
|
|
|
|
|
majorahole
|
|
« Reply #27 on: June 11, 2015, 01:31:25 PM »
|
|
|
the only info I could find for IMX was "i-limiter" and "steady state WG duty cycle" in my logs and testing neither of those definitions were accurate.
so IMX is WG DC after error corrections are back to 0?
|
|
|
|
|
Logged
|
|
|
|
|
majorahole
|
|
« Reply #28 on: June 11, 2015, 01:32:28 PM »
|
|
|
The issue might also be that the stock axis on KFLDIMX on K03 equipped MED9 cars only goes to 1000mbar. Change the axis and the values accordingly and you will have much better control when you reach the requested boost.
thank you, I will try and raise that axis a bit |
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #29 on: June 11, 2015, 06:24:14 PM »
|
|
|
the only info I could find for IMX was "i-limiter" and "steady state WG duty cycle" in my logs and testing neither of those definitions were accurate.
That sentence means nothing. so IMX is WG DC after error corrections are back to 0?
Also pure gibberish. IMX is the predominating factor during steady state after i windup on spool once the D and P contributions go to zero. Obviously, it goes through LDRL first. |
|
|
|
« Last Edit: June 11, 2015, 06:25:52 PM by nyet »
|
Logged
|
|
|
|
|
