|
carsey
|
 |
« Reply #60 on: July 25, 2014, 12:57:00 PM »
|
|
|
Done another change like I said and the adaption gone back up :S yet ive lowered things.
Following specified spot on now though, right up until ~6250rpm where its overboosting and ECU kills throttle off the get it back in check.
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #61 on: July 25, 2014, 01:52:03 PM »
|
|
|
I'd flatten out I-Max from 6000 on up (make it smooth) and have req boost follow the results.
or, alternately, let i-max continue up to 95% and have req boost follow the results...
I really think your imax is moving around too much (way too low at start), you really want to keep it more or less flat (or a simple shape), or the PID will do ugly things...
it has a bunch of odd knees, which you really want to avoid.
get a smooth i-max, and see where your boost ends up. Set req boost accordingly.
A smooth wgdc should end up with a nice smooth boost profile. If you are struggling to get boost to follow req by making imax do weird things, odds are your req boost doesn't make sense.
|
|
|
|
« Last Edit: July 25, 2014, 01:54:47 PM by nyet »
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #62 on: July 25, 2014, 02:06:24 PM »
|
|
|
Thing is, ive never set WGDC to 95% top end, top limit was 79% in IMX and a linear LDRL. Guess the PID was adding extra on because it was seeing the underboost from a desired boost too high.
Reason I started the IMAX low was because the max load was flatlined, so was controlling boost that way through duty cycle. I guess now I have pulled it down to where it should be through desired boost, I can raise the early stages of IMAX.
Im just not 100% sure how I should alter IMAX to get to where I want it. I understand you want 99% duty cycle keeping the wastegate closed on spool, so maybe I could add that until 3000rpm on my setup. Then pull it down to 60-70% duty through the midrange tapering back up top end where k03s is struggling to keep going.
Something a bit like ive attached below. Im not 100% sure tbh.
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #63 on: July 25, 2014, 02:57:34 PM »
|
|
|
TI understand you want 99% duty cycle keeping the wastegate closed on spool, so maybe I could add that until 3000rpm on my setup.
Nope, you don't need Imax to be 99% during spool. During spool the P keeps DC at 95%, unless your LDRXN is too low. Again you do NOT want your i max moving around. You have no way of knowing where to drop imax from 95 to whatever, since actual boost might meet req boost *at any time* depending on when you hit the gas, etc. |
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #64 on: July 25, 2014, 03:07:43 PM »
|
|
|
Ah, I understood that wrong then. You dont happen to have a example of what you mean by any chance do you?  |
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #65 on: July 30, 2014, 05:45:11 AM »
|
|
|
Any ideas on what you mean by keeping my imax flat?
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #66 on: July 30, 2014, 09:37:35 AM »
|
|
|
I mean don't have it woggling around all over the place. Keep it simple.
|
|
|
|
|
Logged
|
|
|
|
|
phila_dot
|
|
« Reply #67 on: July 30, 2014, 12:17:57 PM »
|
|
|
I-max should be the DC required to make the input plsolr for the majority of the rev range.
Keep it smooth or even flat, as in, if you plot the points up the RPM axis it should be a line.
LDDIMXN gives head room and KFLDRL takes care of any non-linearity.
|
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #68 on: July 30, 2014, 12:29:32 PM »
|
|
|
So raise the early stages of it upwards so its more (in line) with the higher duty up top. So instead of 43% up to 3000rpm, make it 60% for example. Am i right in saying that?
|
|
|
|
|
Logged
|
|
|
|
|
phila_dot
|
|
« Reply #69 on: July 30, 2014, 01:55:42 PM »
|
|
|
Are you asking a out KFLDIMX?
You want I to accumulate up to the duty cycle required in steady state, so yes go straight there.
|
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #70 on: July 30, 2014, 02:01:11 PM »
|
|
|
IMX yeah. Will raise the early parts of it and hope it brings things back into check.
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #71 on: July 30, 2014, 02:32:06 PM »
|
|
|
Lets put it this way. Look at the stock IMX. It doesn't do weird things.
|
|
|
|
|
Logged
|
|
|
|
|
carsey
|
|
« Reply #72 on: July 31, 2014, 06:28:03 AM »
|
|
|
had a quick look at the standard one for a k04 (albeit im running a k03s turbo) and can see through the midrange how its a bit more smoother with a nice flowing curve to it.
Without changing individual load sites too much, ive come up with this, Ive ignored the 95% on spool as my current 45% was enough to give me the boost pressure i want... So ive gave that a limit of 65% which it should never really hit if the PID controls spool elsewhere.
Probably still looking at it wrong but surely if the IMX values are higher than the actual duty needed to hit the desired boost, ECU should use however much duty is required to hit that and make it smooth itself?
|
|
|
|
|
Logged
|
|
|
|
|
phila_dot
|
|
« Reply #73 on: July 31, 2014, 07:38:43 AM »
|
|
|
That won't work very well.
I will over accumulate.
It should be the exact DC required to make plsolr_w.
|
|
|
|
|
Logged
|
|
|
|
|
ddillenger
|
|
« Reply #74 on: July 31, 2014, 09:02:34 AM »
|
|
|
Am I the only one that swaps the axis on DIMX?
|
|
|
|
|
Logged
|
Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience!
Email/Google chat:
DDillenger84(at)gmail(dot)com
Email>PM
|
|
|
|
