|
Title: *edited* Post by: mcgas001 on August 20, 2012, 02:44:20 PM *edited*
Title: Re: Help with boost control Post by: nyet on August 20, 2012, 03:52:59 PM Can you post logs of WGDC, req boost, actual boost, and, if possible, the various P, I, D results?
Title: *edited* Post by: mcgas001 on August 21, 2012, 02:38:40 AM *edited*
Title: *edited* Post by: mcgas001 on August 21, 2012, 05:01:18 AM *edited*
Title: *edited* Post by: mcgas001 on August 21, 2012, 10:00:30 AM *edited*
Title: Re: Help with boost control Post by: nyet on August 21, 2012, 10:10:00 AM Add a lot more D, pull a lot more I
that cracking pressure is pretty high; you may need to adjust WG linearization (not sure it exists in ME1.x) and give it more of a very concave bowl shape (drop off very steep from 100%) Title: Re: Help with boost control Post by: mcgas001 on August 21, 2012, 10:14:14 AM *edited*
Title: Re: Help with boost control Post by: nyet on August 21, 2012, 10:21:03 AM Err, sorry i mean I limit (KFLDIMX).. you shouldn't have 0 I, imo.
If your KFLDRL is all zeros, you have completely disabled the PID. It is equivalent to removing the N75 (constant WGDC of 0), which means should only see max boot equal to cracking pressure. I'd say you have a torn wastegate line or broken wastegates. Title: Re: Help with boost control Post by: mcgas001 on August 21, 2012, 10:27:37 AM *edited*
Title: Re: Help with boost control Post by: mcgas001 on August 21, 2012, 11:53:44 AM *edited*
Title: Re: Help with boost control Post by: nyet on August 21, 2012, 11:57:21 AM Honestly, you should'nt have to touch P or I. stock is usually just fine.
generally, all you should ever have to adjust is I llimit (decrease), D (increase), and WG linearization (bring it down some in mid range, but make SURE IT STAYS MONOTONICALLY INCREASING!) A log would be good btw. Please use ECUxPlot if you can. If you can't i can try to make it work with your logger. Title: Re: Help with boost control Post by: mcgas001 on August 21, 2012, 12:24:45 PM *edited*
Title: Re: Help with boost control Post by: nyet on August 21, 2012, 01:07:45 PM I've made KFLDRL and KFLDIMX to make them look like bowls. They start at 95%, drop to 40% in the middle, and then raise back upto 70% by the end. YIKES! No, by bowl i meant concave. You should ALWAYS have your wgdc linearization monotonically increasing. At no point should a higher input dc (for a given point) result in a lower output dc If you aren't sure what that means, step back from your car and read up on control systems :) Title: Re: Help with boost control Post by: Lobuziak on August 21, 2012, 01:10:18 PM I'd love to get a log mate, but seen as Its a vauxhall all I can log with is OPCOM or VauxCom. There is a position for requested and actual boost and WGDC in op-com logs and You can calculate approximated engine load(http://s4wiki.com/wiki/Load) from afm voltage reading converted with MLHFM curve corrected with KFKHFM. What's Your LDRXN? Shouldn't actuator spring be opening wastegate at requested or lower value than needed boost to control it properly... What's Your Afm reading? If boost gets higher the throttle body should close to compensate to get requested load... That's my understanding... Lobuziak Title: Re: Help with boost control Post by: mcgas001 on August 21, 2012, 01:28:18 PM *edited*
Title: Re: Help with boost control Post by: Lobuziak on August 22, 2012, 02:29:33 AM Here are the graphs - actual pressure is going above map sensor's 2,5Bar limit at 4350 rpm and then WGDC is going up and in the same time the ECU is closing throttle plate (max value 5) to limit the mass of air... second graph is using some calculated values. I assumed that afm, injectors and fpr where from Z20Leh(VXR) and KFKHFM adds 4% in high load region. Injectors Duty Cycle is reaching 100% at 5500 rpm and above.
Regards Lobuziak Title: Re: Help with boost control Post by: lulu2003 on August 22, 2012, 02:45:46 AM still building boost with 10% WGDC looks very strange to me. So no need to check PID at this stage.
everything checked mechanically? Title: Re: Help with boost control Post by: prj on August 22, 2012, 07:24:44 AM Your injectors are also maxed out, careful or you will blow this up.
Title: Re: Help with boost control Post by: nyet on August 22, 2012, 09:14:23 AM Like I said before, wastegates or the lines are busted. You shouldn't be making boost past cracking pressure with 10% dc
Title: Re: Help with boost control Post by: mcgas001 on August 22, 2012, 11:48:12 AM *edited*
|