Custom WGDC tables
- Thread starter Sava666
- Start date
Leave P an I stock, modify D factor (lower by 50-55%) to reduce osciliation.
Leave WGDC Base stock and extened MAF range to 350-373
Increase the MAF range on WGDC Ceiling Adder and set the control range to 40-50 (depending on how much control you want automated)
Modify the last two cells my increasing it 10-15% (factor 1.1 or 1.15 multiplicator)
Use MAF Airflow Adder to finetune the underboost areas
Leave WGDC Base stock and extened MAF range to 350-373
Increase the MAF range on WGDC Ceiling Adder and set the control range to 40-50 (depending on how much control you want automated)
Modify the last two cells my increasing it 10-15% (factor 1.1 or 1.15 multiplicator)
Use MAF Airflow Adder to finetune the underboost areas
Thanks but these are MHD+ Custom P/D tables and they are zeroed out. Not Classic control.
If you are using the MHD+ custom setting not just boost ceiling you can scale it in similar fashion as regular controls. Adjust one table at the time. You have oscilation so relax D
Here's a simple graphic showing in general how the P, I, and D parameters influence the output of the process you're trying to control (in this case boost). Watch how the process output curve changes as the P, I, and D parameters are ramped up individually. In general:
-P gets you on target faster
-I reduces offset between target and actual
-D reduces overshoot and ringing (oscillation) in the process output
You can influence oscillation to some extent with all of them depending where you're at, for example, if I is too high, then reducing I can reduce ringing. Same with P. But in general, more D is used to reduce oscillation.
-P gets you on target faster
-I reduces offset between target and actual
-D reduces overshoot and ringing (oscillation) in the process output
You can influence oscillation to some extent with all of them depending where you're at, for example, if I is too high, then reducing I can reduce ringing. Same with P. But in general, more D is used to reduce oscillation.
Attachments
Yeah I know. But have you ever tried to fill in those tables? Have you ever seen them? They are zeroed out. Can you share your tables here?
I spent a lot of time tuning PID on an industrial electronic pressure controller at work many years ago, but not on my car yet - still on an OTS tune until I have more free time to play around. But PID is an extremely common industrial process control methodology, so the principals should be the same.
Were they zerod out in this graph? From this response, I'd add a little I until your boost is oscillating closer to target vs. below it. The oscillation would get worse initially, but just ignore that for now. Once your boost is on target (but still ringing) go start adding D until the oscillation tames down. You're already getting on target really quickly here and the initial overshoot isn't too bad so I'd probably leave P alone for now.
Were they zerod out in this graph? From this response, I'd add a little I until your boost is oscillating closer to target vs. below it. The oscillation would get worse initially, but just ignore that for now. Once your boost is on target (but still ringing) go start adding D until the oscillation tames down. You're already getting on target really quickly here and the initial overshoot isn't too bad so I'd probably leave P alone for now.
Attachments
No doubts on your knowledge mate ) first you need to see them tables in mhd custom wgdc section, then you'll start realizing what I'm talking about. Or you hit me on telegram and we can discuss it later when I start tuning? Fancy?
Not on telegram, but I'd be really interested to see what your tables look like on mhd+ if you don't mind sharing them? I'm n55, and the PID tables have PID parameters split up into a ton of different buckets. I want to get a bigger turbo and try self tuning but honestly that PID is a little intimidating the way it's set up! Gonna be super time consuming...
Ok, I'll DM you tomorrow morning ok? Details to be provided ) left my laptop at work. Cheers
Did you ever progress with this? I was starting to set mine up as well and wasnt' sure how granular you made the breakpoints for boost target
trying to dial in the custom wgdc tables myself.
I got to the point where I zeroed every PID table out and just focused on tuning the base table as best I could. There were certain spikes in boost that couldn't be smoothened out in the base table. I then added some PID control, but no idea where to start with the values I somewhat copied the stock tables. These are for VTT stg2 turbos however so I imagine the stock pid values aren't great. I lowered the D factor, then increased it. Lowered the P factor. I see there is no I factor custom table so I assume it uses the stock one still?
From my experience more D less I can help but its not working here and of course I tried to introduce PID slowly, not looking good.
I got to the point where I zeroed every PID table out and just focused on tuning the base table as best I could. There were certain spikes in boost that couldn't be smoothened out in the base table. I then added some PID control, but no idea where to start with the values I somewhat copied the stock tables. These are for VTT stg2 turbos however so I imagine the stock pid values aren't great. I lowered the D factor, then increased it. Lowered the P factor. I see there is no I factor custom table so I assume it uses the stock one still?
From my experience more D less I can help but its not working here and of course I tried to introduce PID slowly, not looking good.
yeah, i started with a smooth wgdc base and got there after trying to fill in the areas where boost was dropping or spiking.
anyways here it is again with 'smooth' wgdc table. im using adder airflow table now.
last time ive tuned an n54 was pre MHD+ stuff, but i dont remember having such a hard time dialing in a smooth boost curve. I dont want to blame the solenoids though, but the car did come in with crazy boost oscilations to begin with.
how can i embed the graph directly into the post?
yeah, i started with a smooth wgdc base and got there after trying to fill in the areas where boost was dropping or spiking.
anyways here it is again with 'smooth' wgdc table. im using adder airflow table now.
last time ive tuned an n54 was pre MHD+ stuff, but i dont remember having such a hard time dialing in a smooth boost curve. I dont want to blame the solenoids though, but the car did come in with crazy boost oscilations to begin with.
how can i embed the graph directly into the post?
If the car is on twins, just use the normal wgdc tables with custom boost target toggled on. The custom wgdc table imo is more reserved for ST. It's much easier imo to use the normal wgdc tables on twins.trying to dial in the custom wgdc tables myself.
I got to the point where I zeroed every PID table out and just focused on tuning the base table as best I could. There were certain spikes in boost that couldn't be smoothened out in the base table. I then added some PID control, but no idea where to start with the values I somewhat copied the stock tables. These are for VTT stg2 turbos however so I imagine the stock pid values aren't great. I lowered the D factor, then increased it. Lowered the P factor. I see there is no I factor custom table so I assume it uses the stock one still?
From my experience more D less I can help but its not working here and of course I tried to introduce PID slowly, not looking good.
