how-to: ECU-based direct boost control
#197
After setting wgdc properly, like you have, then you fine tune the target-psi and error correction (for psi-target), or fine tune the BDEL and error correction (for load-target)
The problem I have seen in setting a flat WGDC, and using error correction to "fill in" the boost is that A: it is time based and B: it's very hard to keep a stable curve from just using error correction, espically with big added WGDC swings.
If you get your WGDC as close as possible to the curve you want and use error correction to just fill in / take out small amounts, you will have the most stable boost control. At least from my findings. I've tuned over 100 evos with ecu-based boost so far.
One trick to make the boost very stable in most conditions is changing the load it follows to the timing/fuel load which is part of changing to psi-target boost, but it does even more wonders if you are still using load-target boost control.
All this won't help much if you don't have your WGDC dialed in as close as possible to the desired boost curve you want.
Last edited by razorlab; Feb 1, 2008 at 04:38 PM.
#198
You have a flat WGDC curve of 50%. To reach your desired boost at 4000 the WGDC is 50% and your desired boost at 7000 is 70% WGDC. You go full throttle to 7000 and shift. When you hit the throttle again at around 5500, your WGDC is still 70%. This basically means you'll spike pretty much every time you shift.
Razor, I do know what you mean about getting a stable boost curve by setting a nice BWGDC curve. However, the WGDC curve to meet my desired load levels changes on almost a daily basis. If I dial in the BWGDC dead on at 10:00 one night, the next day at noon it's not right anymore.
In a prefect world, the curve itself wouldn't really change .. it would simply move up and down the percentage scale in an equal amount across the rev band to correct for temp changes. This simply doesn't happen for what I've seen.
So, I'm not trying to make the flat BWGDC profile method work better, or even as well as a nicely tuned BWGDC curve with correction. I just want a system that will work year round with stable, predictable results.
#199
Thread Starter
EvoM Guru
iTrader: (50)
Joined: Mar 2006
Posts: 9,675
Likes: 130
From: Tri-Cities, WA // Portland, OR
Its taken some time to get it dialed, bit it was never unpredictable. I keep wondering if my use of a bleeder rather than a modified restrictor pill is part of the reason. It would be great if one of the more experienced tuners here would do a comparison between a bleeder (with no restrictor pill at the BCS) and a modified restrictor pill at the tee. ;-)
Last edited by mrfred; Feb 1, 2008 at 05:58 PM.
#200
#201
Razor, I do know what you mean about getting a stable boost curve by setting a nice BWGDC curve. However, the WGDC curve to meet my desired load levels changes on almost a daily basis. If I dial in the BWGDC dead on at 10:00 one night, the next day at noon it's not right anymore.
Mine is stable. Full error correction on.
If you run just off WGDC without error correction, the boost will change depending on conditions. This is what the error correction is for.
Here are some boost plot overlays from three logs of mine, different days, different conditions:
Last edited by razorlab; Feb 1, 2008 at 08:00 PM.
#202
I keep wondering if my use of a bleeder rather than a modified restrictor pill is part of the reason. It would be great if one of the more experienced tuners here would do a comparison between a bleeder (with no restrictor pill at the BCS) and a modified restrictor pill at the tee. ;-)
#203
Just some creative criticism, but if you set your boost targets unrealistically high such as 20.4 psi from 500-2500 and the WG duty cycle maxed out in those same regions, you are prone to overshooting boost and wavy boost curves.
#204
Are you using the error correction with stock boost load or are you changing the boost load to follow the normal 2byte load?
Mine is stable. Full error correction on.
If you run just off WGDC without error correction, the boost will change depending on conditions. This is what the error correction is for.
Here are some boost plot overlays from three logs of mine, different days, different conditions:
Mine is stable. Full error correction on.
If you run just off WGDC without error correction, the boost will change depending on conditions. This is what the error correction is for.
Here are some boost plot overlays from three logs of mine, different days, different conditions:
For instance .. when I tuned my car in the summer I had a great BWGDC curve tuned and it worked great. Temps changed by about 30* after a few months and my actual WGDC curve changed more than 20% IIRC. When the values are that far off, TBEC is basically controlling the boost anyway, so why even bother with the BWGDC curve?
Tuning with a flat BWGDC could take a bit of the complication and time out of tuning the stock solenoid.
#205
a) The boost target does not matter because boost error does not start working untill it sees -2.3 psi. So I can put 50 psi or 0 psi in 500 rpm, it doesn't make any difference. And if you try to engage boost correction during spool-up in those early rpm areas - i think that would lead to wavy boost.
b) Maxed out WGDC is used for fast spool-up.
I have mentioned before - it would be PERFECT if someone could write a patch to always keep WGDC at 100% until you reach target boost. That way spool up would be the fastest. But noone has done it so far...
#206
I have run the same thing and gotten "stable" boost ... I just don't think you're following me 100% here. I know that TBEC takes up the slack when the temps change. I'm just basically saying why spend hours dialing in the BWGDC curve when the TBEC table is just going to correct anyway? If the TBEC table (essentially) alone can regulate boost, why not just let it?
For instance .. when I tuned my car in the summer I had a great BWGDC curve tuned and it worked great. Temps changed by about 30* after a few months and my actual WGDC curve changed more than 20% IIRC. When the values are that far off, TBEC is basically controlling the boost anyway, so why even bother with the BWGDC curve?
Tuning with a flat BWGDC could take a bit of the complication and time out of tuning the stock solenoid.
For instance .. when I tuned my car in the summer I had a great BWGDC curve tuned and it worked great. Temps changed by about 30* after a few months and my actual WGDC curve changed more than 20% IIRC. When the values are that far off, TBEC is basically controlling the boost anyway, so why even bother with the BWGDC curve?
Tuning with a flat BWGDC could take a bit of the complication and time out of tuning the stock solenoid.
#207
I have run the same thing and gotten "stable" boost ... I just don't think you're following me 100% here. I know that TBEC takes up the slack when the temps change. I'm just basically saying why spend hours dialing in the BWGDC curve when the TBEC table is just going to correct anyway? If the TBEC table (essentially) alone can regulate boost, why not just let it?
For instance .. when I tuned my car in the summer I had a great BWGDC curve tuned and it worked great. Temps changed by about 30* after a few months and my actual WGDC curve changed more than 20% IIRC. When the values are that far off, TBEC is basically controlling the boost anyway, so why even bother with the BWGDC curve?
Tuning with a flat BWGDC could take a bit of the complication and time out of tuning the stock solenoid.
For instance .. when I tuned my car in the summer I had a great BWGDC curve tuned and it worked great. Temps changed by about 30* after a few months and my actual WGDC curve changed more than 20% IIRC. When the values are that far off, TBEC is basically controlling the boost anyway, so why even bother with the BWGDC curve?
Tuning with a flat BWGDC could take a bit of the complication and time out of tuning the stock solenoid.
Have you switched your boost load to 2byte load?
Also, what unstable things are you seeing on your side? Too much boost or too little?
#208
Here's my line of thinking:
b) Maxed out WGDC is used for fast spool-up.
I have mentioned before - it would be PERFECT if someone could write a patch to always keep WGDC at 100% until you reach target boost. That way spool up would be the fastest. But noone has done it so far...
b) Maxed out WGDC is used for fast spool-up.
I have mentioned before - it would be PERFECT if someone could write a patch to always keep WGDC at 100% until you reach target boost. That way spool up would be the fastest. But noone has done it so far...
Really WGDC should be renamed BCS duty cycle.
MB
#209
I have tried both ways.
If wgdc was set at 100 from 0 to 2500 rpm or so & target boost set at 21 in the same regions it would overshoot by about 4-5 psi.
If I ramp into it smoothly with wgdc & target psi then boost is very stable with no overshooting.
#210
Alrighty, I understand what you are saying but the less work the error correction needs to do, the more stable the boost will be, in my experience. WGDC shouldn't take hours to dial in if you know what you are doing (I'm not saying you do not), nor is the stock solenoid or aftermarket solenoid complicated to tune if you know what you are doing, which, to me, it sounds like you do.
Have you switched your boost load to 2byte load?
Also, what unstable things are you seeing on your side? Too much boost or too little?
Have you switched your boost load to 2byte load?
Also, what unstable things are you seeing on your side? Too much boost or too little?
But, my argument is basically this ... you dial in the BWGDC in June and you have a max of +/- 2% TBEC correction ... in January with that same BWGDC curve, you have a correction of -20%. In all honesty ... the fine tuned BWGDC method is not really better than the flat BWGDC method unless you re-tune it every time the temps change by more than 15* or so.
So, my reasoning is that tuning a car with the flat BWGDC at a median value and a good TBEC table should technically be more stable on average year round. However, I've yet to prove that a mostly TBEC controlled boost setup can be stable at all ...
I kind of came to this conclusion months ago when I first started tuning the stock solenoid based on mrfred's findings. It initially took me about a week to really dial in a good BWGDC curve. This was because I was logging for short periods of time (a few pulls a night) and making adjustments. The next night the BWGDC would be off because of the slight differences in baro/temp, so I'd basically have to start over every night.
At that point I decided it would be best to set a nice median curve for the BWGDC and let TBEC take over. This was before mrfred found that the boost load was different from the 2-byte value. When mrfred posted his findings, I immediately switched to the 2-byte load value for boost control. I had everything running correctly and stable with a BWGDC about 8% below actual to accommodate for temperature changes. I assumed that the WGDC swing in colder temps wouldn't be too drastic and assumed that 8% or 10% would be the maximum swing I'd see.
When temps fell in the fall I started paying attention to the WGDC values ... they got as far down as 15% or so below my BWGDC (>20% below actual WGDC in the summer) and I re-tuned the BWGDC curve again. This time I set it pretty much dead on and reintroduced the TBEC assuming that this would be my new median curve.
Since then, I've re-tuned the BWGDC many times and always to a lower WGDC value since the swing between seasons is HUGE.