Stock Boost Control - Closed loop PWM or just Boost On / Off

[cbatters]

I found some excerpts/graphs from the article Jeff is talking about posted at John Monnin's site.

http://johnmonnin.netfirms.com/Boostcontrol.html

MBC has the overboost spike that I expected to see whereas the Profec B has classic critically damped control response.

(I still need to study the entire article to find out exactly how the units were tested/tuned.)





Clint

[TurboBob]

the biggest problem, and why so many aftermarket controllers do such a bad job, is that all the response characteristics of the turbo+engine change so radically with small changes in environmental conditions and/or driving style.

So hard to get them stable and still responsive under all conditions......

B

[cbatters]

Originally Posted by TurboBob

the biggest problem, and why so many aftermarket controllers do such a bad job, is that all the response characteristics of the turbo+engine change so radically with small changes in environmental conditions and/or driving style.

So hard to get them stable and still responsive under all conditions......

B

Still waiting to see the text of the turbo review that jeff has to see the conditions of the test and the procedure they used to tune the controllers.

Fuzzy logic should be able to determine actuator response time and maximum DP/DT to come up with reasonable PD control settings. Non-linear control could keep tabs on DP/DT and any boost overhshoot and modify the PD control settings.

I'll tell you in a few weeks how well my controller simulation works in bang/bang mode (simulating MBC) vesus full PD mode.


Clint

[SHiVeR-91VR4]

clint:

Shoot me an email, I might have some good information for you about this project.

[MythosVR4]

Originally Posted by cbatters

OVERVIEW

1. Stock boost soenoid is powered by the MFI circuit which provides +12V to black lead of the solenoid as soon as the ignition is turned on

2. Boost solenoid is energized when the ECU pulls the white lead of the solenoid to ground.

3. Boost control signal is a constant 16 HZ with varying duty cycle

4. Duty cycle appears to be a function of RPM regardless of load / boost

RPM DUTY CYCLE COMMNENT
2000 100% Constantly energized
3000 70%
4000 60%
5000 55%
6000 50%

What does this mean? When i have power "energized" going through my solenoid you can blow air through it. and when it isn't powered that is when it stops air. Also through the connector it only sends power through the solenoid when you mash the gas. if you slowly increase the rpms the voltage stays off. Is my solenoid wired wrong or what.

[cbatters]

Originally Posted by MythosVR4

What does this mean? When i have power "energized" going through my solenoid you can blow air through it. and when it isn't powered that is when it stops air. Also through the connector it only sends power through the solenoid when you mash the gas. if you slowly increase the rpms the voltage stays off. Is my solenoid wired wrong or what.

One side of the boost solenoid is connected to +12V and the other side is pulled down to ground by the ECU to turn on the solenoid. When the solenoid is inactive (closed) you should see +12V on the ECU boost solenoid wire.

On my car, the solenoid duty cycle varied based on RPM with/without a load.

I will doublecheck the polarity of the duty cycle I was measuring and re-post.



Clint

[FaceGarden]

Hey guys, I was working on my own EBC last year, but since I've taught myself everything about how to do this stuff, I had some problems. Its really quite hard to get an even 5 volts (at ~1A) from the car's wacky voltage when I've never had an analog circuits class (i've never had any circuits class actually). Anyway, there is a link in my sig to the old page for the controller I made. I got it working on the bench but never got the car's power supply working properly. I have been inspired to keep working on it though, as it was basically finished except for the power supply. I had it setup for a very basic boost control method (open solenoid below max pressure, close it above max pressure) just to see how high I could boost with the stock solenoid, but I never got to test it. I also spoke with a guy on 3si who was making his own EBC, but I haven't heard anything about that lately. I would like to do this though, we should all have EBC's, and they should NOT cost $400!
So, one thing I had trouble with was finding cheap solenoids. I think we could all agree that using an inline solenoid instead of a bleeder solenoid is better, but that would add significant cost unless anyone has found a good source (and believe me, I spent the better part of a few weeks searching and calling every place I could find all across the country, the best I could find was around $60 I think.
Anyway, I am going to keep working on my controller. Lets see what we can all do with this!
-Taylor Alexander

[cbatters]

Originally Posted by FaceGarden

Hey guys, I was working on my own EBC last year, but since I've taught myself everything about how to do this stuff, I had some problems. Its really quite hard to get an even 5 volts (at ~1A) from the car's wacky voltage when I've never had an analog circuits class (i've never had any circuits class actually). Anyway, there is a link in my sig to the old page for the controller I made. I got it working on the bench but never got the car's power supply working properly. I have been inspired to keep working on it though, as it was basically finished except for the power supply. I had it setup for a very basic boost control method (open solenoid below max pressure, close it above max pressure) just to see how high I could boost with the stock solenoid, but I never got to test it. I also spoke with a guy on 3si who was making his own EBC, but I haven't heard anything about that lately. I would like to do this though, we should all have EBC's, and they should NOT cost $400!
So, one thing I had trouble with was finding cheap solenoids. I think we could all agree that using an inline solenoid instead of a bleeder solenoid is better, but that would add significant cost unless anyone has found a good source (and believe me, I spent the better part of a few weeks searching and calling every place I could find all across the country, the best I could find was around $60 I think.
Anyway, I am going to keep working on my controller. Lets see what we can all do with this!
-Taylor Alexander

Couple thoguhts:

1. My goal has been to use the existing solenoid to lower costs and simplify installation. No argument that the bleeder config wastes some pressure but in the testing I have done so far, I don't see any control response time advantage of a bleeder versus "inline" configuration.

(I am also not sure how much spool time is adversely affected by the bleeder. At 4000 RPM, 9B's spool VERY fast and are actually pretty tricky to control.)

2. Initial bang-bang control testing was not positive. Actuator response time results in significant overboost. Advantage of EBC is that you can use feed-forward and derivative control to minimize overshoot of setpoint. With bang/bang control mode it is just a fancy MBC with a digital setpoint.


Clint

[FaceGarden]

Originally Posted by cbatters

Couple thoguhts:

1. My goal has been to use the existing solenoid to lower costs and simplify installation. No argument that the bleeder config wastes some pressure but in the testing I have done so far, I don't see any control response time advantage of a bleeder versus "inline" configuration.

(I am also not sure how much spool time is adversely affected by the bleeder. At 4000 RPM, 9B's spool VERY fast and are actually pretty tricky to control.)

2. Initial bang-bang control testing was not positive. Actuator response time results in significant overboost. Advantage of EBC is that you can use feed-forward and derivative control to minimize overshoot of setpoint. With bang/bang control mode it is just a fancy MBC with a digital setpoint.


Clint

Yeah, I intended to use the stock MBC in my final design, but there were some people who felt that an inline valve was better, so I still looked at what that would take. I wanted to test the stock solenoid before I decided otherwise, but I never got the thing finished enough to test.

As far as control goes, I only indended to use the bang/bang method to test the thing. The big problem was that I still haven't had any controls classes, so the derivative control was going to be a bit difficult, although I think I would have been able to do it when the time came. Anyway, I am going to revive my EBC, and see what help I can give here.
-Taylor Alexander