reactone

My manual boost controller spikes really bad so I was brainstorming different ways to control boost and came up with this: (click to enlarge)

A boost signal feeds into an adjustable pressure switch. When the set boost is reached, the switch closes the ground on a normally open solenoid valve, thus feeding the wastegate with full boost pressure signal (I have a 3.6psi spring in there now) which would open the wastegate. When the pressure falls below the set boost, the pressure switch opens,turning off the signal to the wastegate. When boost reaches the set pressure again,the whole process repeats. The only thing I can see that may be a slight problem is that the solenoid would probably flutter back and forth as the pressure is modulated between just above the set boost pressure and just below the boost pressure. A heavy duty solenoid may be in order here.
those who understand this diagram, what are your thoughts??

reactone

No one??

Hitchhikkr

Cant find a problem with it right now. I say go for it, shouldnt cost much at all.

BTW have you ever seen a Saab APC? Google it. JD hooked me up with one but I havent tryed it yet. Pretty straight forward concept, and for those of us without knock sensors an excellent safety net. Ive seen em go for cheap on ebay too.

reactone

Cool, I will Google that ******. My brother is a Saab mechanic, maybe he can score me one on the cheap. I may be able to get my hands on an SMC pressure switch, with a digital readout, So I can set boost pressure at will. I will test this out and let everyone know how it works.

SDRAWKCAB

the valve?

Hitchhikkr

Jus use an old school bosch cycling valve. (like the apc uses)

0b00st0

Not gonna work very well. The boost will fluctuate really bad and inconsistently. Even if the system was really fast at feedback it would still fluctuate. Also the solenoid will fail in no time since mechanical solenoids were never designed for rapid switching. This is why valves are PWM controlled.


The problem is that this control system is all feedback based, and as such will not be able provide consistent response or reasonable steady state error. I have gone over this before in past threads.

I don't know the exact properties of the valve, solenoid, or switch so I can come up with a decent transform, but I can already see that the step response is going to be very bad with massive overshoot and take a long time to come to steady state (which at that point will be a very sharp transitioning wave). i say steady state, but that system will most likely not be able to get there in most gears, maybe 4th or 5th though (ie, time).

reactone

Well thats what I thought the problem would be. It would really never reach a steady state, but rather shoot up and down (over/undershoot). I didn't think it could be that easy..

RotaryGeek

nopee. but dont lisen to me. Im drunk and Doingg BUKS thing on hmt rite now. Lol terrorising hmt ftW

Hitchhikkr

Didnt think about that. I see what your saying though. 10-4.