Supercharger bypass valve operation 5

[jcd06]

Roots superchargers usually have a bypass valve that is open at idle.

I want to start playing with the electric bypass valve that came with my Mercedes kompressor
(mentioned here: thread71-205859 )
and I am wondering which strategy to choose. I can make it TPS or MAP dependent.

I was wondering when it should close.
Should it close gradually from above a certain level of inlet pressure on, or does it shuts close over a fairly short stroke of throttle, located somewhere above the cruise position?

 

[lghikas]

The project is still alive ?

Has anyone tried the MAP-TPS circuit ?

Thanks

Leonidas

 

[markburlington]

Hi Leonidas,

I built the TPS circuit with the same supercharger and valve as Jean (jcd06).
However, the output oscillated on/off at low frequency with the high frequency PWM signal only present during the on phase.
It did control the valve fairly accurately with feedback but at low frequency and ouside the design specification of the valve (>100Hz)
It was also not PWM, simply switching on when the valve moved out of position due to the return spring, and off when it reached the position required.

Possibly I have incorrectly built the circuit or there is a fault.

However, I opted for another approach. I build a driver circuit controlled with a microcontroller, an Arduino Duemilanove which has PWM outputs and I have written a control program using PID feedback. Eventually, the valve will be controlled by the Engine Management Computer (Megasquirt 3)that also has PWM outputs. I have used the microcontroller to learn the PID parameters characteristic of the valve.

 

[bentwings1]

I passingly followed this thread from way back. Pat hit on the roots blower and either FI or carbs on the intake side of the blower.

My streetrod had a 355 cu in SBC with a 6-71 and 2 750 cfm carbs on top of it. No bypas or any of that stuff. This car is as close to a daily driver as you can get. 510 shaft hp at 5500 rpm. I already have well over 13k miles on it in all wether except snow. It's been in the cold down to 10 deg F and heat to 105 deg F the air cond and heat work perfect too. I get 16+ mpg on the highway and 14 cruising around town.

I get 8 pounds of boost from 2000 rpm up and no detonation (8.4 comp ratio). I can run on 87 octane but usually get 91-93. Vacuum runs 12 inches at idle and can get up to 19 coasting. Normal cruise vac is 10-12.

All the parts for the blower are off the shelf items selected on my experience in the hotrod world.

I also had a SC T-Bird which I drove for over 200k with no problems with the motor. The intercooler allowed a little higher compression and the boost was more gradual up to about 2000rpm. From there it came on pretty strong up to 13-14 psi. Ford did a good job on this motor except for head gaskets. Many people blew head gaskets when the blower speed was increased. The fix was to o-ring the block or the heads. The motro was just a little too small and the car just a bit too heavy to be a real super car. It still got 24 mpg most of the time.

There are lots of questions if you are putting throttle control after the supercharger. Fuel distribution and air flow are just 2.

99 Dodge CTD dually.

 

[lghikas]

Is there any way to put a mechanical valve, controlled by vacum ?

Leonidas

 

[jcd06]

Dear all,

Yes, this project is still alive.
It is somewhat on hold because of a pending house refurbishment but in my head it goes on

There is no means to adjust this valve mechanically.
You can see the valve on a picture posted Januari 11 2010.

Mark, it worked for me. It should be possible to have it working over there too but if you have the skills to do it with a microcontroller I wouldn't bother about the opamp stuff. Taking into account the specific valve characteristics is one thing but you still have to tell your system what output you want for a certain set of input parameters. How did you see that?

It is still possible to replace this electric valve by a vacuum operated one but that was not the goal.
For the setup with the new charger I will make sure to leave enough clearance so a vacuum actuator can be fitted.

Train dogs,
Teach people.

 

[markburlington]

Hi Jean,
glad the project is still alive, even if it is only in your head.
I am still playing around with tuning the PID feeback parameters.
(Lots of other stuff going on here too)
I will have to define a look-up table for the boost output. this will be within the ECU program. (Megasquirt III).
I recon TP and MAP as the input parameters, with an overriding max boost limiter (built-in ECU funtion)
I would love to put data-logging on a supercharged Merc engine, to discover how MB do it.
PID feedback control tuning seems to be a black-art but I have a set of parameters that seem to work. Next is the look-up table.
Regards,
Mark

 

[jcd06]

Hi Mark,

Glad to hear about the progress of your project.
PID parameters can be calculated but often this gives useful results only when there is only few external parameters of external influence, which oviously is not the case here.
Time for some trial-and-error engineering and probably being satisfied with some compromises.

If you're wondering how MB did it I can tell you something.
Lately, -finally at last- I had the opportunity to drive a C230 with a M271 engine with exactly the same kompressor as ours.
The game is not to control the bypass valve only, but to involve the electronic throttle as well.
This way they succeeded in avoiding any surge peaks or unwanted engine reaction.

There was no boost gauge in the car so I couldn't see the pressure build-up but the power comes in very smooth, absolutely no sudden kick-in. The same at throttle release; there is clearly a dashpot effect noticeable in the ETC response.
The compromise is that throttle response is not at all what you would expect from a Bowden cable

Regards,

Jean



Train dogs,
Teach people.