how to simulate a proportional valve?

[morens]

does anyone know how to simulate a proportional valve feedback, i wish the information for i can probe a board that drive ones of this valves.

thanks for your help

 

[itsmoked]

morens; not much information to work with... What kind of signals?????? What kind of valve????

 

[morens]

its a hidraulic valve, for a molding injection machine. the board that i want to probe its rexroth mark. and the board uses many valves for injection, for closing clamp.

thanks again

 

[logbook]

If you are talking about SPICE simulation or a similar electronic circuit simulator then get a triangular waveform generator as one input to a comparator and use the error signal as the other input to the comparator. You will have to mess around with the signal offset and gain factors to get the proportional gain settings correct.

This simulation is non-trivial.

I am of course assuming that you are talking about proportional control of an on/off type of valve, where the duty cycle controls the proportional output. If you have a linear actuated valve then the simulation would simply consist of a few opamps, R's, C's and L's, to simulate the various system components.

 

[OperaHouse]

I assume the basic purpose is for testing a board. For that purpose, I would just use a resistor that was twice the DC resistance of the servo coil. Then I would take a 470 ohm resistor from each side of the load resistor and feed that to a 1 uF non polarized capacitor. A DVM could then measure the voltage across that capacitor.

 

[logbook]

Ok, I think I understand better what he means now.

http://www.boschrexroth.com/

I checked out the proportional hydraulic valves on the above site. The fluid flow is apparently monotonically dependant on the analog current into the solenoid coil. One beast in particular takes 30W (12V, 2.5A, 5 ohm coil). Testing the driver board would then consist of putting on a suitable load resistor and perhaps a moving coil meter to measure the output voltage. Obviously the board would have to be put into some sort of open-loop test condition.

 

[morens]

thanks for the comments logbook & operahouse

i dont know much about proportional hidraulics but i saw the drawings for the feedback wiring on a ENGEL manual for a mold injection machine and it seems to be a LVDT or a resolver, more than a variable resistor. i assume by this fact that the valvule hasnt a linear output.

thanks again

 

[OperaHouse]

Most newer designs will have a pulse width modulated output. Even older analog designs will have about 100Hz dither signal added to them to keep the valve from sticking. Doubling the coil DC resistance should approximate the coil AC resistance, might even consider 3X. The added resistors and capacitor will smooth out this signal so it doesn't drive a meter crazy. Tou could even attach a very small geared dc motor in place of the resistance.

On the other end ia a little more difficult. The LVDT may output a signal that will vary the phase of the signal. I you had a spare transducer, you could attach it to the motor via a cam or lever. You would have to allow for continuous rotation. An electronic simularor would likely be beyond explanation here.

 

[morens]

thanks for your advices.

and for the rexroth page

 

[ScottyUK]

Morens,

You might also check out Moog's page. They're well-known in the precision servo valve business.

http://www.moog.com/

----------------------------------

I don't suffer from insanity. I enjoy it...