Shop OBEX P1 Docs P2 Docs Learn Events
Using PWM for valve control — Parallax Forums

Using PWM for valve control

RichardFRichardF Posts: 168
edited 2012-06-10 13:10 in General Discussion
Do you think PWM will work to throttle operate a normally closed solenoid valve? I have an application where I want to control valve position so as to regulate a stream of fluid.
Thanks,
Richard

Comments

  • stamptrolstamptrol Posts: 1,731
    edited 2012-06-09 14:02
    Richard,
    That's how purpose-built hydraulic and other fluid valves work. The big difference is they have been optimized for low inertia, minimum stiction, and fast response. Many are designed just like a speaker coil as opposed to a traditional winding.

    If you do some tests with your valve, you can determine how fast it can cycle. (for a standard valve, I'm guessing a small valve might be able to open and close 2 or 3 times a second) Depending on your ultimate use, that might be sufficient.

    Cheers,
  • RichardFRichardF Posts: 168
    edited 2012-06-09 14:44
    stamptrol wrote: »
    Richard,
    That's how purpose-built hydraulic and other fluid valves work. The big difference is they have been optimized for low inertia, minimum stiction, and fast response. Many are designed just like a speaker coil as opposed to a traditional winding.

    If you do some tests with your valve, you can determine how fast it can cycle. (for a standard valve, I'm guessing a small valve might be able to open and close 2 or 3 times a second) Depending on your ultimate use, that might be sufficient.

    Cheers,

    Thanks, Tom. I will give it a try. What would be a good transistor to use for a buffer between the valve and my Stamp? Valve has a 12 volt solenoid drawing 1/2 amp. Anything come to mind?
    Richard
  • kwinnkwinn Posts: 8,697
    edited 2012-06-09 16:43
    Just about any fet or bipolar transistor with a 1A & 20V or higher rating should work. I prefer to use a fet for the built in protection diode. I have used the SI2302DS and TJS4405NT1G for relay and gas solenoid coils that required more current than a ULN2803 or TPIC6595 could provide.
  • pjvpjv Posts: 1,903
    edited 2012-06-09 16:44
    Hi Richard;

    I believe you success will be quite dependent on the type of valve. You say "normally closed", and that could very well be like the fill valves on washing machines and diswashers, and those are "snap action" valves. I believe you will have difficulty in making those operate in any kind of proportional manner. Other valve designs are meant for proportional control, and othersyet, such as the slow acting zone ball valves for hydronic heating systems (typically 24 VAC) could be faked into operating somewhat linearly. As Tom said, do some experimenting with the intended valve, but I predict that the snap action types will cycle from full on to full of, and you might not want a pulsing kind of flow.

    Just sayin'.

    Cheers

    Peter (pjv)
  • $WMc%$WMc% Posts: 1,884
    edited 2012-06-09 18:28
    Keep the pulse cycles as low as possible.
    '
    The duty cycle of the solenoid is usually very low, It will wear out very fast.
  • Peter KG6LSEPeter KG6LSE Posts: 1,383
    edited 2012-06-09 21:56
    WMC is right

    ....

    People often assume PWM means some KHz speed wave , often for DC motors it is ..
    But in real .. There is no reason not to declare A Frame of 10 Sec and PWM off that as a number .

    it will be ROUGH but In the end you are still doing """ PWM"""""" ..

    I do the Same thing with my E Bike . I pulse the bike to get it going .. Human PWM ..

    to the OP.. Look at how fast your valve can switch .. . Consider Hoses after it as a Snubber to make the flow not so pulsy.

    I can see a 2 Sec Window as your PWM base to work with .

    Peter..
  • ElectricAyeElectricAye Posts: 4,561
    edited 2012-06-10 08:30
    $WMc% wrote: »
    Keep the pulse cycles as low as possible.
    '
    The duty cycle of the solenoid is usually very low, It will wear out very fast.

    I agree. You had better check the specs on whatever valve you are using. Many regular valves have limited cycle life and you can burn through that lifetime pretty fast if you're cycling several times a second.
  • RichardFRichardF Posts: 168
    edited 2012-06-10 09:43
    Based on what has been said I think going with a repeated signal of pulse length just long enough to fully open the valve, rather than PWM, is probably the best starting point. If I put a reservoir downstream to smooth out the pulses then, in the manner of a DAC output capacitor, I should see a smooth change in pressure at the nozzle, which is what I am looking for. I can ramp up the pulse rate until I no longer see an increase in pressure at the outlet nozzle. That should indicate that the valve is starting to just stay open and doesn't have time to close down before the next pulse. As far as life cycle of the solenoid valve, I will just have to see how it plays out. Thanks for your inputs and suggestions.
    Richard
    P.S. This is a "dancing waters" project.
  • kwinnkwinn Posts: 8,697
    edited 2012-06-10 13:10
    It might be possible to use a pressure tank from a well water system and a flow restrictor beyond that to regulate the flow rate. This would reduce the PWM frequency required and prolong the solenoid valve life. Whether this is possible or not would depend on the fluid composition and the application.

    Might be simpler to replace the solenoid with a proportional valve.
Sign In or Register to comment.