• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 28 subscribers
  • Views 2212 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Operating a set of Proportional Solenoid Valves

Caleb Nothnagle
Prodigy 30 points
Other Parts Discussed in Thread: INA138, DRV777, DRV8860, LM3410, DRV8803

Hello! 

I have an application requiring me to individually control up to 5 miniature proportional valves with the output of a microcontroller (0-5V). Preferably this will be done by up to 5 PWM channels or even by serial communication if there exists an IC capable of managing this many valves at a time while only using one or two serial buses. Resolution and response time are both factors to consider as well though I'm not entirely sure of the exact specs required. In any case, my PWM only has 256 possible duty cycle values and I'm fairly certain the mechanical response of the valve will always be much slower than the control signal's ability to change (i.e. <10ms)

The real challenge I'm seeing is that many of the suggested drivers for solenoid control (at least from what I've seen on these forums sofar) lack current control to be able to properly drive these valves proportionally with feedback.  Or it may be that I'm not understanding their capabilities correctly?
I've read discussions on the LM3410 and INA138 as possible solutions; however, these ICs would control only one valve at a time and I was hoping for something more comprehensive. Another driver I was interested in was the DRV8860 which could control up to 8 devices but lacks current control as best as I can tell. Other ICs seem to provide a degree of current control but not with a very decent resolution. Lastly, the DRV777 is another driver mentioned which could fit the bill though the max current output per channel is low and its input to reach this max is only ~2.5V, effectively cutting my resolution in half. 

While I did mention a number of devices here, I am open to any other IC that could work.  Just figured I'd cover where I've already been. 

For the record, the valves I'm interested in using take a supply voltage of approximately 13.5V and require a current of 125mA to achieve full flow. 
The specific valve is attached along with a picture detailing the suggested analog circuit. VSO Data Sheet_1_19_11.pdf

Thank you for reading,

-Caleb 

  • 0
    TI__Mastermind 26065 points
    Hi Caleb

    The value can be driven with both voltage or current. As you said, current control may better to achieve repeatable flow control. But since the value has relative fixed resistance and inductance, also not very small, the pre-calibratied PWM duty (voltage) control should be able to work fine as well.

    So far I think DRV8860 could still be a good option. We can check the internal 8-level PWM duty whether can be directly used as a noncontinually flow level control. Or we can make external PWM control through SPI to get any level duty cycle (Need to test the MAX PWM can be used through SPI, the SPI clk of DRV8860 is MAX 250k).

    Best regards,
  • 0 in reply to Wilson Zuo
    Prodigy 30 points

    Hello Mr. Zuo

    On page 23 of the DRV8860 datasheet it seems to imply that only a few different PWM values may be set within the device.  I assume these are the 8 levels you mentioned.  Could you point me towards the page explaining where the SPI input may be used to provide a continuous current output? I'm also wondering if such an input would affect only a single output channel or all 8 as I am needing to have independent control over each valve.  

    Regarding the current sensing/control,  these valves do heat up significantly with use so a system not taking this into account may lack repeatability.  With that being said, I've yet to measure their change in flow over time with a set voltage as opposed to a set current.  Sounds worth testing to see what I can get away with.

    Thanks! 

    -Caleb

  • 0 in reply to Caleb Nothnagle
    TI__Mastermind 26065 points
    Hi Caleb,

    Yes. The internal 8 level PWM is fixed. What I mean is to do a SPI fast ON/OFF control for each channel. However I just checked it cann't get to higher PWM if we want more duty cycle resolution. For example, 8 bit, 5k PWM with 10 level duty control will need the SPI clock as fast as 400kHz. So DRV8860 may be not able to do it even only for 10 level change. So sorry to bring you the confuse.

    How about using DRV8803 with parallel inputs PWM control? It is with four channels, and maybe we can add one more discrete channel. Also you are right the heat caused current drop when driven with voltage could be very significant.

    Best regards,
  • 0 in reply to Wilson Zuo
    Prodigy 30 points

    Hello Mr. Zuo, 

    My apologies on the slow reply. Certainly I could use two DRV8803 drivers to address 5 outputs, no problem there. 

    I'm more concerned as to what current will be output by the device and how it may use a feedback to toggle the output.  Perhaps there is a separate system which may be used to read the current and provide the necessary PWM? Additionally, I'm not seeing any equations regarding the PWM value and the corresponding current output.  Is the 1A per channel maximum simply used in proportion with the duty cycle? (I.E 50% = 500mA assuming your supply goes that high?)