• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 30 subscribers
  • Views 322 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

DRV8932PEVM: Unable to drive a low frequency 24V solenoid valve with the IC. Valve out of sync /constantly engaged in Cycle-by-Cycle control

Stefano Alfredo La Spina
Prodigy 240 points
Part Number: DRV8932PEVM
Other Parts Discussed in Thread: DRV8932, DRV3946-Q1

Tool/software:

Hi forum.

I need your help since I need to drive a low frequency (1Hz up to 150Hz) 24V - 500mA solenoid valve with the DRV8932 but I am having problem driving it with cycle-by-cycle method.

An MCU is providing a 2Hz digital signal with 40% duty cycle (200ms HIGH / 300ms LOW) applied to IN1.

Wanted to try the method described by datasheet page 16. Where: "Similarly, current through loads connected to ground can be controlled by controlling the INx pin pulse width -
INx = 1 builds up the current, and INx = 0 decays the current."

But following problem occurs:

the valve gets engaged during first cycle and then does not switch off.

Valve has an output LED on the top that shows exact timing in sync with the MCU output but the valve dues not follow the MCU.

the connector has a freewheeling diode inside.

Playing with VREF I have seen that tied to VDD the valve gets first cycle ON/OFF (out of sync) and then remains engaged.

  Please help

  • 0
    Prodigy 240 points

    What I cannot understad is howto pass from  fixed off-time current chopping to cycle-by-cycle mode.

    One last other thing is that I have a preproduction IC as mentioned in this post: https://e2e.ti.com/f/1/t/1054194

    I have seen that pins 20/21 are exposed to header and tagged as DECAY1/BDECAY/EN1 and DECAY0/ADECAY/MODE but on the datasheet it is mentioned as NOT CONNECTED and nothing is found what they effectively do since they are indeed connected to the MSP430 MCU onboard.

    thank You so much.

    Steve

  • +1 in reply to Stefano Alfredo La Spina
    TI__Mastermind 45195 points

    Hi Steve,

    Thank you for your post in this forum.

    Stefano Alfredo La Spina said:
    I need your help since I need to drive a low frequency (1Hz up to 150Hz) 24V - 500mA solenoid valve with the DRV8932 but I am having problem driving it with cycle-by-cycle method.
    Stefano Alfredo La Spina said:

    An MCU is providing a 2Hz digital signal with 40% duty cycle (200ms HIGH / 300ms LOW) applied to IN1.

    Wanted to try the method described by datasheet page 16. Where: "Similarly, current through loads connected to ground can be controlled by controlling the INx pin pulse width -
    INx = 1 builds up the current, and INx = 0 decays the current."

    When the load is connected to GND from OUTx the internal current regulation based on ITRIP will not work - requires load from OUTx to VM. So VREF has no influence in this configuration.

    The average current via the load (solenoid) would be based on the duty-cycle of the PWM drive. Varying the duty would vary the average current. This would be an open loop current control.

    Stefano Alfredo La Spina said:

    But following problem occurs:

    the valve gets engaged during first cycle and then does not switch off.

    Wouldn't this be expected behavior? Unless the current via the solenoid falls below the holding current level the solenoid will remain on. Can we get the current waveform of the coil current using a current probe for review? 

    In order to deactivate the solenoid the off time must be longer than current decay time required to decay the current to < hold current, this would depend on the L/R time constant of the solenoid. 

    What is the application expectation? For example with 2 Hz PWM input is the solenoid expected to turn on and off with PWM on and off? To achieve this you should design solenoid drive with full-bridge such that the current can be forced to decay fast. I hope I understood your application requirement correctly. Please clarify if it is different. 

    Stefano Alfredo La Spina said:

    One last other thing is that I have a preproduction IC as mentioned in this post: https://e2e.ti.com/f/1/t/1054194

    I have seen that pins 20/21 are exposed to header and tagged as DECAY1/BDECAY/EN1 and DECAY0/ADECAY/MODE but on the datasheet it is mentioned as NOT CONNECTED and nothing is found what they effectively do since they are indeed connected to the MSP430 MCU onboard.

    That specific pre-production device in the EVM has the same functionality as the production device. Regarding those pins 20/21, please ignore. This EVM layout was used for several other device variants in the family hence the connections to the MSP430. Thank you.

    Regards, Murugavel 

  • 0 in reply to Murugavel4637
    Prodigy 240 points

    Hi Murugavel, first of all thanks for your exhaustive reply. It took some time for me to make some measurements and come to a conclusion, so sorry for the late reply.

    Yes, the valve is one type of simple PUSH-PULL solenoid valve that has to be engaged and then once released there has to be no delay for the disengagement: it is used in industrial environment to precisely dispense pressurized aereosol so any delay after the turn off would give inaccurate doses; it has to precisely follow the PWM input: this valve that is a 24V AC- 10W valve (with led and embedded freewheel diode in the connector - https://shop.murrelektronik.com/en/Connection-Technology/Valve-Connectors/With-open-ended-wires/MSUD-valve-plug-A-18mm-with-cable-7000-18021-6360500.html)

    I was wrongly thinking that using in an Half-Bridge connection I would get some benefits (8 drivers in 1 chip - lower temperature by providing less hold current, some sort of diagnose - very important - valve is shorted or oc) , but since in a 1:1 PWM-IN PWM-OUT scenario, I should work with higher PWM frequencies since in the original mode as stated in the TI's article: " Recirculating load current in opposing MOSFET may not be enough to decay the current fast enough, which can cause a delay in the plunger releasing".

    Do your think that using a more simple high or low side mosfet is the better solution? does TI's portfolio have some driver that also has diagnose capability?

    Non Working valves like Broken or Clogged valves detection would be an extra value.

    Best regards and thank you again for clearing my doubts

  • 0 in reply to Stefano Alfredo La Spina
    TI__Mastermind 45195 points

    Hi Steve,

    You are welcome! 

    Stefano Alfredo La Spina said:
    Do your think that using a more simple high or low side mosfet is the better solution?

    With an HS or LS FET you would face the same issue of not able to decay the current fast enough to collapse the magnetic field instantly. 

    Stefano Alfredo La Spina said:
    does TI's portfolio have some driver that also has diagnose capability?

    We have the DRV3946-Q1,  a dual channel automotive solenoid driver with Current Regulation, Clamping and Diagnostics features. Thank you.

    Regards, Murugavel