Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Flyback diode on solenoid slows solenoid retract time

  1. Apr 7, 2014 #1

    I'm an electrical engineer dipping my foot into solenoids as part of a project here at work.To monitor the position of the solenoid's armature I have a linear pot installed. My is issue revolves around a flyback diode installed across a DC solenoid I am testing. I am low side switching the solenoid with a cheap DPDT relay. Without the flyback diode installed I see massive negative voltage spikes across V(coil) as I would expect. With the diode installed, no large negative spike occurs. Problem solved right?? Wrong, with the flyback diode installed the retract time, specifically the time after turning off the relay but before I see movement of the armature, jumps from 25 ms without the diode to over 100 ms with the diode. My customer's application is very time sensitive, they won't be happy with a +100 ms delay before the armature begins to move when powering off the solenoid. I am also assuming that my customer will want to use a FET for switching the solenoid on and won't like a 100V surge when turning it off, making the diode and something else to speed up the response time a part of their final solution. Any suggestions would be greatly appreciated! Sorry for the PDF, work copier defaults into PDFs.

    Attached Files:

  2. jcsd
  3. Apr 7, 2014 #2
    Early in my career, I worked on a project for mail sorting and found the same thing. You will need to use a resistor in series with the switching path and a cap in parallel with the resistor. The cap should be of sufficient value to allow the current change to quickly snap the the solenoid in the power on state then ramp the current down to the resistor value limited amount. The resistor limited amount should be the holding current which should be much less than the initial closing current. The diode still needs to be in place, but the smaller current should reduce the holding time by a significant amount. You will need to determine the values based on your current, voltage and coil resistance.
  4. Apr 7, 2014 #3


    User Avatar
    Gold Member

  5. Apr 7, 2014 #4

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    energy to be got rid of is LI2/2

    rate of energy dissipation is VI

    in a resistor V and I are in proportion
    in a zener they are not

    so a zener should give a constant di/dt until the armature starts to move
    that might give you a more predictable time to open, if you need it.

    Is spring tension adjustable, or maybe airgap in operated position? Those are what we tweaked back in days of telephone timing relays(early 60's).
  6. Apr 7, 2014 #5
    I miss Pease Porridge. I got really excited when I saw the title "What's all this ....", so I was proud that I even came close to Bob's wisdom. Bob Pease was a very smart engineer. Emach you would do well to follow or at least understand his designs (first web link in meBigGuy's post).
  7. Apr 8, 2014 #6
    First off, thanks everyone for the replies! I attached screenshots of the data plots illustrating my issue. Now down to business after reading the replies some additional information that would help flesh out the problem further:

    1) You guys already hit on this but I will state it here for the record, we are using a spring return on the solenoid to return it to its initial de-energized state.
    2) I am not kidding that our customer will want to PWM the solenoid when energized in order to save power. Our application is automotive and will see a wide range of temperatures, therefore ideally they will want to develop a temperature vs hold voltage profile in order to ensure it stays actuated but maximizes their power savings.
    3) The diode I have in the test circuit is a RGP30GHE3/73 from Vishay. I am definitely open to a cheaper part that better meets my needs.

    To run down the replies:

    mjhilger & meBigGuy -
    I meet with my customer tomorrow and can inquire if they will do PWM or not. If not I will definitely be looking at Bob's example you and meBigGuy pointed out.

    meBigGuy - I found a reference to driving the coil with an H-bridge and the improved turn off performance that results from it. Thanks, I will definitely be breaking out the old breadboard and buying an H-bridge or two from Digikey.

    jim hardy - the spring tension is not adjustable and there is no airgap in the actuated position.

    Thanks again everyone.

    Attached Files:

    Last edited by a moderator: Apr 10, 2014
  8. Apr 10, 2014 #7

    A follow up. Can anyone help with specifying the power rating or reverse current on the zener diode? I do not know the inductance of the solenoid but I do know the hold current and resistance of the coil.
  9. Apr 10, 2014 #8
    I have not followed up to the extent you are headed, so I don't know all the info you need. But from a physics point of view there is an energy in the magnetic field that will be collapsing that needs to dissipate; coupled with that is the fact that some energy might be added back to the field as the solenoid releases and the energy of the force (K*X) [spring constant * distance] used to hold the plunger is added back in the field, or at least not used.

    I think I would try and identify the solenoid that might be used and run some tests with a current and voltage probes in the flyback diode circuit maybe adding a small resistance 1 ohm or 0.5 ohm, to determine the energy released and consider a 50% safety margin or so and design from there.
  10. Apr 10, 2014 #9


    User Avatar
    Gold Member

    Maybe you can just assume the current is linear starting at the holding current and ending at the time when it voltage collapses across the zener. Calculate the average I^2 and multiply by the zener voltage.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook