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.

DRV110: Current unexpectedly changes with voltage to solenoid

Part Number: DRV110

Hello,

We have built a custom board designed from your reference design for the 24V solenoid.  However, we are using a 28V solenoid and are not powering any other circuitry from the voltage going to Vin of the DRV110.  

The Excel tool (thanks for this by the way) was used to provide values for the components.  The data is shown below:

The schematic for our board is shown below.  Note there are a number of Do Not Installs:

We were surprised that instead of a 1.33A hold current, we are getting about 0.8A at 28V.   More surprising was that as we ramped the voltage down (28, 24, 20, etc), the current increased.  At 18.8V, the 1.33A seemed to be achieved.  Further decreasing this voltage showed the current dropping again.Our solenoid is rated typically for 24V and has about a 9 ohm resistance.  

The variation with solenoid voltage was very confusing.  The only apparent influence would be from the DRV110A_VIN power.  No other pins are referenced to the incoming power supply.

Does anyone have any similar experience or suggestions?

Thanks!

  • Hi Matthew,

    Thanks for your post! I'm recreating your setup and attempting to find the source of the problem. I'm hoping to have an answer for you soon.

    Can you clarify what (if any) source resistance is connected in series with your DC power supply? Also, are you measuring 2.8A peak current or is that lower than expected as well?

    Travis
  • Hi Travis,

    Thanks!  To answer your questions...

    1) No source resistance, the power supply is driving directly

    2) I measure about 2.6 or so.  However, if the solenoid is hooked directly to the power supply, this is the same current I get.  The 2.8A was intended to make sure we max out the current that the solenoid can handle for pull in.

    Thanks,
    Matt

  • Thanks for the information. While I am working on reproducing the problem on our end, you may try adding resistance between the source and the input node on the device. According to the parameter calculator, it appears that 12kohm would serve you best with a 28V input. That should limit the current draw of the internal zener diode and may improve performance. Let me know if you observe any changes.

    Travis
  • Thanks Travis. If you look at the schematic, we have a 20V zener in line with the 28V, then a 4.7k resistor between that and the internal zener in the DRV110a. Is this sufficient or am I missing something?
    Thanks,
    Matt
  • No, I believe you are correct and that should be limiting current sufficiently. I noticed that your MOSFET is rated nominally at 65nC of gate charge when Vgs = 10V, and the application note suggests a maximum gate charge of about 10nC. Since the DRV110A uses PWM to drive the gate, it's possible that your FET isn't switching fast enough for your application. You may take a look at TI's NexFET portfolio for a switching device with lower gate charge. Let me know if this helps.

    Travis
  • We can try to address this.  However, how would that result in a curve as seen below?  This is a measurement we did today on a slightly different solenoid, but the characteristics of the curve are what we don't understand of course...

    Thanks,
    Matt

  • Can you send some oscilloscope snapshots of your output voltage and drain current for various input voltages? 

  • Just to clarify my previous post- can you capture oscilloscope shots of these nodes specifically for 15V and 28V (data points on either side of the vertex in your graph):

    1. Vout (gate-driving voltage) zoomed in enough to see the PWM switching
    2. Vsense (current sensing node)
    3. Idrain (solenoid current) zoomed in to see the peak and hold current levels

    Let me know if you're able to produce these shots, as I think we'll be able to diagnose the cause with them.

    Travis
  • Hi Travis,

    Unfortunately, couldn't get all of them on the same scope.  But did Vout/gate and Vsense on one, then showed the voltage difference across the current sense resistor (0.03 ohm).  We did a duration test on the solenoid yesterday, and it may have been damaged a little.  While the current vs. voltage still is not constant, the 'peak' has shifted down towards the 15V side.  Still, the fact that the current changes at all with voltage is what is a mystery to us...

    Thank you for your help!

  • Thanks for the scope shots Matt!

    The Vout waveform confirms my belief from earlier- the MOSFET you are using is not achieving the rise time necessary for proper operation, as evidenced by the curved rising edge. I have included the spec from the data sheet and a quick calculation that shows the reason why the FET charges slowly.

    Since the DRV can only drive about 10mA of current directly, the calculation yields about a 6us charge time (seen in your scope shot):

    As you can see, the FET is designed for an 18ns rise time when driving with 10V. While this isn't exactly the same spec as the charge time (and the charge time isn't necessarily linear), I think it's still a good indication of how close to the mark we are. 

    I think you will need to do one of two things to achieve the output current you are hoping for.

    Option 1: Replace the MOSFET with one that has a significantly smaller gate charge that should observe a shorter rise time. Just with a short search, I found the following FET to be very reasonable in terms of the spec list: http://www.ti.com/lit/ds/symlink/csd19538q2.pdf 

    Option 2: Add a gate driver between the DRV and the MOSFET in order to charge the gate more quickly. 

    Best of luck!

    Travis 

  • Thank you very much Travis! I'll make these mods and post results...
  • Hi Travis,

    We replaced the FET with this one (4.6nC gate charge):

    FET selection link 

    It clearly made the transitions sharp and fully formed.  However, it did not change the fact that the current changes with voltage.

    Any other suggestions?

    Thanks!

  • Matt,

    Can you replace the link for your FET? I cannot access the website with the link you posted. Do you have a current probe? We would be able to deduce a lot about the behavior of your circuit with some screenshots of the current during the peak and hold times. Due to the high-current nature of your system, there could be a number of reasons why the supply is not outputting the full amount of current. One suggestion I have is to add in the LPF for a crisper signal.

    Let me know if you can get those scope shots, otherwise we will have to try a different approach.

    Travis 

  • Hi Travis,

    Let me work on getting the current info.  in the meantime:

    www.digikey.com/product-detail/en/stmicroelectronics/STS10N3LH5/497-10010-6-ND/2170235