• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 30 subscribers
  • Views 1370 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.

DRV8834 nFAULT signal toggling every 90 microseconds

Gerald Hubbell
Prodigy 55 points
Other Parts Discussed in Thread: DRV8834, DRV8834EVM

Hello all,

 

I have designed a circuit around the DRV8834 PWP and have a prototype circuit card working. The microcontroller is issuing the proper commands in index mode to the stepper motors, STEP and DIR and I set AVREF and BVREF to 300 mA correctly. The nFAULT signal is toggling every 180 microseconds.(See attached). I used the POLOLU carrier for this chip to breadboard with (it worked great), and basically have built the circuit using that as the model on my prototype PCB. All the input signals to the DRV8834 are correct, except the stepper motor is not turning and the nFAULT line is toggling as mentioned. Any ideas?  This is a very small board so I am limited in what I can do to correct this. I am interested in what may cause this.  The chip is not hot, the current is limited, and the VM is okay (5 Vdc). This problem is on 2 of my prototype boards, I have 10, so I expect it is on all of them.

 

Thanks 

  • 0
    TI__Guru** 114905 points

    Hi Gerald,

    Please check your board to determine the reason of nFAULT. One common reason is that VM dips below the operating voltage. This can be caused by the size of the bulk capacitor or the placement of the VM capacitors on the board. Also if the grounds are not connected, this can be a reason.

    Please refer to the schematics of the DRV8834EVM for additional reference. These can be found on the DRV8834EVM product folder. Scroll down on this page to Related Products > Software.

    If after this, you have not been able to identify the problem. Please provide schematics and layout for review.

    Thank you.

  • 0 in reply to Rick Duncan
    Prodigy 55 points

    Hi Rick,

    Thanks for your quick response. I have a 10 uF cap on the VM (5.0 Vdc) and also just to make sure, I placed a 1000 uF electrolytic on the 5 Vdc bus also, and the result was the same. The nFAULT cycle time is 90 microseconds, not 180 (as you probably saw right away, my bad). The grounds are connected as needed I would like to send the schematic and layout to  you directly instead of posting them here on the forum if that is okay.

    Thanks

    Jerry

  • 0 in reply to Rick Duncan
    Prodigy 55 points

    Hi Rick,

    I thought of something else that may or may not impact the operation of the OCP. In my design, I use a PWM DAC RC circuit that I have scaled to output up to around 3 Vdc into the VREF. With the stepper motors I am using, I wanted to limit the current to around 600 mA so I chose an ISENSE resistor value of 1.00 ohm. Is this value too much for the chip. In the example in the datasheet they use a value of 0.400 ohms. Could this be the problem?

    Thanks

    Jerry

  • 0 in reply to Gerald Hubbell
    TI__Guru** 114905 points

    Hi Jerry,

    Setting the sense resistor to 1 Ohm is too high. At 600mA, it violates the absolute maximum of the xISEN pins. Please lower it. Reducing the resistor will also reduce power losses. You can use the xVREF pins with the sense resistor to set the current limit to 600mA.

    Can you look at the VM signal at the inputs to the DRV8834? It is possible that the voltage is dropping there although it looks good upstream.

  • 0 in reply to Rick Duncan
    Prodigy 55 points

    Thanks Rick for your thoughts. I will look again at the VM with the scope, but I have a gut feeling that since the ISENSE   resistor value is too high, that is what the problem is. Thanks again.

    Jerry

  • 0 in reply to Rick Duncan
    Prodigy 55 points

    Hi Rick,

    I looked at VM1 and VM2 and measured 4.971 Vdc and looked at the points with the o'scope and saw just a little bit of noise in the 5 mVdc range but the voltage is steady on the meter. I don't believe the voltage is dipping.

     

    I also put a short across the AISEN and BISEN resistors (1.00 ohm) and did not see any difference in the nFAULT signal, so that may not be the root cause of the problem. Could it be that when I had these boards made, that the DRV8834 chips were damaged in some way causing the fault to be in?  The only recourse I have now is to redesign the circuit around the DRV8834 to replace the ISEN resistors with 0.4 ohm values and also make sure the layout is solid and do another turn of prototypes. Unless you can suggest something else, I am at a loss as to why it is not working.

    Thanks

    Jerry

  • 0 in reply to Gerald Hubbell
    TI__Guru** 114905 points

    Hi Jerry,

    Thank you for sending the schematics and layout. These allowed us to identify two common shortfalls. I am documenting them here so that we will all learn from this.

    First is the lack of adequate decoupling. This is seen quite often. The datasheet does not always show the recommended capacitors on VM (bulk and local). That is why we recommend looking at the EVM developed for the device. The EVM schematics and gerbers can be found in the product folder. For reference, the bulk capacitor is typically larger than required because we are trying to demonstrate a broad variety of applications. The local filter capacitors on VM (both size and placement) are important to minimize local drooping on the VM inputs.

    Second is the missing ground connection to the PPAD. In most cases, the description in the terminal functions section will state how the PPAD should be connected. Again, it is recommended to use the EVM schematics as a guide.

    Thank you again for allowing us to help everyone.