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.

DRV8870: Jittered Output and Lower Output Voltage

Part Number: DRV8870
Other Parts Discussed in Thread: , INA240

Hi Team,

When the coil load is driven by DRV8870 which I bought from TI Store last month, the current of load changes as shown in the following figure:

where the horizontal axis is the measuring time, the vertical axis is the load current and VM = 30V, the load is <1Ω, <1mH.

First, the current rises as expected. Then, after the current reaches the peak point, it starts to decrease slowly, which is unexpected.

I found that this phenomenon is caused by 3 reasons:

  1. The temperature of the load and the chip increase leading to the increase in resistance(load and chip internal RDS);  It is the reason for the current drops like the latter part of the above image.
  2. The output of DRV8870 jitters, and its range accounts for 5% of total output signal, shown as followed. No jitter in input PWM. This is also the reason for the current drops like the former part of the above image(current from 1686mA to 1602mA).
  3. The high-level voltage of the output signal is always lower than VM 5~6V, e.g., Input VM = 30V, my measured output is 24V, wondering where is the 6V difference.

Note that the reason 2 and 3 are found under the measurement without load, so these reasons have nothing to do with the load. 

And I want some explainations about the reason 2 and 3, that is jittered output and lower output voltage while correct input.

Thank for your help.

  • Hi Hector,

    I have a couple questions:

    1. Are you using the TI DRV8870EVM or your own board when you took these measurements?
    2. If you are using your own board, are you able to share the schematic and layout of the board (you can send it via private message)?
    3. Is this issue observed in more than board? 

  • Hi Pablo,

    I'm using my own board:

    The coil load is connected to J1, I use INA240 to measure the current of load. Under the situation without load, it is nothing to do with the INA240.

    I only measured on this board, but I have observed this issue for several DRV8870.

  • Hi Hector,

    Thank you for providing the schematics. I will need more time to look through your schematic. Expect a reply from me by 8/2 US time.

  • Hi Pablo,

    Thank for your help! Is there any results about my problem?

    Hector

  • Hi Hector,

    What is the purpose of the R1 and D3? They are connected in series between the driver outputs OUTA1 and OUTB1 so technically there is a load connected to the driver outputs. Even though the load will not draw much current, we want to completely eliminate the load from the equation. Can you retake the measurements without R1 and D3 populated on the board? 

    What is the forward voltage of the diodes: D1, D2, D4, D5? Whenever the current flows through the diodes, VM will drop slightly by ~2 times the forward voltage of the diodes. This could explain the drop in VM.

  • Hi Pablo,

    Thank for your help. D3 is used to indicate output status and R1 is a current limiting resistor for LED D3. Although I don't think this indicating part is the cause, I will retake the measurements without R1 and D3.

    From the datasheet, I find the forward voltage of the diodes is 0.5V(3A, pulse test: 300 μs pulse width, 1 % duty cycle ). And as you said, the value of the drop in VM is about 1V, which is lower than the measurement result 5-6V. Is there any cause about the drop in VM?

    Hector

  • Hi Hector,

    Let me do a quick correction on my early comment: "Whenever the current flows through the diodes, VM will drop slightly by ~2 times the forward voltage of the diodes. This could explain the drop in VM." Actually, you also have to take into account the voltage drop across the status LED. What is the voltage drop of D3? 

  • Hi Pablo,

    I retook the measurements without R1, D3, D1, D2, D4 and D5 populated on the board, but the problems still exists:

    Without the load, the output of DRV8870 still jitters, about 4% of the whole positive pulse width (0.1us / 2.5us). And the output voltage is significantly less than VM, 16.6V out compared to 20V in.

    Could there be other reasons?

    Regards,

    Hector Jing

  • Hi Hector,

    What is the PWM frequency and duty cycle? The input PWM signals going to the DRV8870 will synchronize with the internal clock. This will cause around a 200ns jitter at the output which is similar to your measurements (Another similar e2e post has been answered here: https://e2e.ti.com/support/motor-drivers/f/38/t/814224?tisearch=e2e-sitesearch&keymatch=jitter ). I believe this is the reason for the jitter.

    In regards to the lower output voltage. I am puzzled as to what could be causing that. Could you send me a scope shot showing the input power supply (VM) together with the output voltage? Seeing the input and output voltages together might give us a clue.

  • Hi Hector,

    The input logic pins of this device have deglitch circuits which synchronize with the internal clock. The OUTx pin state changes on the edge of the deglitcher. The jitter you see is due to the deglitcher catching the signal at different times relative to the internal clock. We have these deglitchers so this device can be robust in a noisy environment. In typical motor applications, this is not an issue since the inductance is high enough. However, you see this issue since the inductance of you load is very low.