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.

DRV8350R: DRV8350R

Part Number: DRV8350R
Other Parts Discussed in Thread: DRV8350

Dear Sir/Madam,

We have a problem about DRV8350 SOx current sense as bellow; Both sense resistor voltage and drive output voltage have a problem. However, our real signal is fine when we measure it by a clamp sensor as shown in figure. Can you explain the What is the problem.

  • Hi Omur,

    Thanks for posting to the MD forum!

    May you share your schematic? Do you have any output filters on the SOx signals? 

  • Hi Aaron,

    We used low pass filter next to the DSP analog inputs. Cut off frequency is about 4.8 kHz.

  • Hi Omur,

     

    Thanks for sharing the schematic and your patience.

     

    I have a few questions and observations from your scope shots and schematics:

    1) Could you provide more info on your scope shots? Specifically, could you measure the peak current and voltages, confirm where they're being probed at on the schematic (i.e. single-ended or differential), and confirm they're biased at 0V?

    2) Are you using the DRV8353RH device? You are using the hardware variant and the CSAs, so would just like to confirm as seeing DRV8350 on schematic can be confusing. 

    3) This shouldn't affect the result much, but typically when choosing a cutoff frequency for SOx filtering, we recommend it being at least twice your PWM frequency, i.e. around 40kHz. If you used a 10n cap rather than 100n, you'll get a cutoff frequency of 48kHz which should pass more important frequencies. 

     

  • Hi Aaron,

    Thanks for your quick reply. You can find the details as below.

    The device code is DRV8353RHRGZT. And we used TPD4E001QDBVRQ as given in figure-1.

     

    Figure 1

    There are two different result for 20kHz and 20MHz scope filter. All probes are isolated. A and B channels are biased to 0V.

    Channel - A: digital signal in the DSP with 12-bit ADC. This signal measured via 12-bit DAC.

    Channel - B: SOA pin of the DRV8353RH.

    Channel - C: The voltage of Rsens (2mohm) 

    Channel - D: Actual current measured via clamp probe.

    Scope used 20kHZ bandwith filter in figure 2.

    Figure 2

    Scope used 20MHZ bandwith filter in figure 3.

    Figure 3

    Best Regards,

    Omur

  • Dear Aaron,

    I changed the filter as 58kHz R=330ohm, C=8.2nF The current waveform is fine. But there is a new problem in the signal as below

    signal crashes periodically.

    Best Regards,

    Omur

  • Hey Omur,

    I'm glad you were able to get a sinusoidal output waveform on the SOB signal. You may need to raise the filtering more, we typically filter in the hundreds of kHz range and even up to 1MHz to only filter out high-frequency noise. 

    Could you turn off the DSP sampling and just measure the SOB signal to see if this effect goes away?

  • Dear Aaron,

    I disconnect the output of DRV from the DSP and the waveforms are given below;

    Best Regards,

    Omur

  • Hi Omur,

    Could you repeat a few of these measurements again without the 2.2k pulldown resistor? The 2.2k pulldown resistor is creating a voltage divider with Rf and changes the cutoff frequency equation to fc = 1/(2*pi*(Rf || 2.2k)*C. Typically for Rf we use small resistances (i.e. 56-ohm) and C = 2.2nF.

  • Thank you Aoron,

    Thank you very much for answering very quickly and for your time.

    I measured again without pull-down resistor. But there is a continuous cutting top of  the signal.

    Best Regards.

  • Hi Omur,

    Just wanted to let you know that I'm still looking into this with my team and I'll get you a response this week.

  • Hi Omur, 

    Can you ensure that you are taking DC measurements of your waveforms? Because your sinusoidal waveforms are biased at 0V, I'm assuming you're taking AC measurements on your oscilloscope.

    Since VREF = 3V, I would expect to see your sinusoidal waveforms biased at 1.5V (VREF/2), and when no current is driving the motor, the SOx signals should be 1.5V. Could you confirm this please?

  • Thank you Aaron. We are waiting your reply. Regards.

  • Hi Omur,

    I replied earlier in the thread, please confirm whether you are taking DC measurements of your waveforms and what your SOx waveforms are when no current is going through the motor.

  • Hi Aaron,

    SOX voltage is zero with some noise. Motor is OFF.

    Channel A 50mV/div

  • Hi Omur,

    This is interesting. I was expecting to see SOx = VREF/2. 

    Was the part powered and enabled when this measurement was taken, and can you confirm that VREF = 3V on the oscilloscope?

    Was this measurement taking during auto-calibration (i.e. <50us after powerup)?

  • Hi Aaron,

    Sorry for my mistake. I forgot enable driver.

    I take a new measurement when driver is enable. 

  • Hi Omur,

    Thanks for confirming this.

    Another option I'd like to try is removing the TPD4E001QDBVRQ if possible and remeasuring the SOx signals. Typically, we use ESD Protection devices for DC sense signals like voltage sensing on phases, and try to avoid adding anything extraneous on the SOx signal lines as they are more sensitive. 

  • We remove the TPD4E001QDBVRQ. But nothing changed. Can we test the DRV. I think there is something wrong about DRV op-amp.

    New waveform

  • Hi Omur,

    Let's recap what we have done so far:

    - The best scope capture was through the DSP DAC, but the signal was crashing periodically

    - The scope captures from when the 2.2k pulldown resistor was removed seemed to "cut off less" and improve as the cutoff frequency is lowered. 

    - Since then, many of the scope captures have been taken using AC coupling. Can you take scope shots from now on using DC coupling because SOx signals should be biased at 1.5V (confirmed from scope capture when motor is off and driver is enabled)

    It looks like the frequency of your sinusoidal current is 10Hz. Could you try selecting a filtering cap to get a low cutoff frequency, i.e. 47Hz to see if the signal improves? Typically we don't have many customers driving sinusoidal current through the DRV835x devices, so I'm curious if high frequencies from noise or some external source are causing to periodically cut off the signal. 

    What are you supplying for your VREF = 3V? Typically it needs to be a low noise source, i.e. a voltage reference IC. 

    I appreciate your patience as we work through this issue.

  • Hi Aaron,

    Actually, I appreciate your patience as we work through this issue. It was a great pleasure working with a professional like you. 

    I gave up using SOx in our application, instead I used a hall effect current sensor. We obtained the signal we wanted as below.


    I am very grateful for your time. 

    Best Regards,

    Omur

    New measurement related to SOx your last requested.

    Below you can see the result we obtained with the hall effect current sensor.