DRV8353R:GDUV error active when using DRV8353RS

youngtak han

Part Number: DRV8353R
Other Parts Discussed in Thread: CSD19532Q5B, DRV8353

Tool/software:

This is a continuation of the previous question.
I checked the datasheet of DRV8353RS and it says that GDUV should be 0 at Addr 1 for SPI communication.
However, in the same datasheet, it says something like the picture below.
Which is correct, 0 or 1? I ask because sometimes there are typos in the SPI communication data as well.

On power-up, the GDUV is enabled as shown below.
Replacing the DRV8353RS and the MOSFET causes the same symptoms.
In your previous question, you suggested using CLR_FLT to perform an initialisation, I tried it but the same result occurs.

The MOSFET is using CSD19532Q5B. Have a nice day and look forward to your reply.

2 months ago

0 Davis Jackson 2 months ago

TI__Intellectual 2940 points

Hi Youngtak,

There are two relevant registers here, one being Fault Status Register 2 which reports whether GDUV has occurred (0 if not and 1 if it has), and the other being Driver Control Register which allows you to enable or disable the GDUV fault through DIS_GDUV as shown in the table you attached.

Could you please provide a waveform for the VCP and VGLS pins? These determine the GDUV fault.

Best,

Davis

0 youngtak han 2 months ago in reply to Davis Jackson

Prodigy 90 points

Thank you for your response.
I will measure and attach the waveforms for the VCP and VGLS pins as soon as possible.
So, if VCP and VGLS are UV when DIS_GDUV is 0, does that mean GDUV is enabled to 1 when VCP and VGLS are UV?

0 Davis Jackson 2 months ago in reply to youngtak han

TI__Intellectual 2940 points

Yes, that is correct. GDUV reports the fault over SPI and the nFAULT pin should latch.

0 youngtak han 2 months ago in reply to Davis Jackson

Prodigy 90 points

The waveforms for VCP and VGLS are shown below.
When 48V is applied to the VCP, it is observed on the oscilloscope as 48~50V, but there is no change in the VGLS. Is it possible to check if there is a problem with the circuit on the VGLS side?

[VCP]

[VGLS]

5582.DRV8353RS.pdf

0 Davis Jackson 2 months ago in reply to youngtak han

TI__Intellectual 2940 points

It would appear that there is an undervoltage condition on both VCP and VGLS based on the ratings in the datasheet. I haven't found any issues looking at your schematics, but please check if there is any short from VGLS to ground. Additionally, the section on High and low side pre-driver undervoltage in this FAQ might be useful. Please let me know if this helps.

0 youngtak han 2 months ago in reply to Davis Jackson

Prodigy 90 points

Thank you for your response.
Based on the datasheet below, we determined that V_VCP should be 7.5-11.5V when V_VM is 12V and V_VGLS should be 10-12.5V, so we don't have an issue with V_VCP.

In your answer, you said that two items were under voltage, so you looked at the datasheet again.
I understood that for VGLS, it is an undervoltage condition.
So if Vdrain is 48V, should VCP be 48+5V?

0 Davis Jackson 2 months ago in reply to youngtak han

TI__Intellectual 2940 points

The V_VCP spec is measured with respect to VDRAIN. Since VDRAIN looks to be 48V from your schematic, the measured 48-50V on VCP in the scope image is undervoltage given the second spec you sent. The two possible causes I can think of at the moment might be a short between VGLS and ground or the placement of the capacitors on the PCB layout. Please let me know if you are able to find anything based on the FAQ I linked in my previous response.

0 youngtak han 2 months ago in reply to Davis Jackson

Prodigy 90 points

Thank you for your response, and I apologize for asking the same question multiple times.
Is that section "High and low side pre-driver undervoltage" included in the datasheet for the DRV8353?
Can you point me to the section material?
If GDUV is active and persistent, can it damage the motordriver?
I ask because I want to know if the problem is caused by driving the motor with GDUV enabled.

0 Davis Jackson 2 months ago in reply to youngtak han

TI__Intellectual 2940 points

No worries at all, sorry for the confusion.

Here is the FAQ I'm referring to: https://e2e.ti.com/support/motor-drivers-group/motor-drivers/f/motor-drivers-forum/1176539/faq-quick-guide-to-debugging-common-issues-in-bldc-motor-drivers. It contains the section on High and low-side pre-driver undervoltage (section 9).

Section 8.3.6.2 in the datasheet contains more information on the fault reporting, but the FAQ link provides more details on debugging this fault.

Could you also please provide scope images of VM, VCP, VGLS, and VBUCK_12V at the startup of the device? This will help me understand all the dependencies that determine the outputs of the charge pump and low side regulator.

Lastly, can you confirm if this device is already damaged? It's possible that device damage might be causing the charge pump and regulator to be undervoltage.

0 youngtak han 2 months ago in reply to Davis Jackson

Prodigy 90 points

The motor driver was replaced with a new motor driver, but the same symptom occurred.
However, it seems that the A-phase mosfet is currently damaged and it is difficult to provide waveforms. It seems that there is a problem with the design circuit, so we will analyze and fix it.

VCP and VGLS are not working properly even though I replaced the motor driver before the mosfet was damaged. In the case of VM, 12V is being input normally.

Can you check the waveforms of the VM, VCP, VGLS, and VBUCK_12V you mentioned to make sure they don't go below the low voltage threshold in the datasheet? I will check after repairing the board.