DRV8353: 电机驱动器论坛

We use drv8353 in our motor control system, vbuck is OK, output signals also OK. We use 4 GPIOs to connect to DRV8353 SPI pins, trying to control DRV8353 by SPI interface, but SDO always output 0xff, could you please help to check is there any problem in schemetic and final signal, thanks a lot in advance

For the first waveform, yellow is nSCS, blue is clock.
For the second waveform, yellow is clock and blue is MOSI
There is no need for the third waveform because MISO is always high.

Tandi

Di Tan：

We are using DRV8353RS,  trying to read from address 04h, but the output from MISO is always 0xffff, it is quite urgent, any suggestion is appreciated, thanks a lot

,

Cherry Zhou：

Hi Tandi,

We're looking into this issue and please expect the response.

Thanks

,

Cherry Zhou：

Hi Tandi,

Is the device getting enabled  well before you read the spi signals? Are you waiting for tWake before sampling?

Also I checked the register 4h

All the default values are 1 so maybe try writing a different setting with 0s in it to see if the devices is reading and writing properly.

Thanks.

,

Di Tan：

Hi Cherry,

Thanks for your reply. 

Yes, DRV8353RS is enabled, and we waited a long time (more than 10us) after system start and then keep on reading DRV8353RS register by spi. We also tried other registers, like 01h, 05h, still got all high output, thanks 

,

Cherry Zhou：

Hi,

The twake for this device is 1ms.

It seems like there might be some issue with the order in which the signals are being sent/read. I suggest consulting section 8.5 of the DS for SPI timings

Is the PVDD voltage stable during this event? Has the customer tested our EVM for this device, since that would be useful design and evaluation tool.

Thanks

,

Di Tan：

Cherry, 您好，twake 是 10ms， 不是 10us，写错了。

我测试到第三块开发板时，spi 通讯成功，前两块一块是全新生产出来的板子，第二块是 drv8353 被烧坏了，重新手工换过 8353 的板子。前两块板子也许是硬件的问题，或者是驱动的时序处于临界状态，不能确保每块板子都OK，我自己再仔细查一下。您看这个波形假如还有改进建议的话也请帮忙提一下，谢谢

,

Cherry Zhou：

已跟进给工程师，应该会在下周工作日给到您答复。

,

Cherry Zhou：

您好，请看以下答复：

It is likely that the board damage could be a potential reason for the SPi issue seen.

What did the customer change in the third design to prevent he burn down observed in the first two boards?

Please check out the common debug FAQ on tips on MOSFET damage and overheating.

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

Please check out the following guide for tips on layout. This can help reduce ground noise seen in the waveforms.

https://www.ti.com/lit/an/slva959b/slva959b.pdf

,

Di Tan：

Hi Cherry，

我们测试了更多的板子，相信应该是SMT焊接，rework 等其他原因造成的通讯不良，多谢您这边的建议和参考资料。

Thanks