TPS65987D: Problems writing to flash

Darren Jenkins

Part Number: TPS65987D

USB PD team,

Is there a write protect feature on the TPS65987D? Our customer is unable to write to the external serial flash with their host processor. They know the host processor is communicating with the TPS65987D.

They had an earlier set of boards work properly, and they were able to write the flash on those boards. Now a new set of boards is seeing this problem.

Thanks,
Darren

over 4 years ago

0 Michael Emanuel35 over 4 years ago

TI__Mastermind 40575 points

Dear Darren,

What are the differences in the flashing process between the new and old boards? Has the layout/schematic changed?

Thanks,

Mike Emanuel

0 Darren Jenkins over 4 years ago in reply to Michael Emanuel35

TI__Genius 12511 points

Hi Mike,

The only thing was that PP_HV2 was tied to 5V when the boards were first built. They changed this to pull to ground (as on the original board), and the problem remains the same.

Thanks,

Darren

0 Michael Emanuel35 over 4 years ago in reply to Darren Jenkins

TI__Mastermind 40575 points

Darren,

I am having trouble understanding your setup. You said your earlier boards were working, but you also say that the problem remains when you undo the only change. Are you sure you're using the same flashing process? Is the PD controller being powered properly? Please let me know if there are any other differences between your working and non-working board.

Thanks,

Mike Emanuel

0 Darren Jenkins over 4 years ago in reply to Michael Emanuel35

TI__Genius 12511 points

Hi Mike,

I clicked the wrong button. I didn't mean to close this thread!

I understand your point, so I'm trying to figure out the right area to investigate here. There must be another factor. The customer has asked if it could be related to a new date code of our part, and if that new date code is treating write protection differently.

They are programming the TPS65987D through the I2C port, and are using the 4CC Flash erase and write commands.

When they issue a write or erase command, they get a normal response of 0000 indicating the command was successful. They have experimented with issuing the LOCK command, and haven't seen any impact on the response to an erase/write (0000).

Thanks,

Darren

0 Michael Emanuel35 over 4 years ago in reply to Darren Jenkins

TI__Mastermind 40575 points

Darren,

Please get back to me with the specific differences between the boards, setups, etc. This will help further debug.

Thanks,

Mike Emanuel

0 Darren Jenkins over 4 years ago in reply to Michael Emanuel35

TI__Genius 12511 points

Hi Mike,

Where would we find source code for software on a host controller that erases and writes a new PD Controller application to the SPI Flash tied to the TPS65987D?

Thanks,

Darren

0 Michael Emanuel35 over 4 years ago in reply to Darren Jenkins

TI__Mastermind 40575 points

Dear Darren,

If you want to flash from a host controller you will have to write your own code to flash the PD controller using I2C. Depending on your setup, you can also flash from the Aardvark using the full flash binary or flash using a USB from the GUI tool.

Thanks,

Mike Emanuel

Please click "Resolved" if this answered your question.

0 Darren Jenkins over 4 years ago in reply to Michael Emanuel35

TI__Genius 12511 points

Mike,

The problem ended up being caused by the fact the FLem command does not cause the PD controller to wait for the Flash Busy Flag to be cleared while a sector is being erased. I'll email you some more details to discuss if documentation can be updated.

Thanks,
Darren

Interface

Interface forum

TPS65987D: Problems writing to flash