IWR6843: Cannot open COM port

Hello

Can you please confirm if you have set the device in the right mode for flashing before attempting the loading of image.

Thank you,

Vaibhav

Yes I did. SOP was set to 101.

Hello Mustafa, 

Please make sure the COM port you're trying to use is available, sometimes it engaged somewhere else which causes this issue. Restarting the PC sometimes helps here in case you aren't sure where the COM port is engaged.

You can also take a look at our uniflash FAQ : (+) [FAQ] mmWave Sensor- Uniflash related queries - Sensors forum - Sensors - TI E2E support forums

Mustafa Homsi said:

More details were sent over email

Would it possible to attach those details (if relevant) here as well?

Regards,

Ishita

Hello Ishita,

We made sure the COM port is available and tried restarting the pc multiple times. Please keep in mind that the same pc setup can flash the code to a TI IWR6843 ISK mounted on a boost dev board. We also confirmed that the Si LAB USB bridge is functional by sending characters on the UART port and receiving it on the USB side.

Regarding the additional info, I can share it, but can you convert the ticket to a private ticket. Thanks!

Hello Mustafa,

Can you connect to the UART COM port using Teraterm or similar emulation software?

Regards,

Ishita

Yes, I am able to open the port using PuTTy.

Hello,

Okay, so the COM port is accessible. Please see is SOP Mode 5 (101) is set up correctly.

You may have to kill all the previous failed attempts of Uniflash programs from the Task manager (Including Flash python programs). Any previous failed attempt background application could block the UART port for the fresh Uniflash programming. 

As recommended by the EVM user-guide, It is important to apply reset immediately after the SOP mode change (Or power recycle), Otherwise right boot mode is not registered. 

If all of that doesn't work, you could also try this on a different PC/laptop as well.

Regards,

Ishita

Hello Ishita,

As mentioned before, I have already confirmed the SOP mode. I always kill all related tasks, and I have tried multiple times on a fresh restart and different PCs. I was never successful. We always apply resets before loading the image. 

Not sure how we can troubleshoot this further. Is it possible to send a simple command over PuTTy and have the radar chip send a response back to confirm communication is successful? As mentioned, this is a custom board, and we are assuming that IWR6843 is shipped with the UART bootloader flashed. We haven't loaded any program over the JTAG connection. Thanks!

Hi Mustafa,

 Can you please confirm the flash part number ? 

regards
AK

Sure. Here is the one we are using.

Thanks. The flash part MX25U1633FZUI shared is supported.  then one thing that i suspect is the SOP[2:0] -101 not being registered by the device. Please make sure that the SOP - pull ups reference to 3.3V and weak pull ups are avoided. You may try to hardwire the SOP pins  to flashing mode and test to rule out pullup/pulldown circuit.  

Abdulraheem, in our current setup, we have SOP2 (P9) shorted to 1.8V, SOP1 (G13) connected to gnd through 100kOhm and SOP0 (N13) connected to 1.8V through 100kOhm. Are you saying we should replace the 100kOhms with shorts? Please note VIOIN is connected to 1.8V and not 3.3V.

Is there a way to communicate with the bootloader over UART and query its mode of operation? Wouldn't this address the concern of whether the chip entering flashing mode?

Thanks Mustafa. Not the shorts but a 10KOhm would work. And once you set the SOP-101 and try to program the flash, does the Uniflash utility hang? Can you probe nError_out (TP213) signal state at this point ? If asserted then SOP are not set correctly. 

regards
AK

We will change SOP0 and SOP1 resistors to 10kOhm. Is it okay that we are shorting SOP2 to 1.8V or do you want us to use a 10kOhm instead? Please keep in mind that we have to keep a pull down resistor in order to bring it back to SOP001 mode. Right now it is 100kOhm.

The Uniflash utility does hang. nError-out is high but haven't checked while trying to load code. We will check it. We shared the hardware bring up checklist results over emails. It has info about nError_out and nRST.

There are no commands we can send manually over UART?  

Hi Mustafa,

As discussed, the power sequence should be looked at here. Also, can you probe the VBGAP pin after powering to make sure this comes up and indicates powered on state?

Regards,

Jackson

Hi Jackson. What is the recommended power-up sequence? All I was able to see in the datasheet is the following comment "The IWR6843 device expects all external voltage rails to be stable before reset is deasserted. " This is in section 8.10.1.

Today. we power up 1.2V first, then 1.8V, then 1V and once we confirm all are stable, we pull nRST high. Is this acceptable?

We measured 0.95V at VBGAP.

Hello Mustafa, 

I believe the debug for this is going on offline as well with Lindsey Fike over email. We will post regular updates here as well if you want. 

Regards,

Ishita