AM62A3: Boot Sequence and proceedure for custom board

Hello Jianzhong,

Thank you for the input.

Do I need to flash the U-Boot binaries separately via UART, or is it sufficient to flash the .wic image to the SD card and boot the board by configuring the boot pins for SD card boot? On the EVK, we were successfully booting from the SD card using this method.

I’ve already tried booting from the SD card on our custom board, but there were no console logs indicating that the boot process had started.

One notable difference in our hardware is that the SD card slot does not have a card detect pin. To handle this, we’ve manually pulled down the MMC1_SDCD (E18) line so that it always remains low.

Thank you.

shaik adnan pasha said:

Do I need to flash the U-Boot binaries separately via UART,

No, you don't need to do this.

shaik adnan pasha said:

One notable difference in our hardware is that the SD card slot does not have a card detect pin. To handle this, we’ve manually pulled down the MMC1_SDCD (E18) line so that it always remains low.

This may be the reason why your board doesn't boot up. Let me loop in our EVM expert for further help.

Regards,

Jianzhong

Hello Adnan,

Are the power supplies as per the requirement, do you have the clock and RESETSTATz output as required.

You need to share the schematics to have a quick look.

Regards,

Sreenivasa

I have attched the schematics for sd-card please have a look.

Hello Adnan,

Thank you.

SD card interface includes SOC + power + SD card.

Part of the schematics does not help in review.

I can't see pulldown on the clock.

The SD card supply seems to be connected to a fixed 3.3V SYS voltage.

for SD card boot, the boot starts at 3.3V and the IO is switched to 1.8V. A power switch is required. 

Can you check if the software has disabled switching to 1.8V and operates only at legacy speed.

Regards,

Sreenivasa

sch-jul9.zip

i have attached the schematics, let me know if you need anything else.

Hello Adnan,

Thank you.

Let me review and comeback.

Regards,

Sreenivasa

Hello Adnan,

Looks like you attached individual pictures.

I need a searchable PDF to review.

Feel free to share over the private chat.

Regards,

Sreenivasa

Hello Sreenivasa,

I’ve shared the schematics PDF provided by our hardware team with you via private chat. Please review it and let me know if you need any additional information.

Thank.

Hello Adnan,

Thank you for the schematics.

The SOC MMC1 interface is not visible.

With the provided schematics, i see that the SD card power supply is fixed and does not have a power switch connected.

Please refer SK schematics.

You need to change the software to fix the SD card IO supply to 3.3V and operate at legacy speed.

Regards,

Sreenivasa

Hello Adnan,

The below could help

how to make the SD card use only 3.3V?

(+) SK-AM62P-LP: How to Set SD Card to 3.3v only (low speed) - Processors forum - Processors - TI E2E support forums

https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1413596/sk-am62p-lp-how-to-set-sd-card-to-3-3v-only-low-speed

Regards,

Sreenivasa

Hello Sreenivasa,

Thanks for your input i'll make the device tree changes and let you know my observations.

Regards,
Adnan

Hello Adnan,

Thank you. I assume you have the pulldown for the clock and the series resistor provision if provided is populated with 0R.

Regards,

Sreenivasa

Hello Sreenivasa,

I just had a confirmation from our team there is a 49.9k pulldown on clock and a series resistor is also provided and populated with 0R.

Regards,
Adnan

Hello Adnan,

Thank you.

If possible, please change to 10K.

Regards,

Sreenivasa 

Hello Sreenivasa,

We’ve updated the pull-down resistor on the clk line to 10k as recommended, and I’ve also made the necessary device tree changes in both the U-Boot and kernel sources. After building a new image and flashing it to the SD card, the board still doesn’t output any logs over the serial console (UART) during boot from the SD card.

Regards,
Adnan

Hi Jianzhong Xu Kallikuppa Sreenivasa 

The DDR part number used on our custom board is MT53E256M32D1KS-046 WT:L. Could you please confirm if any specific DDR configuration is required for this, or if additional setup is needed?

Regards,
Adnan

hello Adnan, 

The schematics is partial and we can only provide limited support.

shaik adnan pasha said:

We’ve updated the pull-down resistor on the clk line to 10k as recommended, and I’ve also made the necessary device tree changes in both the U-Boot and kernel sources. After building a new image and flashing it to the SD card, the board still doesn’t output any logs over the serial console (UART) during boot from the SD card.

remove C306 and test
can you measure MCU_PORz, PORz_OUT and RESETSTATz

Can you read the marking on the processor and share - did you get the samples recently 

Do you have provision to configure UART boot mode?

Regards,

Sreenivasa

Hello Adnan, 

Missed asking - did you measure the SOC supply rails 

Regards,

Sreenivasa

Hello Sreenivasa,

We will remove C306 and test and let you know the status.
I have also shared the complete schematics in your private message.

We have gotten the samples recently.SITARA AM62A32AOMHIAMB (3BP06MS).

We have provision to configure UART boot mode.

Hello Adnan, 

Thank you.

Please check if the supplies are OK and RESETSTATz/PORz is high with a delay of ~10ms after MCU_PORz goes high.

If the above looks Ok, Please configure UART mode and see if you get some logs.

Regards,

Sreenivasa

Hello Sreenivasa,

The supplies are okay but VDDSH_VSDIO (LDO1) was not coming so we have tied that pin to 3.3V.
We are getting the below log when i give reset in sd card boot mode and stops.

U-Boot SPL 2023.04-gb0d717b732 (Dec 06 2023 - 15:11:45 +0000)
SYSFW ABI: 3.1 (firmware rev 0x0009 '9.1.8--v09.01.08 (Kool Koala)')
am62a_init: board_init_f done
SPL initial stack usage: 17040 bytes

Hello Sreenivasa,

Below is the data we get while booting from uart

02000000011a0000616d3632610000000000000048534653000800000008000002a6000000000000cb39ee39c52d0469806636ff350520fcf7065cbec5cdddfea08863506c2be9f2242ff3207f919c2edcff407261f0908459139f3c153770f3024c4eae43a71151ad0bc40b0000000000000000000000000000000C

Hello,

shaik adnan pasha said:

We are getting the below log when i give reset in sd card boot mode and stops.

I see the ROM is able to successfully boot from the SD card now.

The latest issue is because of the custom DDR you are using. The default WIC image is built for the TI EVM only. For custom boards, you would have to port the U-Boot and Linux as described in the following application note.

Minimal Platform Development on AM62x Devices (Rev. A)

Please check the section "4.3.2 Generate Devicetree for DDR Configuration." for porting the DDR in the U-Boot.

Hello Adnan, 

Thank you.

shaik adnan pasha said:

The supplies are okay but VDDSH_VSDIO (LDO1) was not coming so we have tied that pin to 3.3V.

The PMIC GPIO11 is used to enable the SD card output and is required to transition from 0 to 1 to enable the SC card output.

The GPIO11 is not connected in the design. So the PMIC SD card output will not be available.

Regards,

Sreenivasa

Hello Adnan,

shaik adnan pasha said:

The DDR part number used on our custom board is MT53E256M32D1KS-046 WT:L. Could you please confirm if any specific DDR configuration is required for this, or if additional setup is needed?

The DDR configuration is for the memory on the SK.

There could be difference in the configuration.

Please start a new thread for the DDR expert to support.

Please ensure the title includes LPDDR configuration.

Regards,

Sreenivasa

Hello Adnan,

shaik adnan pasha said:

The DDR part number used on our custom board is MT53E256M32D1KS-046 WT:L. Could you please confirm if any specific DDR configuration is required for this, or if additional setup is needed?

Please change R28 2.2K/0.125W to 10K.

Regards,

Sreenivasa

Hello Sreenivasa,

Thanks a lot for your inputs.

I will raise another query on ddr configuration.

we will try chnaging the R28 to 10K as you have suggested.

Best Regards,
Adnan

Hello Adnan, 

Thank you.

Please add this thread reference when you submit the thread for LODDR4.

Regards,

Sreenivasa

Hello Sreenivasa,

I have added a link to this thread in other query.

Below is the link to the new thread for ddr configuration.

e2e.ti.com/.../am62a3-ddr-configuration-for-lpddr4-mt53e256m32d1ks-046-wt-l-on-custom-am62a-board-boot-stuck-at-u-boot-spl

Best Regards,
Adnan

Hello Sreenivasa,

After performing multiple resets, I no longer see the usual boot logs that were previously visible with the same OS image.

Initially, the first few SPL boot log lines appeared, but after 3–4 resets, no output is visible on the serial console. Even a full power cycle (turning the device off and on again) does not restore any console output.

Even without ddr configuration i should be seeing the rom code initialisation logs.

I even tried with the updated ddr configuration and still dont see any logs i have updated the same in other thread also.

In below screen shot you can see after 4 resets i can only see first 2 lines from spl rom boot and there is no line saying am62a_init_board_init done after which we are not able to get any log even after restarting the device or trying to reset also don't help?



Best Regards,
Adnan

Hello Adnan, 

Thank you.

Is this the SD card boot.

I assume you disconnected the connection to the PMIC when you connected an external 3.3V to the SD IO supply.

Regards,

Sreenivasa

Hello Sreenivasa,

We disconnected the VDDSHV_SDIO trace from the PMIC and supplied it externally with a 3.3V (3V3_SYS) source.

Since GPIO11 had no external pull-up resistor connected, there was no voltage level defined for the pull-up. To resolve this, we disconnected the previous connection and provided 3.3V from the 3V3_SYS rail.

 Regards,
Adnan

Hello Sreenivasa,

I tried changing the boot mode to uart and dont see any logs.

if give below command i get ACK time out?

Sending tiboot3.bin, 2483 blocks: Give your local XMODEM receive command now.
Xmodem sectors/kbytes sent: 0/ 0k
Retry 0: Timeout on sector ACK
Retry 0: Timeout on sector ACK

Hello Adnan, 

Thank you.

The bootmode may not have DDR dependency.

Did you try the bootmode on multiple boards.

Can you probe the bootmode configuration.

Are all the supplies as expected and is the PORz_out high.

Regards,

Sreenivasa

Hello Sreenivasa,

Would you like us to probe all 16 boot mode configuration pins? If so, we can check them and update you on the status tomorrow along with PORz_out.

Also, just to confirm — are you saying that the board should be able to boot in UART boot mode even without any DDR configuration or even with default DDR configuration?

Lastly, we’d like to know if we should expect to see any UART output log when switching to UART boot mode before sending any binaries (i.e., without having flashed anything yet).

Regards,
Adnan

Hello Adnan, 

You had previously configured the UART boot. I assume it was primary boot, please check only relevant pins.

shaik adnan pasha said:

Lastly, we’d like to know if we should expect to see any UART output log when switching to UART boot mode before sending any binaries (i.e., without having flashed anything yet).

Here is the bootlog you were seeing before

Since the issue is seen after some time, it is important to probe the supplies.

Please also probe the clock output 25MHz.

You can probe during reset on WKUP_CLKOUT0 pin

Regards,

Sreenivasa

Hello Sreenivasa,

We have verified that the 25 MHz clock is present and stable.

However, we are encountering issues with the following reset signals:

• RESETSTATz
• PORz
• MCU_PORz
  All of these signals remain LOW.

Specifically, the MCU_PORz signal is always low. According to the schematic (Page 16), this signal is driven through an AND gate that combines the PMIC_POWERGOOD and an external switch signal. The PMIC_POWERGOOD line is pulled up to 3.3V_SYS, but the PMIC appears to be actively pulling it low.

We have confirmed that all power rails from the PMIC and buck converters are at correct voltage levels.

Also let us know what should be the status of MCU_INTn pin in PMIC.

Could you please review the PMIC-related section of the schematic and advise on any possible root causes or debug steps to help resolve this issue?

Best Regards,
Adnan

Hello Adnan, 

Please probe the supplies with an oscilloscope and confirm if the PMIC is starting for some reason.

Are you checking on multiple boards.

Regards,

Sreenivasa

Hello Sreenivasa,

We are checking on single board.

Below are some more observations along with the previous observations shared.

Regarding probing which supplies do you want us to probe input supply or output voltage of pmic?


Regards,
Adnan

Hello Adnan, 

This was a working board that has issues now.

Is there some external input that is permanently connected to any of the IOs before the board power-up.

Regards,

Sreenivasa

Hello Adnan, 

I see provision to isolate the 3V3 Main and the core voltage output from the PMIC.

Can you measure the current by removing the series resistor or adding a shunt resistor instead of 0R 

Please isolate the core voltage and try powering the PMIC with core voltage isolated.

Regards,

Sreenivasa

Hello Adnan, 

Checking if you are able to make some progress?

Regards,

Sreenivasa

Hello Sreenivasa,

Thank you for your inputs.

We managed to boot one board after modifying the DDR configurations.

Regards,
Adnan

Hello Adnan, 

Thank you.

Regards,

Sreenivasa