tx1 boot failure with self-made carrier board


I made a tx1 carrier board myself.and When I use my self-made board with tx1 core boad, the jetson tx1 can not boot.

I measured the voltage, included Vin 19V, 5v_sys, 3v3_sys, 3v3_slip and the 12v_sip, and all those voltage are correct. Also I measured the boot pin voltage at initial time, and those are correct.
But I found that the power monitor I2C bus is about 2.2v, not 3.3v.-- when use the nv carrier board, the monitor I2C bus is 2.2v for a while and later change to 3.3v, this is a diffierent.

Below is the uart log, the tx1 stopped at the final and never change latter.

[0000.759] [TegraBoot] (version 24.00.2015.42-mobile-ec3b827e)
[0000.764] Processing in cold boot mode Bootloader 2
[0000.769] A02 Bootrom Patch rev = 63
[0000.772] Power-up reason: on button
[0000.776] No Battery Present
[0000.778] RamCode = 0
[0000.781] Platform has Ddr4 type ram
[0000.784] max77620 disabling SD1 Remote Sense
[0000.788] Setting Ddr voltage to 1125mv
[0000.792] Serial Number of Pmic Max77663: 0x1a04af
[0000.800] Entering ramdump check
[0000.803] Get RamDumpCarveOut = 0x0
[0000.806] RamDumpCarveOut=0x0, RamDumperFlag=0xe59ff3f8
[0000.811] Last reboot was clean, booting normally!
[0000.816] Sdram initialization is successful
[0000.820] SecureOs Carveout Base=0xff800000 Size=0x00800000
[0000.826] GSC1 Carveout Base=0xff700000 Size=0x00100000
[0000.831] GSC2 Carveout Base=0xff600000 Size=0x00100000
[0000.836] GSC3 Carveout Base=0xff500000 Size=0x00100000
[0000.841] GSC4 Carveout Base=0
xff400000 Size=0x00100000
.846] GSC5 Carveout Base=0xff300000
[0000.851] Bpm
pFw Carveout Base=0xff2c0000 S

[0000.857] Lp0 C
arveout Base=
0xff2bf000 Size=0x

[0000.872] RamDump Carveout Base=0xff23f000 Size=0x00080000
[0000.877] Platform-DebugCarveout: 0
[0000.881] Nck Carveout Base=0xff03f000 Size=0x00200000
[0000.886] Non secure mode. Disable rollback prevention
[0000.890] AOTAG Init Done

[0000.938] Using GPT Primary to query partitions
[0000.944] Loading Tboot-CPU binary

[0000.993] Verifying bootloader in OdmNonSecureSBK mode
[0001.003] Bootloader load address is 0xa0000000, entry address is 0xa0000258
[0001.010] Bootloader downloaded successfully.
[0001.014] Downloaded Tboot-CPU binary to 0xa0000258
[0001.019] MAX77620_GPIO1 Configured.
[0001.022] MAX77620_GPIO5 Configured.
[0001.026] CPU power rail is up
[0001.028] CPU clock enabled
[0001.032] Performing RAM repair

Can anyone help me analysis why this happened? thanks very much!

Did you see any abnormal status on VDD_3V3_SYS, the pull-up source of power monitor I2C lines? If not, seems some devices drive the I2C lines to lower than 3.3V, you might need to check all devices on your I2C even by removing them one by one…

Hi Trumany,

Thanks for you relay.

And I don’t think I2C voltage is the reason, because the nv carrier board also make the voltage 2.2v for a while, the difference is in my case the voltage doesn’t rise up finally.-- I think this is the result of boot failure, not reason.(and the i2c pullup voltage is not VDD_3V3_SYS on carrier board but from tx1 core board).

Compare with the nv correct boot message, I know that after the “Performing RAM repair”, should be something about warm reset, but I haven’t any idea about that.

Finally I know what’s the reason.

I stuff the r52 on my board, which is left nc on nvidia board so the jtag_trst pin is pull up.

When I remove r52, the system boot correctly.

Good to hear that, R52 could cause that, it is really a difficult to be discovered issue.