flash failure on Kasli_SOC

sum_s

I got flash failure on kasli_soc board.
I used https://nixbld.m-labs.hk/build/106791 's sd boot file for flashing
but I got crashed message through Putty when booting the board.
here is that pic

And also can not connect default ip ( : 192.168.1.56 ) then run blink_forever.py or jtag programmig

Here is pic of my board connections

If failures are due to my faulty borad,
Is there any way I can check the board is fully fuctional?

sb10q

Another anomaly is the FPGA DONE LED is off. The ARTIQ/SZL firmware is supposed to configure the FPGA and that LED would turn on.
Did you check all power supply voltages?
Did you set the correct baud rate? 115200-8-N-1.
You don't need PuTTY, you can simply do:

stty 115200 < /dev/ttyUSB2
cat /dev/ttyUSB2

Unplug and replug the USB cable to reset the USB UART, PuTTY may have selected some options you do not want. Also make sure programs like ModemManager are not interfering.
If you cannot make progress with the serial port, then you need to use a JTAG debugger (OpenOCD and such).

sum_s

I tested all test point voltages and found wrong outputs from TI's LDO outputs.
I checked schematic file then might find a problem.

Comment says "While using configuration pins, unmount FB resistors and ...."
But my board fulled with FBs and 0R configuration jumpers.
I guess that's why LDO outputs are excessive.
like this

So, l think there is 2 ways to fix it

unmount SNS 0R jumpers and all configuration pins jumpers
unmount FB resistors, then unmount selective comfuguration pins consider LDO outputs

Could you teach me which one is Kasli_SOC's reference?
Or recommend me what to do?

sb10q

I think either option would be fine: SNS mounted + configuration jumpers mounted correctly for each voltage, OR FB mounted.
But the excessive voltages might have caused some damage already.

sb10q

It's probably simpler for you to remove SNS and all configuration jumpers.

sum_s

After works on resistors, I can get right voltage outputs but still fpga done led is not on.
I think there might be a damage.
I guess SDRAM ( =MT41K256M16HA-125:E) or Zynq chip, maybe both could got damages.

here is what I got after artiq_flash start

Is there way I can know that Zynq chip is fuctional or not?

sb10q

artiq_flash is not the correct tool to use for Kasli-SoC. Please see the instructions in the artiq-zynq repository as I pointed out in your other forum thread.
https://forum.m-labs.hk/d/363-how-do-i-get-gatewarefirmware-for-kasli-soc/8

sum_s

The result seems scan works anyway.
Can I consider that Zynq chip is still alive?
Or that result is nothing matter?

sb10q

Seems that at least part of it is still alive, yes.
Please follow the instructions given here and in your other thread.

sum_s

Still don't work after both repository's instruction SD booting way and jtag programing way.

`Here is what I got after trying local_run jtag

Now I'm plan to replace mounted ddr memory chip most vulnerable to excessive voltage I guess

sb10q

Again, please follow the instructions, in particular those that pertain to the serial console. A DDR memory failure would have clear symptoms there.

sb10q

And a DDR failure would not cause garbled console output, at least not during the initial phases of boot.

sum_s

Still get broken respond from serial

sb10q

Wrong clock?