T17/S17 malfunction: cases, solutions, remedies, RMA history

[unisahmed]

I've got the same issue.
I have 2 T17 42T, one is running with 1 board and the other with 2 boards.
Can i swap one board to the other machine?

[1714989095]

1714989095

[1714989095]

1714989095

[1714989095]

1714989095

[mikeywith]

I've got the same issue.
I have 2 T17 42T, one is running with 1 board and the other with 2 boards.
Can i swap one board to the other machine?

If you use the stock firmware you will have to try, there is a combination of voltages stored in the EPROM of each hash board, some will match well some won't, if you use custom firmware, you can simply mix and match as they don't care about the voltage readings.

However, why do you want to do that in the first place? unless you are going to sell the other PSU and Control board, or you are running out of space, it's better to keep 1 working board in 1 miner and 2 in the other, the 3 working boards will run a lot cooler sitting next to dead hash boards.

[unisahmed]

Hi Mike, thanks for the reply.

The very first day i ran these machines they were working with all 30 chips and 3 boards. After 30 minutes one Chain failed and then 2nd failed. so i decided to test the 2nd machine and the same shit repeated but only one chain failed on this machine.  I moved the running board on to the 2nd machine and ran the stock firmware and surprisingly its been running pretty well for the past 4 days and 17hrs. I'm getting 42T with the stock firmware.

As for the 1st machine, ill try to find someone locally and get it fixed.

https://www.linkpicture.com/q/T17-42T-4Days-Mining.jpg

[BitMaxz]

Hi Mike, thanks for the reply.

The very first day i ran these machines they were working with all 30 chips and 3 boards. After 30 minutes one Chain failed and then 2nd failed. so i decided to test the 2nd machine and the same shit repeated but only one chain failed on this machine.  I moved the running board on to the 2nd machine and ran the stock firmware and surprisingly its been running pretty well for the past 4 days and 17hrs. I'm getting 42T with the stock firmware.

As for the 1st machine, ill try to find someone locally and get it fixed.

If you can take the kernel logs from your first machine with the 2nd hashboard(Failed) maybe we can see some other reason under your kernel logs why the 2nd hashboard chain fail.

You can post the whole kernel logs here and make sure to paste it inside the insert code tag(A # button when editing/posting) or you can use pastebin.com and share the URL here.
Sometimes the issue is due to sensors or overheating and kernel logs will explain everything that happens in your machine.

[mikeywith]

I moved the running board on to the 2nd machine and ran the stock firmware and surprisingly its been running pretty well for the past 4 days and 17hrs.

Like I explained in the first post, you could get lucky, I don't know how many combinations are there "I remember 3 of them for S9k", they could be the same", so every board you mix has a 1/3 chance, so more often than not, it doesn't work, but seems like you are pretty lucky, maybe mine to a solo pool and hit a whole block with that gear?


I would also be interested in the new kernel log for the bad machine.

[wndsnb]

The voltage setting stored in the eeproms (for the current mode if a s17) needs to be the same on all hashboards in a miner for stock firmware to work. If you have logs from the miners with the boards you want to mix, you can tell in advance if they will work together. You just need to find the following section in the logs and check that the voltage setting is the same in the two miners you want to mix and match boards from/to.

Code:

022-05-17 22:44:27 uart.c:71:set_baud: set fpga_baud to 115200
2022-05-17 22:44:38 driver-btm-api.c:915:check_asic_number_with_power_on: Chain[0]: find 48 asic, times 0
2022-05-17 22:44:49 driver-btm-api.c:915:check_asic_number_with_power_on: Chain[1]: find 48 asic, times 0
2022-05-17 22:45:00 driver-btm-api.c:915:check_asic_number_with_power_on: Chain[2]: find 48 asic, times 0
2022-05-17 22:45:05 driver-btm-api.c:345:set_order_clock: chain[0]: set order clock, stragegy 3
2022-05-17 22:45:05 driver-btm-api.c:345:set_order_clock: chain[1]: set order clock, stragegy 3
2022-05-17 22:45:05 driver-btm-api.c:345:set_order_clock: chain[2]: set order clock, stragegy 3
2022-05-17 22:45:05 driver-hash-chip.c:502:set_clock_delay_control: core_data = 0x34
2022-05-17 22:45:05 uart.c:71:set_baud: set fpga_baud to 3000000
2022-05-17 22:45:05 driver-btm-api.c:1574:check_clock_counter: freq 50 clock_counter_limit 6
2022-05-17 22:45:07 uart.c:71:set_baud: set fpga_baud to 115200
2022-05-17 22:45:07 voltage[0] = 1780
2022-05-17 22:45:07 voltage[1] = 1780
2022-05-17 22:45:07 voltage[2] = 1780
2022-05-17 22:45:07 power_api.c:219:set_working_voltage_raw: working_voltage_raw = 1780