Quote from: musetheater on March 27, 2016, 11:13:09 PM

Hi all,

I've posted before on the Ultra Under-overclock thread by emdje, but I thought I should post here since I'm using this image on my A2 Mega. Bought it this past January, finally got everything set up and working after replacing the power supply, and it's been working smoothly for about 3 months. Tonight, however, I received one of the automated emails from Mining Rig Rentals saying the rig is offline, which shouldn't be the case. I pulled up the control interface on my computer and saw something a bit unsettling. The cgminer log kept repeating the same message which was followed 10 seconds later with a forced restart, which kept repeating until I stopped the miner.

This is the most recent display in cgminer:

Code:

[2016-03-27 18:48:45] Started cgminer 3.9.0 innosilicon-12chip-v0p5
[2016-03-27 18:48:45] Run Reset=1
[2016-03-27 18:48:45] ST MCU hardware reset start
[2016-03-27 18:48:49] SPI Speed 4000 kHz
[2016-03-27 18:48:49] ST MCU - Enable (Pre-header)
[2016-03-27 18:48:49] Run temperature 35C
[2016-03-27 18:48:49] A1 = 1200,7
[2016-03-27 18:48:49] A1 PLL Clock = 1200MHz
[2016-03-27 18:48:49] A1 = 1200,7
[2016-03-27 18:48:49] A1 PLL Clock = 1200MHz
[2016-03-27 18:48:49] A1 = 1200,7
[2016-03-27 18:48:49] A1 PLL Clock = 1200MHz
[2016-03-27 18:48:49] A1 = 1200,7
[2016-03-27 18:48:49] A1 PLL Clock = 1200MHz
[2016-03-27 18:48:49] A1 = 1200,7
[2016-03-27 18:48:49] A1 PLL Clock = 1200MHz
[2016-03-27 18:48:49] A1 = 1200,7
[2016-03-27 18:48:49] A1 PLL Clock = 1200MHz
[2016-03-27 18:48:49] AUTO GPIO CS
[2016-03-27 18:48:49] Failure(cs0)(2): missing ACK for cmd 0x02
[2016-03-27 18:48:49] ACK(cs0) timeout:cmd_POWER_ON_BCAST-0.0553s
[2016-03-27 18:48:49] SPI(cs0) no device
[2016-03-27 18:48:49] ACK(cs0) timeout:cmd_RESET_BCAST - 0.27 ms
[2016-03-27 18:48:49] Failure(cs1)(2): missing ACK for cmd 0x02
[2016-03-27 18:48:49] ACK(cs1) timeout:cmd_POWER_ON_BCAST-0.0575s
[2016-03-27 18:48:49] SPI(cs1) no device
[2016-03-27 18:48:49] ACK(cs1) timeout:cmd_RESET_BCAST - 0.27 ms
[2016-03-27 18:48:49] Failure(cs2)(2): missing ACK for cmd 0x02
[2016-03-27 18:48:49] ACK(cs2) timeout:cmd_POWER_ON_BCAST-0.0487s
[2016-03-27 18:48:49] SPI(cs2) no device
[2016-03-27 18:48:49] ACK(cs2) timeout:cmd_RESET_BCAST - 0.27 ms
[2016-03-27 18:48:49] Failure(cs3)(2): missing ACK for cmd 0x02
[2016-03-27 18:48:49] ACK(cs3) timeout:cmd_POWER_ON_BCAST-0.0499s
[2016-03-27 18:48:49] SPI(cs3) no device
[2016-03-27 18:48:49] ACK(cs3) timeout:cmd_RESET_BCAST - 0.27 ms
[2016-03-27 18:48:49] Failure(cs4)(2): missing ACK for cmd 0x02
[2016-03-27 18:48:49] ACK(cs4) timeout:cmd_POWER_ON_BCAST-0.0489s
[2016-03-27 18:48:49] SPI(cs4) no device
[2016-03-27 18:48:49] ACK(cs4) timeout:cmd_RESET_BCAST - 0.27 ms
[2016-03-27 18:48:49] Failure(cs5)(2): missing ACK for cmd 0x02
[2016-03-27 18:48:49] ACK(cs5) timeout:cmd_POWER_ON_BCAST-0.0509s
[2016-03-27 18:48:49] SPI(cs5) no device
[2016-03-27 18:48:49] ACK(cs5) timeout:cmd_RESET_BCAST - 0.27 ms
[2016-03-27 18:48:49] No any A1 board

I've seen in another thread that a possible reason to this might be the power supply or a bad SD card. Since I'm not the best at computer coding, especially Linux, I'm not sure how to tell what the problem is. The power supply shouldn't be a problem since I put a brand new EVGA Suprnova 1300W Gold in about 2 months ago. The SD card came with the miner, so I'm not sure about it's reliability, but it works enough that I can still pull up the interface.

Also, another thing to note, right when the miner stopped working and displayed the message above, the board temperatures SKYROCKETED almost instantly. During the day they average about 52*C, but within 2 minutes of the error message, the boards (0 through 5) had temps of 68*C, 59*C, 80*C, 101*C, 97*C, and 55*C, respectively. They are receiving a strong airflow and no problem with air circulation to get rid of the heat, so I'm hoping that factor isn't a bad sign, and that it's just caused by the malfunction with the image, or whatever might be wrong. Hoping to hear back very soon with a possible solution or any ideas. Thanks

I can't really see the all the hashing boards going bad, while it *could* be a problem with the image normally that results in things not booting , do the LEDs on the hashing boards just flash the whole time? Although just one part on the controller board going bad or breaking can cause the whole mess to appear dead. As for testing I guess a reflash of the image would be the easiest to do, as for more advanced testing going to be hard to do without soldering irons, cables etc...