Quote from: Phoenix1969 on December 25, 2013, 07:01:58 PM

Quote from: Bogart on December 25, 2013, 06:57:08 PM

Quote from: Phoenix1969 on December 25, 2013, 06:49:46 PM

Quote from: Bogart on December 25, 2013, 06:43:39 PM

Quote from: sickpig on December 25, 2013, 06:33:42 PM

Quote from: Bogart on December 25, 2013, 06:31:22 PM

My November Jupiter hashed at 670GH/s when it was working perfectly. Last week something went wrong in it and I've got a die off and it gets 630GH/s now.

did you try to increase the voltage for switched off die?

There is no voltage adjustment available on the November units, so no.

and you are running kncminer-0.99.1-E_November.bin ?

https://docs.google.com/file/d/0BxXlEOb-bf3AVUFRSHNLRHVudGc

I am. I was running 0.99-E when the failure occurred. 0.99.1-E did not fix the issue.

Buggah! Thought I'd ask... Just to check.... you also went thru & did the pci plug wiggles, tried the Temp-Wake method, tried pressing on the bar-clamp...(on that asic)..and tried re-flashing?... all the norms.....?

Son-of-a-Josh. I tried some of those things (swapping the power and data cables around) and now I really killed it to where it won't boot. Sad

The control board LEDs don't come on at all now, though all the fans run and the ethernet port link LEDs come on. The BBB shows a single blue LED but never anything else.

I tried disconnecting all the modules' data cables, but the behavior did not change.

I think I may have the corrupt flash storage problem. I guess I'll hunt around for an SD card now and try to boot off that.

Update: I tried holding down the small button on the front of the control board while powering the unit on, and it came to life now and booted normally. I don't know if the button press did it, or if it was just luck. I'm afraid to touch it again now.

My cable swapping did not solve the dead die problem.