Quote from: davmp on January 07, 2014, 05:05:18 PM

Quote from: flemeister on January 07, 2014, 04:44:17 PM

Quote from: davmp on January 07, 2014, 01:58:22 PM

I'm trying 53:2:950 right now

Your board may need 950mV to get rid of most of those HW errors. You could also try int:53:2:850. I get lower a higher hash rate and lower HW errors (virtually none) on my boards with 54:2:850 than I do at default settings (40:1:850).

And definitely get another fan cooling the line regulator heatsink as well as those teeny little heatsinks, especially if you want to use 950mV.

I *am* using 950mV. I said so twice already. :-)

I tried int:53:2:850 earlier, the error rate was higher.

I'm using a large floor fan blowing over the whole board. Pretty sure it is cooling the whole board pretty well. Without a temperature sensor gun, is there a way to tell if components are getting hot and causing the HW errors?

I'm getting a HW error rate of about 3% at int:53:2:950. The board is cooled by 3 12cm fans - 1 at the heatsink side, 1 at the component side and one from the side with the airflow along the heatsink fins. The reported temp is around 32C. The long term avg hash rate is about 22.4 GH/sec WU 314/m. It does take the board a while to achieve steady state in hashing, take into account the pool applying var diff. I'm running the drillbit CGMiner 3.8.5 on win7/64. The board f/w is at the latest 'test firmware'.

Cheers