Quote from: SebastianJu on October 22, 2013, 11:11:11 AM

Im a bit puzzled now. I host some miners for another user. 3 bbxx + 1 bbx in one stack. It runs at 26.5GH/s, thats 7.57GH/s for one bbxx. Not much. I previously had 5 miners and they ran at 43.75GH/s, which is 8.75GH/s per bbxx.

Now i got 10 more miners that run in another stack. Its 10 bbxx. And they only run at 64GH/s. Thats 6.4GH/s per bbxx only.

I dont know whats wrong. All run in one cgminer-instance, so two miners are recognized. All miner are blinking with all colors.

I ran my old 5 bbxx, that i dont have anymore, at the following settings:

--avalon-options 115200:10:10:28:450 --bitburner-voltage 1334 --avalon-temp 30 --queue 6

I ran the 3.5 bb at

--avalon-options 115200:8:10:28:450 --bitburner-voltage 1334 --avalon-temp 30 --queue 6

and the 13.5 now at

--avalon-options 115200:28:10:28:450 --bitburner-voltage 1334 --avalon-temp 30 --queue 6

I found that the second value in options should not be too low, but it doesnt matter if its too high. Since i dont know exactly how to setup for 2 stacking towers i simply used 28.

But i dont really get whats the problem. Why should my old miners run faster and the newer ones slower? queue 10 didnt change anything too. HW is under 1%.

7.5 miners are powered with a psu from burnin, the other 6 with an Seasonic Platinum SS-1000XP. The workload of the raspi is at 7%. Could it be differences in pools used? I used bitparking.com with my own miners and now bitminter.com is used. All boards are on firmware 1.0.2.

Do i really have to start testing and tweaking everything from the start because every setup needs its own settings? I mean it cant be normal that 10 miners only bring 73% of the hashrate of 5 miners got before.

cgminer didnt change in an unfortunate way right? 1334mV and 450MHz i used before can be used now too. In older cgminer-versions i had to manually edit cgminer sourcecode.

Someone knows?

Why don´t you use the --avalon-auto ?
So you don´t have to look the whole day at your bitburners. cgminer runs every cluster at the optimum.