Here is something for the electronically enabled:
Here is an obscure thread regarding the mod of a power supply to adjust the bus voltages. Discussed in this thread is the fact that some power supplies have onboard pots to tweak the 3.3, 5 and 12 DC outputs
http://hardforum.com/showthread.php?t=765175Now opening up an ATX power supply is a potentially hazardous endeavor and messing about inside could kill you, so the responsibility is your own.
The big capacitors must be discharged WITH THE UNIT OFF AND UNPLUGGED from everything.
If there are multiple pots, you have to figure out which one controls the +12V output. You will also have to figure out which one increases or decreases the 12V output.
This will upend power supply selection somewhat if there are particular brands that have a board level pot to adjust the +12V output.
The thread has useful safety information which should be consulted.
The thread is 11 years old. The safety info is okay but most likely the psu's mentioned are not for sale or they have been redesigned.
I have yet to find a 2014 ,2015, 2016 model atx with a pot for sale in the USA.
Someone linked one but it was not available in the usa. If you have 120 volt power you just won't be able to overclock.
There are more factors with the avalon it has 4 pcie jacks you need to supply 275-300 watts a jack to get close to 3900 ghs.
This means 16 or even 15 gauge cables not many atx psu's give you 16 gauge cables.
Right now it has been hot again in NJ it is 55f my garage is 87f my bedroom is 81f so I had to lower the 4000 watt dual psu from 12.55 to 12.23
I get 3500 for each avalon 6 and I get strange volt readings.
12.2 and 12.0 power comes from same psu's with same length cables
I am pulling 2300 watts to get 7150gh about .32 watts a gh I have a lot of overhead on the psu's
say 2300/4000 = 57.5% used if you are pulling 1100 watts on your 2 psu's you are 1100/1500 = .733% and your psu may be 750 watts at 240 volts and only 700 watts at 120 volts so you could be 1100/1400 = .7857%
that is a safe number but may result in volts sag.
can you measure your volts? and not the software one since I really do not trust the software readings mine should be = and they are not.