0 sats \ 4 replies \ @cascdr 24 Mar \ on: Mining Bitcoin with Off-grid Solar: Experiences So Far DIY
The amount of power is partially based on intensity (which you don't control) and partially based on angle of incidence (which you only control in some scenarios). 90 degree incidence is ideal (it is a cosine function).
Some people build robotic systems that actually track the sun but they are complex and not cheap. I've dabbled in this but it'd be interesting to see if there's different strategies on optimizing your output more.
I had not thought about just cutting out the AC and going straight to DC. I am curious if you have any data on about how much efficiency you saved.
I've been working on improving the angle of incidence. Results so far are greatly improved. I'll share more here when I have more data.
I don't yet have data on efficiency gains from going direct-dc, but Dirty from 100acresranch says he generally sees between 16-25 j/Th on a jPro. Again, I'll share what I'm seeing when I get the panel angle optimized.
reply
says he generally sees between 16-25 j/Th on a jPro
What does this mean in the overall picture of efficiency? I'm an EE but haven't really played with mining just yet.
Even really good AC to DC converters will lose you 8-10% and state of the art is probably like 3-5% under perfect conditions. I wonder if commercial miners will get on this tip and invest in DC microgrids
reply
Out of the box the jPro advertises 30 j/Th, but I see folks using aftermarket firmware getting 26, maybe even a bit less, with a jPro. So sub 20 j/Th is a significant improvement.
Building out a microgrid is pretty capital intensive, so I imagine miners will continue trying to find places where cheap power is stranded that doesn't require as much capex to get off the ground. Although adding this to existing solar farms could be a great way to use electricity when there's excess. I'd guess it's already being done.