I was fighting high temps on my #Bitaxe Ultras (1366 ASIC), so I used some thermal paste from Noctua (the fan people) and the temp dropped by 10+ degrees C. I was previously underclocking to get it to 63C. Now I'm running defaults and getting 51C. That's measured by the temp sensor on the board behind the 1366. There was plenty of paste on there that looked clean and well-applied, but it just didn't perform as well as the Noctua stuff.

Now, interestingly, the thermal camera shows me some things I didn't consider.

First, the power bricks are hot AF. It's probably worth getting something more efficient rather than the cheapest 5V supply mankind can produce.

Then I notice that Q1 and Q2 MOSFETs are also hot AF. They're heating themselves, the board, and everything around them. The printed standing case for the Bitaxe doesn't allow for any airflow back here, so I may adjust that. Maybe it doesn't matter, maybe it does. Right now they're the hottest thing on the board.

I also have on hand 3 boards from d-central.tech to compare.

The first has spade connectors on the board and came with a power supply putting out a hair over 5V. This one seems to run the coolest overall. The temp sensor on this board version is nowhere near the ASIC, so it says something like 32C all the time. If you have this board, don't be fooled and don't use automatic fan control.

The third photo shows some heating around the DC plug. I'm assuming this is from a sub-optimal connection dealing with 2.4-ish amps running through it all the time. I'll probably replace that connector with something better or wire directly to the board. This board and its twin both have 5V supplies delivering more like 4.8V at the connector, and they run hotter in general. Hash rate between the 3 boards is close enough that I can't tell the difference.

Now, none of that is probably earth-shattering news, but it gave me something to do while I drink my #coffee.

#coffeechain #bitcoin #mining #plebmining #nostr #grownostr #plebchain