I used the software that @jstefanop has put out here on 2 L3+ units last night! Excellent!
I used the WinSCP to do it all - I have Putty on my system, so not sure if it picked up on that or if it has it embedded in the software or install, but once I established a connection to the miner with WinSCP, I was able to open a Putty window from a button in WinSCP that opened right to the miner already connected to - just entered password and followed the directions.
Anyway - I will write more on it later but the end result was that I got them both running at the lowest power (0xfe) and was able to clock them both up to 394MHz (~515MH) without any issue. I recorded temps at each step and after clocking up to 394, the temps were back to about the same as they were running at 0x80 (normal voltage factory setting). So I ended up at about 15% less power use with about a 2% increase in hash rate with temps staying in the same ballpark as they started at. I was getting no more HW errors that before with this config - I was targeting lowering power use, then improving hash rate - while maintaining similar temps. I'm sure I could have increased hash rates by more than 2%, but I didn't want the additional heat. My miners are running overall hotter right now than I want - getting ready to move them to a cooler place with better airflow, but until then, I need to keep the temps in check.
Very Good!
Now - I have two L3++ units as well. They are running at 450MHz from Bitmain and running generally in the 590-600MH range. So my question on these - is there already any kind of automatic voltage adjustments going on with these L3++ units? Their MH number seem to be fluctuating much more and much more constantly - like something is being adjusted constantly. My L3+ units stay around 502MH to 503MH very consistently (well, now 514MH to 517MH
) - but these L3++ units are bouncing around probably 570MH to 615MH constantly. I mean I just site there watching to numbers fluctuating none stop (using Awesome Miner). So I just wanted to see what you guys think - I know I saw someone else had used this on the L3++, but just wondering if it might interfere with Bitmain's settings if they may be messing with an auto voltage system design as well.
Also, one of my L3++ units has 1 hash board that is running a little more than, but pretty close to, 1 HW per minute with factory settings. Is this worth contacting Bitmain over? I'm wondering if dropping the voltage on this unit with this software from @jstefanop will potentially ease that number down - less power - less heat - maybe help improve the HW rate?
And lastly - I had a thought/concern about the lower power settings. When running electrical - it is always better to use heavier wire as it is more efficient at carrying the electricity: less power loss due to resistance (generally lost in the form of heat) and improves the watts used by the connected equipment, etc... When doing electrical runs that are too long for the gauge wire used, you frequently hear about how that can damage equipment due to the power reaching the connected devices not being strong enough - too much lost in the transmission and the overly low power somehow causing damage to the connected devices. Is this is a concern here? If we lower the voltage down to lowest setting - is this somehow stressing the miner's components in a way that might cause them to fail earlier? Should we look at lowering the voltage some, but not to the lowest settings? I don't know and am truly asking. What are your thoughts on this? I am very happy with the idea of lowering heat, power costs, and/or even getting improved hash rates - but...
Thanks!
