Bitcoin Forum
March 17, 2026, 05:59:06 PM *
News: Latest Bitcoin Core release: 30.2 [Torrent]
 
   Home   Help Search Login Register More  
Pages: « 1 [2]  All
  Print  
Author Topic: Bitaxe Longevity: A QHSE Perspective on Underclocking for 24/7 Solo Mining  (Read 208 times)
Nexus9090
Full Member
***
Offline Offline

Activity: 534
Merit: 181


So many numbers and so little time


View Profile
March 16, 2026, 10:19:12 PM
Merited by philipma1957 (1)
 #21

-SNIP-

Following the interesting feedback from Nexus9090 regarding thermal management, I’ve been analyzing the Bitaxe setup through the lens of my professional background in QHSE (Quality, Health, Safety, and Environment).

While many focus solely on hashrate, we often overlook the Mean Time Between Failures (MTBF) of these small DIY boards. Here is a brief QHSE perspective on why the "Thermal Paste" upgrade is a safety and quality necessity, not just a performance tweak...

My recommendation for the community:
Always monitor your VRM temps alongside your ASIC. If you repaste, don't just chase the 11.25TH/s; use that thermal headroom to lower the overall stress on the board.

Just a little info.

Most semiconductors have a guaranteed working temperature of 125°C, but this is typically package and PCB layout dependent.

The VRM's that have sync FET switching on the NerdOAxe are rated to 125°C each guaranteed recommended max working temperature. If you're interested the datasheet for the sync rectifiers can be found here https://www.ti.com/product/CSD95472Q5MC

The ASIC's (BM1370) on the other hand given the FR4 substrate packaging are more likely to have an upper limit of around 90°C, this is just because of the package and its poor thermal characteristics.

In my own tests with the (BM1397) ASIC a predecessor of the BM1370, I found that they became unreliable once you exceeded 70°C and very often would stop hashing all together beyond that temperature.

So for the most part, the VRM power stages definitely have plenty of headroom.

The ASIC on the other hand, less so.

The problem here is the packaging that BitMain chose to use for their ASICs, its cheap and prone to moisture ingress MSL3 would be my guess as to the package standard for moisture sensitivity. This is more of a problem at the assembly stage since moisture ingress can cause pop-corn or internal fractures that only become apparent after many thermal cycles. Given the build quality of these boards there's no guarantee that the material handling has been up to standard; I very much doubt the ASICs used are baked before assembly to drive out moisture so that'll have an impact on the devices life time.

In addition, the small exposed die area has limited thermal conductivity simply because of its size at 64mm.Sq. pushing 20Watts through that and dissipating it is hard work which is why good thermal paste is an absolute must for these devices.

In conclusion, Yes on the whole running the devices at a lower clock speed will prolong the lifetime of the devices, lower stresses throughout. But is it really that big a worry? probably not.

For me I'm not at all worried about pushing the board hard, that's what I bought it for. If it breaks I'll fix it, because I have the required skills in that department.

But for most people, keeping things steady at a lower speed is probably a safer bet.




philipma1957
Legendary
*
Online Online

Activity: 4788
Merit: 11619


'The right to privacy matters'


View Profile WWW
Today at 01:00:28 AM
 #22

-SNIP-

Following the interesting feedback from Nexus9090 regarding thermal management, I’ve been analyzing the Bitaxe setup through the lens of my professional background in QHSE (Quality, Health, Safety, and Environment).

While many focus solely on hashrate, we often overlook the Mean Time Between Failures (MTBF) of these small DIY boards. Here is a brief QHSE perspective on why the "Thermal Paste" upgrade is a safety and quality necessity, not just a performance tweak...

My recommendation for the community:
Always monitor your VRM temps alongside your ASIC. If you repaste, don't just chase the 11.25TH/s; use that thermal headroom to lower the overall stress on the board.

Just a little info.

Most semiconductors have a guaranteed working temperature of 125°C, but this is typically package and PCB layout dependent.

The VRM's that have sync FET switching on the NerdOAxe are rated to 125°C each guaranteed recommended max working temperature. If you're interested the datasheet for the sync rectifiers can be found here https://www.ti.com/product/CSD95472Q5MC

The ASIC's (BM1370) on the other hand given the FR4 substrate packaging are more likely to have an upper limit of around 90°C, this is just because of the package and its poor thermal characteristics.

In my own tests with the (BM1397) ASIC a predecessor of the BM1370, I found that they became unreliable once you exceeded 70°C and very often would stop hashing all together beyond that temperature.

So for the most part, the VRM power stages definitely have plenty of headroom.

The ASIC on the other hand, less so.

The problem here is the packaging that BitMain chose to use for their ASICs, its cheap and prone to moisture ingress MSL3 would be my guess as to the package standard for moisture sensitivity. This is more of a problem at the assembly stage since moisture ingress can cause pop-corn or internal fractures that only become apparent after many thermal cycles. Given the build quality of these boards there's no guarantee that the material handling has been up to standard; I very much doubt the ASICs used are baked before assembly to drive out moisture so that'll have an impact on the devices life time.

In addition, the small exposed die area has limited thermal conductivity simply because of its size at 64mm.Sq. pushing 20Watts through that and dissipating it is hard work which is why good thermal paste is an absolute must for these devices.

In conclusion, Yes on the whole running the devices at a lower clock speed will prolong the lifetime of the devices, lower stresses throughout. But is it really that big a worry? probably not.

For me I'm not at all worried about pushing the board hard, that's what I bought it for. If it breaks I'll fix it, because I have the required skills in that department.

But for most people, keeping things steady at a lower speed is probably a safer bet.






yeah my 69 year old eye with a lens replaced due to cataracts in both eyes means don't over clock cause you won't see it good enough to solder it.

So I under clock on a steady basis and most every mining gear I underclock does well.

Factually speaking my bitaxe is not hitting a block at 800gh or 1.1th

So I rather not buy another one cause I broke it. sidekick did a good rework and the heatsink he added is good.

▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
████████████████████████████████▀
██████████████████████████████▀██▄█
████████████████████████████▀██████
█████████████████████████▀█████████
██████████████████████▀████████████
█▄██▀▀█████████████▀███████▄▄▄█████
███▄████▀▀██████▀▀█████▄▄▀▀▀███████
█████▄▄█████▀▀█▀██████████▄████████
████████▀▀███▄███████████▄█████████
█████████▄██▀▀▀▀███▀▀██████████████
███████████▄▄█▀████▄███████████████
███████████████▄▄██████████████████

 AltairTech.io    Miners  Parts 🖰 Accessories 
_______Based in Missouri, USA._________________Your One-Stop Shop for Bitcoin Mining Solutions_____________________Mining Farm Consulting__________
.
.🛒SHOP NOW .
Nexus9090
Full Member
***
Offline Offline

Activity: 534
Merit: 181


So many numbers and so little time


View Profile
Today at 01:17:18 AM
 #23

-SNIP-

yeah my 69 year old eye with a lens replaced due to cataracts in both eyes means don't over clock cause you won't see it good enough to solder it.

So I under clock on a steady basis and most every mining gear I underclock does well.

Factually speaking my bitaxe is not hitting a block at 800gh or 1.1th

So I rather not buy another one cause I broke it. sidekick did a good rework and the heatsink he added is good.

Yeah rework on these things is not for the fainthearted or as you say, for someone who's ocular challenged.

I'm fortunate enough that I have all of the required rework tools, microscope, hot air, reflow oven, hot plate, pick and place machine etc etc, since that's part of my business.

Even so these ASIC's are tricky little buggers to solder down in rework especially on the BitMain boards reworking them is hard. I've only had to do a few S17 repairs for a friend, they're truly awful to work on.

The BitAxe Gamma is a breeze to work on in comparison.

Pages: « 1 [2]  All
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!