L3+ heat question

Question for someone. I live in Montana and most of us don't have AC here. My L3's are in the garage. We have several weeks where it does get in the 90's here, cools back in the evenings to the 40s or 50s. I was wondering at what point you shut your units down if you feel its getting to hot. I realize they have an auto shut down but would really prefer they not do that on me.

Thanks in advance.

mopst of us have cooling inferstructure in place so we dont have to turn it off, 75C is warning zone

Wow, that's 167F, it doesn't get near that hot here. I was thinking 26 Celsius - 80F was hot and taxing them.

Is that the board temp or the chip temp.


It's NOT ambiant temp. Just because it's only 90'F outside doesn't mean the chips or the boards in the L3+s arent approaching 120'+

Thanks for the clarification Ziggy. It will be 80F ambient this weekend. I will monitor.

even 90dF air moving over the chips will cool below 90F

Fluid and thermo Dynamics

cc4506 wrote:Question for someone. I live in Montana and most of us don't have AC here. My L3's are in the garage. We have several weeks where it does get in the 90's here, cools back in the evenings to the 40s or 50s. I was wondering at what point you shut your units down if you feel its getting to hot. I realize they have an auto shut down but would really prefer they not do that on me.

Thanks in advance.

i have my miners located in a garage as well. we have had numerous days of 85 degree heat already. the warmest any of the L3's have gotten has been 73 Chip Temp, which is approaching the warning zone, as Gregory points out. i have a three room space constructed... entry, mining and heat disipation. as such, the machines are not breathing their own generated heat. another factor is the sun and how it traverses the house. the sun rises in the back of the house and doesn't really shine on the garage door/driveway at the garage door until around 2pm or so. when it was 85 and the sun was still not shining on the front, the chips were running about 5/7 degrees Celcius cooler than when the sun was shining on the front with the same temperature.

the S9's are a different story. they naturally run hotter and you can run them comfortably into the low 90's (chip temp). however when they get that hot, the fans run full blast (6k) and they have a bit of a howl to them, which depending upon where one lives, may be an issue.

GregoryGHarding wrote:even 90dF air moving over the chips will cool below 90F

Fluid and thermo Dynamics

Even when the chip or board temps are above 90F?

I get the concept of windchill, the removing of moisture and cooling effect, but the same way plants are immune to the effect, aren't computer chips(hopefully moisture free) also immune to cooling of airflow via evaporation.

So cooling miners must be convection based. Air isn't a very good heat transfer, which is why all the boards in the asics have little tiny heat sinks soldered to them. Im not seeing how 90'F air can be pulled across 170'F chips and make them go under 90'F
Seems like its asking an aweful lot of cooling from the warm air.

If there was no new heat applied, I could see them cooling off to 90 overtime.
Moving air with a fan does not change the temperature of the air that is being moved.
So even with no new heat applied how would the chips or board go under the ambient temp?

I have a room thermometer. Turning on a fan and blowing it at the thermometer doesn't change the temperature in my room.

Thanks for the help. Eventually that side of the garage will be a "clean room" and add cooling. But because of how the garage is located its not as critical as you might think - except for our short summer. Learned a bit about thermo dynamics along the way as well. Was surprised to learn its more important to move the warm air off, (and not have them rebreathe it), then it is to add cool air.

Thanks everyone.

CSZiggy wrote:

GregoryGHarding wrote:even 90dF air moving over the chips will cool below 90F

Fluid and thermo Dynamics

Even when the chip or board temps are above 90F?

I get the concept of windchill, the removing of moisture and cooling effect, but the same way plants are immune to the effect, aren't computer chips(hopefully moisture free) also immune to cooling of airflow via evaporation.

So cooling miners must be convection based. Air isn't a very good heat transfer, which is why all the boards in the asics have little tiny heat sinks soldered to them. Im not seeing how 90'F air can be pulled across 170'F chips and make them go under 90'F
Seems like its asking an aweful lot of cooling from the warm air.

If there was no new heat applied, I could see them cooling off to 90 overtime.
Moving air with a fan does not change the temperature of the air that is being moved.
So even with no new heat applied how would the chips or board go under the ambient temp?

I have a room thermometer. Turning on a fan and blowing it at the thermometer doesn't change the temperature in my room.

the temps in the UI are in Celsius, right? I think you are crossing the two calculation units.

you said, "Even when the chip or board temps are above 90F?" and "Im not seeing how 90'F air can be pulled across 170'F chips and make them go under 90'F". this is wrong but perhaps you meant, 'Even when the chip or board temps are above 90C?', and ' Im not seeing how 90'F air can be pulled across 170'F chips and make them go under 90'C'. yes, 90 degrees Fahrenheit is 104 degrees Fahrenheit cooler than 90 degrees Celsius.

also, if air didn't cool the chips, why are fans installed on ASICs? not only do the heat sinks dissipate the heat, but they also are cooled by the driven air which allows them to take on more heat to dissipate. the ASIC's utilize not only convection but conduction cooling.