We did some testing in climate chamber and we noticed quite high non-linearity with the temperature (we get much higher delta T at below zero temperatures than plus temperatures)
=> I’m wondering if you can share more information about the technology of the heatpipe used. Namely what is the medium in the heatpipe?
You wrote above “There is water in the heat pipe”, is it really water or some other medium (ie. methanol based)? According to info I found the water would not work under some 25C…
Can you share the logs during tests so we can monitor the temperature, as the BPMP will raise the voltage below 25C, which may account for the symptoms.
Attached is the table with measured values from Xavier internal sensor for 3 test cases after temperature stabilization - same setup and same power load profile (ID3 30W) tested at 3 ambient temperatures (-40C, 25C, 70C).
You can see that behavior is quite linear for 25C and 70C test case (delta to ambient is similar), but for -40C case the delta to ambient is 2.5-3x bigger, so it is very high non-linearity that indicates that the heatpipe at low temperatures works much much worse => ie. my speculation that the medium in the heat pipe does not work at low temperatures.
The first thing to note is that the module is not designed to operate at -40C. (Temperature Range: -25C – 80C)
The second observation is that the system does not appear to be exceeding its operating limits.
There will be part to part variance based on the where it sits in the process spread, please note that below silicon temperature of 25C, it operates at elevated voltages.
Yes, I know. We intentionally tested outside the spec to see if there is any margin in your spec. And I was also bit curious what is behind that -25C limit. Is there any technical reason for that or you simply did not tested it?
The datasheet is bit interesting - in one place saying -25C operational, -25C storage but few pages further you do 72hours storage tests at -40C. Also you have recent announcement about planned industrial grade version. => All this gives reasonable prospects that there is good chance that -40C will be ok.
About exceeding the operational limits - you’re completely right. Test at low temperature does not exceed any temperature limits, so I don’t have any concerns about this.
My concerns were this observed non-linear thermal behavior. We need to understand if this is ok and can be explained by some regular mechanisms that are part of normal operation of the Xavier SoM or if this is indication of some failure. …simply to be sure that the product that we’re designing will be working ok.
As you can see in data sheet, both operational and storage temp range are -25C - 80C, the -40C storage 72hr test is non-operational. The limit is from the request of some components in module which should not be violated and no need to test outside the spec as it has been validated already. The industrial module will use the corresponding components to expand the temp range. Did you do test in -25C? Basically if the system does not exceed limits, there should be no concern even such non-linear observed.
