Ambient vs computer temps...direct relation.

[Bluefront]

Well we had a little cold snap today......the ambient in my house dropped from 26/27C to22C. And my computer temps across the board dropped the same amount. CPU idle went from 41C to 36C. HD temps dropped from 34C to 29C.

I haven't adjusted the base fan speeds....but I will in a few weeks. I've never seen such a direct relationship to a changing ambient in any other computer of mine.

I wonder if this means this particular computer is at it's optimal setup? That'ed be nice.....

[Reachable]

In a thread here at the beginning of the summer, somebody who seemed to know what he was talking about stated that the change in the temperature of the components would be 1:1 with the change in the ambient temperature (inside the case, presumably.) 'Certainly makes sense.

If the change in the temps of your components complied so easily with the change in the ambient of your house, it must mean, I would think, that your system has noticeably good airflow.

[dukla2000]

I wonder if this means this particular computer is at it's optimal setup? That'ed be nice.....

Good question - the answer depends on what you are trying to optimise!

Actually your config (and all of ours) are just conforming to standard thermal equations, in particular
dT = (0.05 * H) / Q
where dT is rise in temp above ambient in C, H is heat in Watts & Q is airflow in m3/minute. Except the equation applies from room ambient to internal case ambient to CPU temp (and possibly others).

Then again you could take the (SPCR influenced) view that optimal has got more to do with the acoustics than the temperatures, so the fact your CPU temp dropped is in fact sub-optimal because in fact it would have been 'better' for your fan speeds (and noise) to automatically drop to maintain the same temp

[AZBrandon]

Don't power supplies get less efficient as the ambient temperature goes up? That would make for at least one component that would have a non-linear temperature gain due to consuming more power at higher temps because of lower conversion efficiency.

[dukla2000]

Interesting point In one Seasonic doc I have they just say

The rated power will derate from 100% to 80% from 40oC to 50 oC Linearly

without explanation. My assumption is a cooling limitation (they can no longer keep the temp: ambient + dT within spec based on the available airflow) but there could be efficiency implications as well.

[Bluefront]

dukla2000.....My Black Knight case actually came with an automatic fan controller that you could enter a CPU temp (say 42C), and the controller would change the fan speed up/down at 2.8v to about 11.8v to maintain that set temp.

The controller I use now works somewhat differently. I set a CPU idle temp by adjusting the base idle (software).....then enter an optimal CPU load temp. The controller then adjusts the fan speed up as the temp rises.

This works ok....but I will have to lower the base idle manually, to take advantage of the lower ambient temp. Takes a few seconds.....

[Reachable]

Interesting point In one Seasonic doc I have they just say

The rated power will derate from 100% to 80% from 40oC to 50 oC Linearly

without explanation. My assumption is a cooling limitation (they can no longer keep the temp: ambient + dT within spec based on the available airflow) but there could be efficiency implications as well.

Are they talking about ambient temperatures here or temperatures measured inside the PSU? I don't see how the PSU can 'know' what the ambient of the case is.

They seem to be concerned about damage, so they engineered a gradual decline in the unit's output potential between 40C and 50C. The word "linearly" would further indicate that it was engineered.

So the quote doesn't indicate that the unit loses efficiency at higher temperatures. That would be disastrous, anyway. At lower efficiencies the unit would work harder and produce more heat, leading to a further rise in ambient temperature and on and on in a vicious cycle until the unit overheats.

[Tephras]

Are they talking about ambient temperatures here or temperatures measured inside the PSU? I don't see how the PSU can 'know' what the ambient of the case is.

They seem to be concerned about damage, so they engineered a gradual decline in the unit's output potential between 40C and 50C. The word "linearly" would further indicate that it was engineered.

So the quote doesn't indicate that the unit loses efficiency at higher temperatures. That would be disastrous, anyway. At lower efficiencies the unit would work harder and produce more heat, leading to a further rise in ambient temperature and on and on in a vicious cycle until the unit overheats.

All PSUs, AFAIK, are rated to deliver full output only within a certain temperature range, that is, or at least should be, listed in the specs as the "operating temperature". I don't know much about electronics but I think the above is because of the physical laws and nothing that are engineered.

[Reachable]

All PSUs, AFAIK, are rated to deliver full output only within a certain temperature range, that is, or at least should be, listed in the specs as the "operating temperature". I don't know much about electronics but I think the above is because of the physical laws and nothing that are engineered.

From what I've encountered, the term "operating temperature" in the specs of electronic devices means the range of ambient temperatures at which the device is assured to work.

I'm not an engineer, but it really does seem reasonable that there would be safeguards engineered into these devices to protect them from damage from heat.

[MarvW]

Ambient temperature: Is the temperature surrounding an object (room temperature).
Temperatures recorded in a system a normally the board temp and CPU temp.
Board temp is supplied by a thermal transducer mounted under the CPU.
Cpu temp Is supplied by an imbedded transducer in the Cpu.
Unless there is another transducer involved ,fan controllers, case sencers, such as the Antec P160 come with some sensors to place where ever you want.
Ambient, board and case temperature refers to the same thing.
Which is normally slightly higher than room temp.
ANY cooling is only going to be as efficient as it's suroundings. For example if the ambient temp is 80°F
You can't expect to get less than that anywhere in your system.

[Tephras]

From what I've encountered, the term "operating temperature" in the specs of electronic devices means the range of ambient temperatures at which the device is assured to work.

I believe you are right on that one, but in the ATX12V Power Supply Design Guide it's stated that the power supply efficiency should be tested under the recommended ambient temperature which is +10C to +50C, or in other words, the temperature range that most PSU specs I've seen specify as the "operating temperature" (0C to 40C is also quite usual). So the PSU manufacturers normally only guarantees that the PSU will deliver what it's rated for within that temperature range. However, what the actual testing conditions were during the test is something I've never seen mentioned in the specs. After all, it's only a recommended temperature so the PSU might be tested by the manufacturer under a low operating temperature, something that PSU manufacturers might do "to make their units look more powerful on paper" as MikeC wrote in this thread.
BTW, regarding the lower efficiency at higher temps, he also wrote, in the same thread, that "Efficiency of all electronic devices varies with temperature. Transistors, other semiconductors, caps etc. There appears to be an ideal temp/load where highest efficiency is reached for most electronics devices, but on either side of this ideal range, it falls away, and especially as you go beyond the upper limits, it falls away very steeply."

I'm not an engineer, but it really does seem reasonable that there would be safeguards engineered into these devices to protect them from damage from heat.

Yes, at least some PSUs have over-temperature protection.

[Ralf Hutter]

Ambient, board and case temperature refers to the same thing.

FWIW, I disagree.

Ambient temp = temperature of the room that your system is in.

Board temp = temp of whatever sensor on your board is being reported as the "board temp". This could be located in many places on the board. Some are better than others, virtually all are essentially meaningless, unless they're placed on or [/]very[/i] near a critical component such as VRM or NB.

Case temp = temp of the air in your case. Quite commonly measured with an external temp sensor placed about 1-1.5" above the CPU HSF fan.

[Jonas/Viper]

Ambient temp = temperature of the room that your system is in.

I agree on your definitions except that I think that ambient temp can vary depending on the context. If you are talking about which ambient temperatures a PSU can operate in, I would specify that temperature as the intake temperature, not the temperature of the room around it. If it's about a whole computer it would be the temperature of the room though.

[StarfishChris]

Or more simply,
ambient • adjective 1 relating to the immediate surroundings of something.
Case temperature isn't uniform - i.e. ambient temp at the CPU heatsink is higher than around the drives/intake point.
Also, rate of heat loss increases as the difference between heatsink and air temperature increases. Whether it's noticable or not is a different matter