From: cr88192@gmail.com
On 9/26/2025 9:28 AM, Scott Lurndal wrote:
> BGB writes:
>> On 9/25/2025 9:03 PM, Scott Lurndal wrote:
>>> MitchAlsup writes:
>>>>
>>
>>>
>>> Consider that there are losses converting from the
>>> primary (e.g. 22kv) to 480v (2%), and additional losses
>>> converting to 208v (3%) to the UPS. That's before any
>>> rectification losses (6% to 12%). With various optimizations,
>>> they reduced total losses to 7.5%, including rectification
>>> and transformation from the primary voltage.
>>>
>>
>> Hmm...
>>
>> Brings up a thought: 960VDC is a semi-common voltage in industrial
>> applications IIRC.
>
>> What if, opposed to each computer using its own power-supply (from 120
>> or 240 VAC), it uses a buck converter, say, 960VDC -> 12VDC.
>
>
> In those datacenters, the UPS distributes 48VDC to the rack components
> (computers, network switches, storage devices, etc).
OK.
I had thought they were usually 120VAC or 240VAC.
At least, what rack-servers I had encountered were usually one of these
(sometimes they had the little switch on the power-supply set to 240V
even in the US).
Then again, can also note that when setting up my milling machine,
lathe, and plasma table, that these were all using 240VAC for the power
distribution to the various components. These were all Tormach machines
though, so can't say for others.
48VDC also makes sense, as it is common in other contexts. I sorta
figured a higher voltage would have been used to reduce the wire
thickness needed.
Though, I don't actually know how real datacenters work here, just sort
of coming up with something assuming optimizing for the target goals
(powering all this stuff while minimizing electrical losses and cost).
I did realize after posting that, if the main power rails were organized
as a grid, the whole building could be done probably with 1.25" aluminum
bars.
Could power the grid of bars at each of the 4 corners, with maybe some
central diagonal bars (which cross and intersect with the central part
of the grid, and an additional square around the perimeter). Each corner
supply could drive 512A, and with this layout, no bar or segment should
exceed 128A.
Assuming if they were using 240VAC, seems like the typical housing setup
(12AWG wire) would be woefully insufficient. Would either need to be
heavily built up and/or use much heavier gauge wiring.
Or also solid copper or aluminum bars. Not sure if I had heard of this,
usual idea IIRC was that people always use wire for AC power, except
that if pushing a continuous load of several hundred amps, wire seems
less practical (would need to be very thick, hard to work with, and
expensive).
Granted, more likely they would run the cable closer to the rated values
and accept more energy loss due to electrical resistance (since, yeah, a
1.25" bar or similar for 128A is a little excessive).
Though, it seems likely that in this case, solid metal bars might be
cheaper than using a whole lot of heavy gauge wire. And, repurposing
generic aluminum bar-stock might be the cheapest option here (with joins
either as aluminum clamps or via welding).
If operating closer to conventional electrical ratings, could drop to
0.375" bars for 128A. Going much thinner, voltage drops and heat would
become an issue.
So, say:
0.250" likely high resistive loss.
0.375" roughly nominal.
0.750" maybe sufficiently low resistance
(could likely handle 500A before significant heat)
1.250" maybe overkill
Well, and could maybe put a plastic coating or similar on the bars to
limit accidental short-circuits. Decided to leave out analysis, but the
most likely option (to balance cost and effectiveness) would likely be a
post-install application of acrylic paint (latex paint would be
insufficient, epoxy likely too expensive, ...).
...
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
