From: cr88192@gmail.com
On 9/26/2025 8:32 AM, Michael S wrote:
> On Fri, 26 Sep 2025 12:10:41 -0000 (UTC)
> Thomas Koenig wrote:
>
>> BGB schrieb:
>>
>>> Brings up a thought: 960VDC is a semi-common voltage in industrial
>>> applications IIRC.
>>
>> I've never encountered that voltage. Direct current motors are
>> also mostly being phased out (pun intended) by asynchronous motors
>> with frequency inverters.
>>
>
> Are you sure?
> Indeed, in industry, outside of transportation, asynchronous AC motors
> were that most wide-spread motors by far up to 25-30 years ago. But my
> imressioon was that today various type of electric motors (DC, esp.
> brushlees, AC sync, AC async) enjoy similar popularity.
>
IIRC, reluctance motors are also popular here. They are sorta like BLDC,
but cheaper due to not needing big magnets (though, BLDC motors can give
more power in a physically smaller package if compared with reluctance
motors; but reluctance motors are still more compact if compared with AC
induction motors).
Like BLDC, it is possible to run reluctance motors at an exact speed.
This is unlike AC induction motors where, although speed can be adjusted
with a VFD, it isn't particularly exact as it depends on the load on the
motor and similar. Accurate speed control on an induction motor will
still require using an encoder, but they are still not good for
positional control (and the effective "holding torque" of an AC
induction motor is very low).
Where more accuracy is needed, something like a big BLDC or reluctance
motor with a servo-drive might be used (typically with hall-effect
sensors in the stator).
Generally, these motors can't be driven open-loop, as they are prone to
"drop out" at relatively little load in these cases.
Technically, the stator construction for a reluctance motor can be
nearly identical to an induction motor, the main differences are in the
design of the rotor.
Where, say, an induction motor typically has a hollow rotor consisting
of layered steel plates with an embedded copper or aluminum "squirrel
cage" (a ring of bars around the perimeter, all shorted together at the
top and bottom).
The reluctance motor can use a solid steel rotor, with gaps machined in
to control where magnetic flux will go.
A typical BLDC motor either has a ring of permanent magnets, or
alternating poles (from the top/bottom) with a central ring magnet.
I had before imagined it should be possible to make a hybrid of a
reluctance and induction rotor for intermediate effects; partly by
filling the gaps in the reluctance rotor with aluminum in place of air.
This could still operate synchronously, but could have better torque
under load and less issue with drop out. If it drops below synchronous
speed, it would instead induce eddy currents in the aluminum parts of
the rotor; rather than the air being "basically useless". However,
aluminum would still behave more like air as far as the magnetic flux
lines are concerned.
Though, some commercial designs had instead gone the other way,
hybridizing the reluctance rotor with a BLDC rotor, and using (cheaper)
ceramic magnets in place of rare-earth magnets (as typical in a BLDC).
One variant here resembling a reluctance motor with a split rotor, with
the top/bottom rotated relative to each other, and a central ceramic
ring magnet. Though, I think this pushes it more into the BLDC category.
Also common, on the AC side, are 440 and 208 3-phase.
Many traditional AC induction motors operate on 440VAC 3-phase.
A lot of traditional industrial machines were also 440VAC.
There is some stuff I saw about electrostatic motors gaining popularity
in some areas, but these tend to operate at high voltages but very
little amperage. They are comparably weak compared with magnetic motors,
but can be more energy efficient.
>>> What if, opposed to each computer using its own power-supply (from
>>> 120 or 240 VAC), it uses a buck converter, say, 960VDC -> 12VDC.
>>
>> That makes little sense. If you're going to distribute power,
>> distribute it as AC so you save one transformer.
>>
>
> I never was in big datacenter, but heard that they prefer DC.
>
DC -> DC allows higher conversion efficiency compared to AC.
Higher voltage distribution also allows more efficiency.
Higher voltage would be needed with DC vs AC, as DC is more subject to
resistive losses. Though, more efficiency on the AC side would be
possible by increasing line frequency, say, using 240Hz rather than
60Hz; but don't want to push the frequency too high as then the wires
would start working like antennas and radiating the power into space.
A higher line frequency would increase the relative efficiency of
electrical transformers. Higher voltage AC also has a higher conversion
efficiency than lower voltage.
In theory, assuming the AC comes in at 60Hz, could have a sort of rotary
converter to boost the line frequency (could have a vaguely similar
construction to an AC motor, but where input power uses 6 coils, and the
output side has 12 or 24 coils; likely also operating like a boost
transformer).
Not sure if anyone already builds this, or the conversion efficiency of
such a device. Would need to hopefully have a high conversion efficiency
(otherwise it would not offset the losses in all of any smaller
transformers).
Though, wouldn't really gain anything if just going directly to DC via
bridge rectifiers (with no intermediate transformers), and then using
DC-DC conversion.
So, say 1320VAC 3-phase could likely be rectified into 960VDC, where,
assuming the presence of big capacitors, the voltage would drop slightly
in conversion due to phase ripple (the "peaks" getting flattened out).
Or, in theory, I have little idea where people would get diodes and
capacitors big enough for this. Presumably giant industrial-sized diodes
and capacitors could exist though (well, and/or PCBs with craptons of
smaller components).
Then again, in a relative sense, boards with 1000s of diodes and
capacitors wouldn't cost much relative to the cost of the building and
servers.
...
>>>
>>> Or, 2-stage, say:
>>> 960V -> 192V (with 960V to each rack).
>>> 192V -> 12V (with 192V to each server).
>>>
>>> Where the second stage drop could use slightly cheaper transistors,
>>>
>>
>> Transistors?
>
> Yes, transistors. DC-to-DC convertors are made of FETs. FETs are
> transistors.
>
Yes, pretty much.
MOSFET, diode (from ground), inductor, and a capacitor;
Then you need a controller circuit to keep track of the voltage and
[continued in next message]
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
