h a l f b a k e r yTempus fudge-it.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Consider the process of supplying power to the
average computer.
Mains voltage goes into a UPS where it is converted
to 12V, then
converted back to mains voltage before being
converted to 12V, 5V,
and 3.3V. Doesn't that seem like a total waste of at
least one step?
(Granted, that's
a bit of a simplification of how most
UPSes work, but
it's close enough for most purposes.)
There's really no good reason for power supplies to
be inside the
computer case anymore. The UPS already converts
mains voltage to
12V, so at the very least that output should simply be
fed into the
computer case, where it is converted to whatever
voltages are
necessary. The problem is quite simple technically,
but the trouble is
getting everyone to agree on a standard. So I'm
proposing one.
Both computers and (external) power supplies would
have a "power
class" rating from 1 to 3:
Class 1 computers would accept mains voltage, and
class 1 power
supplies would basically be, well, an outlet. A
standard UPS that
supplies mains voltage would also be a class 1 power
supply (but with
battery backup). Class 1 is pretty much the way
things are now.
Class 2 computers would accept 12V, and convert it
internally to any
other required voltages. Class 2 power supplies
would supply 12V.
Optionally, class 2 connectors can support (but are
not required to
have) a data connection such as USB, for
communicating information
such as battery status, power draw, temperature, and
so forth. Class 2
would really be a transitional class, so that existing
UPS and computer
designs could be tweaked to simply eliminate a single
step on each
side. The key thing is that to meet the class 2
standard, any given
device would have to use a standardized connector.
Class 3 computers would accept a connector that
carries 12V, 5V, and
3.3V, along with a mandatory data connection. Class
3 power supplies
would similarly provide these voltages plus the data
line. The data
connection would be required to be able to talk to
multiple systems
at once, as under the standard one power supply
could be used to
power multiple computers. Additionally, class 3
computers could be
able to (but are not required to) accept multiple
power supplies, for
redundancy or load distribution purposes. Again, to
meet the
standard a standardized connector would have to be
used on all
equipment.
The various power classes would not be compatible
with each other in
any way, but any device could be certified for
multiple power classes.
For example, a UPS that provided both mains voltage
and 12V with a
data connection would be certified as power class
1/2. As long as the
power supply and computer have at least one
matching power class,
you can be certain that they will work with each
other.
DC PC PSUs
http://www.powerstream.com/DC-PC-48V.htm [Wrongfellow, Feb 14 2011]
[link]
|
|
What counts as a computer? A desktop terminal? A slimline laptop? A solid state tablet? A smartphone? The embedded chip in your washing machine? |
|
|
I'm really using "computer" here to refer to the power supply
found inside the computer case. I'm using that convention to
avoid confusion between the external power supply or UPS
and the internal power supply similar to what is almost
universally found on desktop computers. So to answer your
question, anything that has a power supply COULD technically
be classified under this system, but it's not really useful for
power systems that wouldn't normally make use of an external
UPS, such as laptops, phones, or anything else with an
internal battery. It would probably only be used for desktop
computers, servers, and so on. |
|
|
You can get PC power supplies that run off various DC inputs, including 12V (link). |
|
|
The point isn't simply to get a power supply that can
take DC voltages, but rather to eliminate the redundant
conversion of AC to DC to AC to DC as is normally the
case with most computers with a UPS attached. Since
the UPS is already converting mains voltage to 12V
anyway, why not let it power the computer directly?
Power supplies could be made significantly cheaper due
to fewer parts, and peak power ratings could be
increased for the same equipment because the battery
would be available as a fallback in case the power supply
output rating is temporarily exceeded. |
|
|
[bigsleep], I'm not sure what you mean. It states in the
description that in order to meet the standard, all
equipment would need to use a standardized connector
for a given power class. |
|
|
I haven't changed the description since I wrote it. I hardly
blame you for not noticing it thoughit's my own fault for
making the description so verbose. Truth be told, my eyes
always glaze over a bit when I read some of the lengthier
descriptions on here. |
|
|
The answer is I probably should quit Halfbaking at 3AM,
but when inspiration strikes... |
|
|
//The point isn't simply to get a power supply that can take DC voltages, but rather to eliminate the redundant conversion of AC to DC to AC to DC// |
|
|
Well, that's exactly why people make power supplies that can take DC. |
|
|
The power circuitry for a phone exchange goes roughly |
|
|
mains in ----> BIG HONKING SCARY BATTERY ----> PCs and stuff |
|
|
So where exactly is the standard for "BIG HONKING
SCARY BATTERY" specified? I have no doubt that it can
be (and is already) done. That's what I meant by //The
problem is quite simple technically, but the trouble is
getting everyone to agree on a standard. So I'm
proposing one.// |
|
|
There are tensperhaps hundredsof millions of desktop
computers and servers out there attached to UPSes that
are undergoing the wasteful and ridiculous process of
converting AC to DC only to convert it back to AC then
back to DC. What's needed to solve this problem is a
/standard/, not a one off solution. |
|
|
So desktop PC's with built-in laptop-battery PSU's... as a bonus your laptop battery gets charged up. |
|
| |