DC-DC UPS?

[Noozer]

Might be a dumb question, but the expensive part of a UPS is converting the
AC to DC for storage, then back to AC for use.

Why not put the UPS *after* the PSU... three (?) mini UPS's actually - 5v,
12v and 3.3v... and the -5v, etc if necessary. It could possibly even mount
into a 5.25" drive bay.

Sounds simple, doesn't it?

 

[kony]

Might be a dumb question, but the expensive part of a UPS is converting the
AC to DC for storage, then back to AC for use.

Why not put the UPS *after* the PSU... three (?) mini UPS's actually - 5v,
12v and 3.3v... and the -5v, etc if necessary. It could possibly even mount
into a 5.25" drive bay.

Sounds simple, doesn't it?

You write about putting it in a 5.25" bay which would
require very small batteries, meaning it couldn't have much
runtime. If long runtime for a system (only, including no
monitor or other peripherals if not powered by the system)
is not required, then we avoid the expensive larger,
traditional UPS, and can instead use one of the basic
commodity $40 types.

Building something that will sell only in low volume and
that requires more technical skill to implement by the
system integrator or end user will end up costing more.

A (computer, switching) PSU has feedback, such that it not
only regulates but in regulating, allows for the voltage to
remain reasonably constant (within ATX specs, hopefully)
regardless of the load on it. With a battery array set to
n.nV, the voltage will drop significantly as the load
increases. You would need this UPS-after-PSU to regulate
the output, and in doing so you have created a second power
supply in addition to the first one, making it all that much
harder to cram it into a small space or keep cost low.

It could be implemented a different way, if the mainboard
only needed 12V input (regulating that down to lesser
voltages on the mainboard subcircuits) or the main system
PSU was one like those used in Epia car-PCs that takes the ~
+12V input and regulates to all the needed rail voltages,
then the primary AC-DC PSU for the system could be a 12V
output (only) type that is then charging only a 12V battery
which is then directly powering the car-PC PSU in event of
power outtage.

This will also cost more since it's a niche product, and is
stil not likely to fit within a 5.25" drive bay unless only
a very short runtime is needed and the system doesn't use
much current. Although avoiding the DC-AC conversion would
make it smaller than a traditional UPS and could be more
efficient, it wouldn't be THAT much more efficient so the
battery size wouldn't decrease very much. Just buying a
proprietary sized battery in itself instead of a common type
used in an UPS would probably cost as much as the entire $40
UPS I'd mentioned above.

 

[Noozer]

You write about putting it in a 5.25" bay which would
require very small batteries, meaning it couldn't have much
runtime. If long runtime for a system (only, including no
monitor or other peripherals if not powered by the system)
is not required, then we avoid the expensive larger,
traditional UPS, and can instead use one of the basic
commodity $40 types.

True, but I'm thinking of something that would let the PC run for a 10
second interruption, then signal the PC to hibernate. Probably looking at
about 2 minutes of run time. No extra loads like monitors or printers to
worry about either.

Considering how much people rely on PC's I'm surprised that there isn't a
better power solution than the 120v AC UPS's.

....not to mention that most consumer UPSs put out really crappy power and
don't do ANYTHING to protect against power surges/etc.

 

[John McGaw]

Might be a dumb question, but the expensive part of a UPS is converting the
AC to DC for storage, then back to AC for use.

Why not put the UPS *after* the PSU... three (?) mini UPS's actually - 5v,
12v and 3.3v... and the -5v, etc if necessary. It could possibly even mount
into a 5.25" drive bay.

Sounds simple, doesn't it?

Some years back a friend and I started working on an uber-UPS which
would solve the runtime problem quite effectively by floating a
diode-isolated bank of series-connected 12V batteries across the
rectified primary DC of a computer's PS. We ran into concerns, not the
least of which was potential safety problems, and never got much past
the measuring and dreaming stage.

Given the weight and volume of even the small 6AH alarm gel batteries we
were intending to use it would not have been a lightweight solution (10
or 11 batteries) and the volume would have necessitated an external
battery enclosure in all except the most humongous AT tower cases of the
time but it would probably yield outstanding runtime. In modern
universal supplies the primary voltage is twice what we were assuming so
the number of batteries would have to be correspondingly doubled also.

 

[Guest]

|
| |> On Fri, 07 Sep 2007 09:04:51 GMT, "Noozer"
|>
|>>Might be a dumb question, but the expensive part of a UPS is converting
|>>the
|>>AC to DC for storage, then back to AC for use.
|>>
|>>Why not put the UPS *after* the PSU... three (?) mini UPS's actually - 5v,
|>>12v and 3.3v... and the -5v, etc if necessary. It could possibly even
|>>mount
|>>into a 5.25" drive bay.
|>>
|>>Sounds simple, doesn't it?
|>>
|>
|>
|> You write about putting it in a 5.25" bay which would
|> require very small batteries, meaning it couldn't have much
|> runtime. If long runtime for a system (only, including no
|> monitor or other peripherals if not powered by the system)
|> is not required, then we avoid the expensive larger,
|> traditional UPS, and can instead use one of the basic
|> commodity $40 types.
|
| True, but I'm thinking of something that would let the PC run for a 10
| second interruption, then signal the PC to hibernate. Probably looking at
| about 2 minutes of run time. No extra loads like monitors or printers to
| worry about either.
|
| Considering how much people rely on PC's I'm surprised that there isn't a
| better power solution than the 120v AC UPS's.
|
| ...not to mention that most consumer UPSs put out really crappy power and
| don't do ANYTHING to protect against power surges/etc.

That's becaus emost consumers don't want to pay all the extra cost of a
better power supply. They want the lowest prices. Manufacturers want to
keep getting a profit. The end result is you get what you pay for.

Some better power supplies _will_ ride through very short (less than one
second) power blinks just from the capacitors inside. If you overrate
the power supply (for example use an 800 watt one when 250 would have
been enough) you'll get bigger capacitors and maybe even 2 or 3 seconds
ride time. That's only good if the power comes back on real fast, not
enough for a clean hibernate.

There are power supply units that use DC input, like 12 volts or 48 volts.
Other voltages used in other markets might also be available (28, 36, and
72 volts). Some of them can take a wide voltage swing as well, giving you
plenty of range by using batteries aimed at the top voltage.

Note that these voltages are not the choice for most power efficiency.
Designing a data center for the most efficient use of power would likely
involve a 360 volt DC bus feeding each machine. The lower current would
mean less loss on the bus. Because power supplies actually convert the
voltage UP to around this level first, before converting back down to the
voltages the mainboard and devices use, a smaller more efficient power
supply could be used in each machine when the source voltage is in this
range. But such a setup would be far too hazardous for typical home use
(that is, not something sold commercially for a consumer market, even
though an engineer hobbyist might have no problem with it).

 

[Guest]

| Noozer wrote:
|> Might be a dumb question, but the expensive part of a UPS is converting the
|> AC to DC for storage, then back to AC for use.
|>
|> Why not put the UPS *after* the PSU... three (?) mini UPS's actually - 5v,
|> 12v and 3.3v... and the -5v, etc if necessary. It could possibly even mount
|> into a 5.25" drive bay.
|>
|> Sounds simple, doesn't it?
|>
|>
|
| Some years back a friend and I started working on an uber-UPS which
| would solve the runtime problem quite effectively by floating a
| diode-isolated bank of series-connected 12V batteries across the
| rectified primary DC of a computer's PS. We ran into concerns, not the
| least of which was potential safety problems, and never got much past
| the measuring and dreaming stage.

The design I thinking of a while back involved gating (using gate-turn-off
thyristors) each battery into a large capacitor, timing it so only one was
ever on at any one time, and wiring the capacitors themselves in series.
The thyristors would have to see the sum voltage, but the batteries would
not.


| Given the weight and volume of even the small 6AH alarm gel batteries we
| were intending to use it would not have been a lightweight solution (10
| or 11 batteries) and the volume would have necessitated an external
| battery enclosure in all except the most humongous AT tower cases of the
| time but it would probably yield outstanding runtime. In modern
| universal supplies the primary voltage is twice what we were assuming so
| the number of batteries would have to be correspondingly doubled also.

Legacy inverter design requires that. By developing your output voltage
across N capacitors in series, you can pump each capacitor one at a time
(at a high frequency) from a lower voltage.

 

[kony]

On 29 Sep 2007 13:45:41 GMT, (e-mail address removed)
wrote:

Because power supplies actually convert the
voltage UP to around this level first, before converting back down to the
voltages the mainboard and devices use, a smaller more efficient power
supply could be used in each machine when the source voltage is in this
range.

The difference would be negligible, when a PSU uses a
roughly 360VDC input to the chopper transistors it is merely
a bridge rectifier followed by capacitors. Active PFC units
have to maintain a level above the peak rectified DC so 360V
wouldn't be enough margin. The PSU will still need
capacitor(s) for this 360V so the only thing removed would
be the bridge rectifier, roughly 1% efficiency difference
and practically no difference in PSU size.

Further, commodity grade electrical supply line wire is not
very expensive relative to most other components, the loss
on the (existing AC, or per your idea going to DC) supply
wiring is also negligible.

But such a setup would be far too hazardous for typical home use
(that is, not something sold commercially for a consumer market, even
though an engineer hobbyist might have no problem with it).

It would be impractical and expensive, but the hazard not
significantly different than present 110/220V AC delivery
systems. All that to save a trivial amount of power.

 

[meow2222]

Might be a dumb question, but the expensive part of a UPS is converting the
AC to DC for storage, then back to AC for use.

Why not put the UPS *after* the PSU... three (?) mini UPS's actually - 5v,
12v and 3.3v... and the -5v, etc if necessary. It could possibly even mount
into a 5.25" drive bay.

Sounds simple, doesn't it?

If you could charge a 12v battery from 12v it would be extremely
simple. You cant though, so the ups would have to charge itself
with a convertor.

But given the tiny capacity and low charge speed, such a unit
would be quite doable. The miniature battery bank would need to
use something with fast discharge, maybe NiFe cells, maybe NiCds.

The one downer is that the user would need to open the pc up to fit
it, and thus it would never see the sales volume of plug-in units,
and the resultant low prices. This is a rather big downer! That could
maybe be solved by creating a British Standard for a connector
position in a 5.x" bay, so that such units could just slot in, no
case
opening required.


NT

 

[meow2222]

PS for 10 seconds of use one could possibly go with no
output regulation and stay in spec.
12v is 10x nicds, (and is non critical anyway.)
5v is 4x nicds = 4.8v
3.3v is 3x nicds = 3.6v - a schottky diode.
It would only need to stay in spec for enough seconds to hibernate.

To minimise costs one could charge the 5v pack from 12v line and
the 3.3v pack from the 5v line, so only the 12v pack would need a
small upconvertor.

That just leaves the q of controlling the UPS's output.


NT

 

[kony]

PS for 10 seconds of use one could possibly go with no
output regulation and stay in spec.
12v is 10x nicds, (and is non critical anyway.)

It's not so easy to assume a cell is 1.2V. They range from
1.45 or so (float, typical might be closer to 1.35V with a
load) down to 0.8 per cell depending on remaining capacity
and cell impedance vs discharge rate. This could easily put
the pack outside of a stable (ATX spec) voltage range.

 

[meow2222]

It's not so easy to assume a cell is 1.2V. They range from
1.45 or so (float, typical might be closer to 1.35V with a
load) down to 0.8 per cell depending on remaining capacity
and cell impedance vs discharge rate. This could easily put
the pack outside of a stable (ATX spec) voltage range.

If they only need power the pc for 10 seconds, and are forever on
trickle charge, when in use theyre always going to be fully charged,
under load, and with no charging source running. That gives a much
narrower voltage window. Whether its narrow enough is another
question, as you say.


NT

 

[kony]

If they only need power the pc for 10 seconds, and are forever on
trickle charge, when in use theyre always going to be fully charged,
under load, and with no charging source running. That gives a much
narrower voltage window. Whether its narrow enough is another
question, as you say.

I would expect it to take a bare minimum of sub-C sized
NiCad, if not larger, to maintain the high current required
without dipping too far under a stable voltage range.
That's at least 17 cells for the 3 major rails, it probably
wouldn't fit in a 5 1/4" drive bay, and the total project
would probably cost as much if not more than a 500VA UPS one
might find for about $30 after rebate. It might be most
worthwhile when size or weight were the most critical
factors, though Li-Ion cells might be even better for that
if the budget stretched to cover a more sophisticated
charging circuit. Enough Li-Ion cells & current capability
might fit in a 5 1/4" bay space.

 

[kony]

I would expect it to take a bare minimum of sub-C sized
NiCad, if not larger, to maintain the high current required
without dipping too far under a stable voltage range.
That's at least 17 cells for the 3 major rails, it probably
wouldn't fit in a 5 1/4" drive bay, and the total project
would probably cost as much if not more than a 500VA UPS one
might find for about $30 after rebate. It might be most
worthwhile when size or weight were the most critical
factors, though Li-Ion cells might be even better for that
if the budget stretched to cover a more sophisticated
charging circuit. Enough Li-Ion cells & current capability
might fit in a 5 1/4" bay space.

.... but of course, we've said nothing about the particulars
of the system. It'd be quite a bit easier to keep a Via C3
system w/integrated everything and running off a CF card
(maybe 40W peak power) up for 10 seconds than a Pentium 4
based gaming rig or fileserver with several drives. To add
to the irony in many cases windows decides to spin up
sleeping drives right before powering off a system. There
must have been some genius put into that (not!).

 

[meow2222]

I would expect it to take a bare minimum of sub-C sized
NiCad, if not larger, to maintain the high current required
without dipping too far under a stable voltage range.
That's at least 17 cells for the 3 major rails,

I make it 10 at nominal voltage, but with the high load current you
might be closer.

it probably
wouldn't fit in a 5 1/4" drive bay,

I see no probem in getting it in a standard CDROM enclosure.
Sub-C batteries, dropper for charging, and a little detection
electronics would fit in half the size.

and the total project
would probably cost as much if not more than a 500VA UPS one
might find for about $30 after rebate.

Cost depends entirely on sales volume. Eliminating both the mains
psu and the invertor are causes for a real price drop, but real world
cost depends as much on business and market factors.

It might be most
worthwhile when size or weight were the most critical
factors,

true, but if it contains much less stuff and costs less it could
potentially take over market dominance from the current plugin jobs.

One could possibly be daring and go with an SLA battery in a
plastic liner on the basis that by the time its rotted through the
ups
case the pc will be 10 years old and of little value anyway. A mini
SLA would be cheaper than NiCds.

Using NiFe would enable even smaller cells.

though Li-Ion cells might be even better for that
if the budget stretched to cover a more sophisticated
charging circuit. Enough Li-Ion cells & current capability
might fit in a 5 1/4" bay space.

If you up the price its range of application dwindles, and small
sales
volume equals high price - just not the way to go imho.


NT

 

[kony]

kony wrote:

I make it 10 at nominal voltage, but with the high load current you
might be closer.

I was speaking of the cells for the 5V and 3.3V rails in
addition to the 12V.

I see no probem in getting it in a standard CDROM enclosure.
Sub-C batteries, dropper for charging, and a little detection
electronics would fit in half the size.

With Sub-C, yes that would be possible but I was thinking
along the lines of it being fairly expensive just to get a
few dozen seconds of runtime (if that), it would be better
to have larger than Sub-C and then discharge the cells less
to decrease the voltage drop, plus the larger cells
typically have a lower impedance as well. There should
probably be some cutoff circuit for when they fall below a
certain threshold as having the system write to disk when
instable could be even worse than just abruptly shutting off
in some cases.

Cost depends entirely on sales volume. Eliminating both the mains
psu and the invertor are causes for a real price drop, but real world
cost depends as much on business and market factors.

Yes, the cost has almost nothing to do with parts
elimination, only market demand for volume of the product.
You could probably buy a whole CDROM drive and gut it, that
costing less money than it would cost for an empty
electronics enclosure of the right dimensions, for example.

true, but if it contains much less stuff and costs less it could
potentially take over market dominance from the current plugin jobs.

I can't see that happening, I wouldn't expect it to run for
over a minute if that long, contrasted with 15 minutes for a
device that can power more than just the computer. Remember
that often people might be actively using the system and may
want the monitor to work, or an inkjet printer or
modem/router/etc. depending on the circumstances. There's
enough of a price disparity between SLA and NiCad that it
wouldn't end up significantly less expensive but would have
significantly shorter runtime (and versatility since it
lacks any 110-220 VAC output).

These other devices could also have their own individual
battery packs but then the cost continues to rise.

Also we haven't even considered the little odds and ends to
implement this, like the wiring harnesses, connectors, etc,
requiring the primary AC-DC PSU to input to this battery
pack for a switchover circuit or else a more crude and lossy
isolation would have to be used.

One could possibly be daring and go with an SLA battery in a
plastic liner on the basis that by the time its rotted through the
ups
case the pc will be 10 years old and of little value anyway. A mini
SLA would be cheaper than NiCds.

Using NiFe would enable even smaller cells.



If you up the price its range of application dwindles, and small
sales
volume equals high price - just not the way to go imho.

I think we'd already reached that point when the
conversation about an internal battery began. It's not as
though the tech to make it happen didn't exist, it's that
nobody perceives a large enough market for it to make it
practical. There may well be such a device that already
exists but that it's priced itself into obscurity already.
I vaguely really I saw some power supplies somewhere that
were half-height internally (I mean std. ATX form factor)
and had a layer of Ni-Cad batteries under the PCB. The main
problem with that today is the ever increasing current
requirements of modern systems, there's just not enough
empty space in decent ATX PSU for this today.

 

[meow2222]

I was speaking of the cells for the 5V and 3.3V rails in
addition to the 12V.

Tap those off the 12v pack, no need to duplicate.

With Sub-C, yes that would be possible but I was thinking
along the lines of it being fairly expensive just to get a
few dozen seconds of runtime (if that), it would be better
to have larger than Sub-C and then discharge the cells less
to decrease the voltage drop, plus the larger cells
typically have a lower impedance as well.

IIRC radio control enthusiasts use NiCd D cells to run 60A motors,
and half that would be enough for a PC.

There should
probably be some cutoff circuit for when they fall below a
certain threshold as having the system write to disk when
instable could be even worse than just abruptly shutting off
in some cases.



Yes, the cost has almost nothing to do with parts
elimination, only market demand for volume of the product.
You could probably buy a whole CDROM drive and gut it, that
costing less money than it would cost for an empty
electronics enclosure of the right dimensions, for example.



I can't see that happening, I wouldn't expect it to run for
over a minute if that long, contrasted with 15 minutes for a
device that can power more than just the computer. Remember
that often people might be actively using the system and may
want the monitor to work, or an inkjet printer or
modem/router/etc. depending on the circumstances.

Maybe so, though I assumed most folk just wanted to avoid losing
data on sudden power down. Even if its 50/50 that would still give
such a mini-ups half the potential UPS market - but perhaps theyre
bought mainly for other jobs.

There's
enough of a price disparity between SLA and NiCad that it
wouldn't end up significantly less expensive but would have
significantly shorter runtime (and versatility since it
lacks any 110-220 VAC output).

These other devices could also have their own individual
battery packs but then the cost continues to rise.

Also we haven't even considered the little odds and ends to
implement this, like the wiring harnesses, connectors, etc,
requiring the primary AC-DC PSU to input to this battery
pack

I think we may have been talking at cross purposes then. I was
proposing doing away with all that, just trickle charging the battery
pack from the PC's molex connector. Would need nothing more
than a basic bottom price linear regulator. Same molex serves the
PC with power during the mini-ups's brief run. I dont know what the
continuous current rating of those molexes is, but if the UPS only
powers the system for 20 seconds its not going to be a problem.

for a switchover circuit or else a more crude and lossy
isolation would have to be used.





I think we'd already reached that point when the
conversation about an internal battery began. It's not as
though the tech to make it happen didn't exist, it's that
nobody perceives a large enough market for it to make it
practical.

You think so? I dont know, but I'd be mildly surprised if a large PC
mfr didnt think there was any mileage in offering a mini built in UPS
as an optional extra. Theres certainly a market for UPSes.

There may well be such a device that already
exists but that it's priced itself into obscurity already.
I vaguely really I saw some power supplies somewhere that
were half-height internally (I mean std. ATX form factor)
and had a layer of Ni-Cad batteries under the PCB. The main
problem with that today is the ever increasing current
requirements of modern systems, there's just not enough
empty space in decent ATX PSU for this today.

NT

 

[kony]

Tap those off the 12v pack, no need to duplicate.

To arrive at ~ 5V alone it could be done, but to remain at ~
5V under high current may not be so easy with sub-C cells.

IIRC radio control enthusiasts use NiCd D cells to run 60A motors,
and half that would be enough for a PC.

The motors don't care if the voltage drops while a PC does.
Running a pack with a tap in it to derive 5V and 3V power
might be too much current to keep voltage up. Even so, D is
definitely an improvement and yet getting ever closer to
exceeding available space in a drive bay. Remember it's not
just the batteries but the circuit, insulating material or
standoffs, the incoming and outgoing connectors or wiring
harness.

Maybe so, though I assumed most folk just wanted to avoid losing
data on sudden power down. Even if its 50/50 that would still give
such a mini-ups half the potential UPS market - but perhaps theyre
bought mainly for other jobs.

Some might only need to avoid losing data, but how many
would take a device with lesser versatility and capacity
just to save a bit of space outside of the PC while taking
up space and having to route a second thicker wiring harness
inside the PC? There's also another problem in that PC PSU
with rail sense leads use these to sense at the ATX
connector under the presumption it is plugged into the load,
while the resultant voltage at the load will be lower than
it senses due to the intermediate stage of having to go
through the battery pack subcircuit.

I think we may have been talking at cross purposes then. I was
proposing doing away with all that, just trickle charging the battery
pack from the PC's molex connector. Would need nothing more
than a basic bottom price linear regulator.

Pack input isn't the problem, that'll work for charging but
you still need to isolate the two supplies from each other.
You have two power sources and you have input to the pack
through this regulator, but it still has to get to the
motherboard without being a common rail to the PSU, and both
have to be plugged into everything so the original PSU plus
then has to be unplugged (else a lot of ugly grafting of
leads onto the existing PSU wiring harness) . Thus, you
have to unplug the PSU from the mobo and put it's connector
onto another circuit board (or dongle style split) along
with the output from the battery pack to the motheboard and
other components needing power.

Same molex serves the
PC with power during the mini-ups's brief run. I dont know what the
continuous current rating of those molexes is, but if the UPS only
powers the system for 20 seconds its not going to be a problem.

6A, IIRC, but there's the voltage drop problem across the
connector and single supply leads making it too lossy.
You'll also have to switch the lead from input to output
around the charging circuit if I understand what you're
suggesting, then have the power going through these single
supply leads up to the PSU wiring harness distribution point
and down into the main PSU harness, which is quite a
distance to travel for the currents in a PC.

You think so? I dont know, but I'd be mildly surprised if a large PC
mfr didnt think there was any mileage in offering a mini built in UPS
as an optional extra. Theres certainly a market for UPSes.

An OEM could come closer to the goal because they can limit
the unit's compatibility with certain systems' unique
current requirements. On the other hand we'd be cursing all
that much more about the proprietary way they ended up
getting the job done, OEMs don't need much of an excuse to
do screwy things that make it either impossible or very
costly to replace anything... keeping in mind that even if
the mini-UPS were to work there will always be a certain
failure rate and replacement requirement.

 

[Guest]

| The difference would be negligible, when a PSU uses a
| roughly 360VDC input to the chopper transistors it is merely
| a bridge rectifier followed by capacitors. Active PFC units
| have to maintain a level above the peak rectified DC so 360V
| wouldn't be enough margin. The PSU will still need
| capacitor(s) for this 360V so the only thing removed would
| be the bridge rectifier, roughly 1% efficiency difference
| and practically no difference in PSU size.

So standardize on 400V or more if needed. I do believe a bit more than 1%
of the power loss comes from the initial upconversion in a typical PSU,
based on how hot I've seen said parts get (and in a couple of my many PSU
burnups, on those parts).

Where there's dissipation, there is efficiency gained if that part can be
eliminated without incurring that cost elsewhere (but that may be an issue
in the data center to convert its incoming 3 phase 480VAC to 400VDC and
managing the paralleling with the battery banks).


| Further, commodity grade electrical supply line wire is not
| very expensive relative to most other components, the loss
| on the (existing AC, or per your idea going to DC) supply
| wiring is also negligible.

It's not too bad at 120VAC, though a it better at 240VAC (which I would
prefer to use). But a design around 12VDC would be some very big wires.
It's both the cost of the copper and the heat loss, which is proportional
to the _square_ of the current for the same size conductor (no, you would
not do that), which means it is proportional to the current when the wire
size is also proportional to the current (e.g. much more wire with the
same current density).


|>But such a setup would be far too hazardous for typical home use
|>(that is, not something sold commercially for a consumer market, even
|>though an engineer hobbyist might have no problem with it).
|
|
| It would be impractical and expensive, but the hazard not
| significantly different than present 110/220V AC delivery
| systems. All that to save a trivial amount of power.

Such a system can be made safe by competent electricians. But for homes,
the standards have to consider do-it-yourself-ers doing the wiring. DC
is not as easy to interrupt as AC. And your fault current from a central
PSU might not have enough to work the breakers properly.

Even common DC to 120VAC inverters are dangerous for homes because such
inverters can't properly trip circuit breakers when there is a short on
the AC side.

 

[kony]

| The difference would be negligible, when a PSU uses a
| roughly 360VDC input to the chopper transistors it is merely
| a bridge rectifier followed by capacitors. Active PFC units
| have to maintain a level above the peak rectified DC so 360V
| wouldn't be enough margin. The PSU will still need
| capacitor(s) for this 360V so the only thing removed would
| be the bridge rectifier, roughly 1% efficiency difference
| and practically no difference in PSU size.

So standardize on 400V or more if needed. I do believe a bit more than 1%
of the power loss comes from the initial upconversion in a typical PSU,
based on how hot I've seen said parts get (and in a couple of my many PSU
burnups, on those parts).

How did you "see" these parts getting hot? Some parts can't
be eliminated in the switch from AC to HV DC, like current
inrush limiting.

Believe whatever you like, but typically it's just a bridge
rectifier, then a voltage doubler in 110VAC locations, so
it's 2 x diode drop of about 0.7V. You still have to have a
capacitor, though perhaps a smaller capacity would suffice
if the power were already DC. That combined with removal of
the bridge rectifier isn't much cost or space savings. Some
AC filtration circuitry could also be reduced or removed but
not all of it. These parts are only a small % occupancy of
the real-estate in a PSU.

Where there's dissipation, there is efficiency gained if that part can be
eliminated without incurring that cost elsewhere (but that may be an issue
in the data center to convert its incoming 3 phase 480VAC to 400VDC and
managing the paralleling with the battery banks).

yes but only a minimal amount of efficiency gained. There
are lots of other ways to reduce power consumption so it is
stretching a point too far.

| Further, commodity grade electrical supply line wire is not
| very expensive relative to most other components, the loss
| on the (existing AC, or per your idea going to DC) supply
| wiring is also negligible.

It's not too bad at 120VAC, though a it better at 240VAC (which I would
prefer to use). But a design around 12VDC would be some very big wires.

??? Who suggested using a 12VDC whole-site supply? It's
non-applicable.

The topic was using existing mains supply infrastructure or
moving to something proprietary for a supposed gain. Also,
you are ignoring that converting from the mains AC to this
HV, ~ 400VDC is not without losses. It still has to be
rectified, did you propose using synchronous rectification
(which has more benefit vs cost in lower voltage
applications) or continuing to use a bridge rectifier? If a
bridge, you have lower power loss with individual bridge
rectifiers in each system PSU than only one for the entire
site HD DC, since voltage drop rises with a current
increase.

|>But such a setup would be far too hazardous for typical home use
|>(that is, not something sold commercially for a consumer market, even
|>though an engineer hobbyist might have no problem with it).
|
|
| It would be impractical and expensive, but the hazard not
| significantly different than present 110/220V AC delivery
| systems. All that to save a trivial amount of power.

Such a system can be made safe by competent electricians. But for homes,
the standards have to consider do-it-yourself-ers doing the wiring.

It would not be more hazardous to any significant extent,
it's not as though a DIYer doesn't have to use same methods
for 110/220V.

DC is not as easy to interrupt as AC. And your fault current from a central
PSU might not have enough to work the breakers properly.

The current isn't THAT much lower, if all these proprietary
supply parts were being installed it would simply be a
matter of also sourcing suitable breakers.

Even common DC to 120VAC inverters are dangerous for homes because such
inverters can't properly trip circuit breakers when there is a short on
the AC side.

There's no need for them to trip a home circuit breaker
until the current exceeds a safe level to that inverter.
Instead they have internal fuses, it is shutting off the
device with a problem instead of the entire AC circuit,
exactly the solution you'd want.

Not dangerous, optimal.

 

[meow2222]

To arrive at ~ 5V alone it could be done, but to remain at ~
5V under high current may not be so easy with sub-C cells.

Neither of us really knows what size cells, but of course theyll be
sized for the job. However given that D cells can knock out 60A
there doesnt seem much doubt that it'll all fit in a standard 5.25"
case.

I'd look at NiFe rather than NiCd though. With their very high
current ability they can be made smaller and more V stable.

The motors don't care if the voltage drops while a PC does.
Running a pack with a tap in it to derive 5V and 3V power
might be too much current to keep voltage up. Even so, D is
definitely an improvement and yet getting ever closer to
exceeding available space in a drive bay. Remember it's not
just the batteries but the circuit, insulating material or
standoffs, the incoming and outgoing connectors or wiring
harness.

those other parts are fairly small volume-wise.

Some might only need to avoid losing data, but how many
would take a device with lesser versatility and capacity
just to save a bit of space outside of the PC while taking
up space and having to route a second thicker wiring harness
inside the PC?

I'm not planning any extra wiring required inside the PC. Why do you
think there would need to be? V_drop too high?

Why would a business user choose an inbuilt miniups?
1. Because they dont want to fork out on a large whole site UPS, nor
have loads of mains plugin UPSes cluttering up the place.
2. Lower cost than an external plugin UPS
3. Ease: just tick the box on the form and you get a data-safe PC.
4. Any company PCs used offsite will always be UPS protected
this way

There's also another problem in that PC PSU
with rail sense leads use these to sense at the ATX
connector under the presumption it is plugged into the load,
while the resultant voltage at the load will be lower than
it senses due to the intermediate stage of having to go
through the battery pack subcircuit.

main psu doesnt go through the batt pack, the wiring in the pc is
not affected, the miniups just plugs into a molex connector, end of
job. If running the Pc's original wiring loom proves too high a
V_drop
to set the batt pack out a little higher and run it open loop, then
there would also need to be an exrta sense wire tagged on. Very
easy.

Pack input isn't the problem, that'll work for charging but
you still need to isolate the two supplies from each other.
You have two power sources and you have input to the pack
through this regulator, but it still has to get to the
motherboard without being a common rail to the PSU, and both
have to be plugged into everything so the original PSU plus
then has to be unplugged (else a lot of ugly grafting of
leads onto the existing PSU wiring harness) . Thus, you
have to unplug the PSU from the mobo and put it's connector
onto another circuit board (or dongle style split) along
with the output from the battery pack to the motheboard and
other components needing power.

You're way overcomplicating this. The battery pack output stays
connected via a relay whenever the machine is on. At power down
it senses that and continues supplying power for 15s (ie it doesnt
distinguish mains failure from normal power down) then cuts out.

A possible way to control charging is to connect the half way point
on the batt pack to each rail (0v or 12v) via a basic reg, thus
charging each half of the pack alternately. Doide & R ensures the
pack and its halves can discharge very fast but only charge at a
trickle.

6A, IIRC, but there's the voltage drop problem across the
connector and single supply leads making it too lossy.

Do you know what the R or range of R is? I've not measured it. The
ATX Vspec allows a fair amount of swing IIRC

You'll also have to switch the lead from input to output
around the charging circuit if I understand what you're
suggesting,

trying to avoid any such switching if poss. If we switched the batt
pack between charge and out, a power failure would mess with the
pc before the relay clicked in.

then have the power going through these single
supply leads up to the PSU wiring harness distribution point
and down into the main PSU harness, which is quite a
distance to travel for the currents in a PC.

yeah. If too much R and load variation we could resort to an extra
sense wire to fix it. That would require a high current reg though,
best avoided if poss.

An OEM could come closer to the goal because they can limit
the unit's compatibility with certain systems' unique
current requirements. On the other hand we'd be cursing all
that much more about the proprietary way they ended up
getting the job done, OEMs don't need much of an excuse to
do screwy things that make it either impossible or very
costly to replace anything... keeping in mind that even if
the mini-UPS were to work there will always be a certain
failure rate and replacement requirement.

They do things proprietary ways mainly because they can save
money. 20p off 10,000 units is £200, os if they can do it non
standardly for a few p less they will. But I dont see proprietary vs
not having a big effect on viability. One PC mfr would only capture
part of the market, but people buying new systems are an attractive
target for addons, especially if they promise to safeguard against
business data corruption or loss, and do it at lower cost than
anyone else's UPS, and with less hassle..


NT