How smart is a computer-type UPS?

[Rowbotth]

I have one of those grey box UPS that is supposed to take over in case
of a power outage, connected to the VCR in the lving room. Recently, I
replaced the batteries due to problems when the power went out. (The UPS
did not keep power top the VCR.)

Since then, I've noticd a peculiarity which I don't fully understand.

If I shut off the circuit at the breaker panel which also feeds the UPS
& VCR, the resultant power outage - even if less than 15 minutes, and
with nothing being recordrd on the VCR - invariably leads to all power
being removed from the VCR, and the clock as well as all programming all
go to Electron Pergatory.

But if there is a power outage, lasting about the same time, the UPS
does lick in and saves the clock and will continue to record whatever is
being taped by the UPS.

As the Ancient Chinese Detective Harry Hoo used to say: "Two
Possibilities." Either my UPS wants to see the power drop off in a
semi-relaxed slope, and the abrupt slope of the panel breaker opening is
just too fast for the thing to respond; or is there another explanation?

HR.

 

[Charles Perry]

I have one of those grey box UPS that is supposed to take over in case
of a power outage, connected to the VCR in the lving room. Recently, I
replaced the batteries due to problems when the power went out. (The UPS
did not keep power top the VCR.)

Since then, I've noticd a peculiarity which I don't fully understand.

If I shut off the circuit at the breaker panel which also feeds the UPS
& VCR, the resultant power outage - even if less than 15 minutes, and
with nothing being recordrd on the VCR - invariably leads to all power
being removed from the VCR, and the clock as well as all programming all
go to Electron Pergatory.

But if there is a power outage, lasting about the same time, the UPS
does lick in and saves the clock and will continue to record whatever is
being taped by the UPS.

As the Ancient Chinese Detective Harry Hoo used to say: "Two
Possibilities." Either my UPS wants to see the power drop off in a
semi-relaxed slope, and the abrupt slope of the panel breaker opening is
just too fast for the thing to respond; or is there another explanation?

HR.

I vote for another explanation. I don't know what it is, but knowing how
UPS work (and having tested many) the explanations you presented do not
work.

Charles Perry P.E.

 

[Charles Perry]

I'm not understanding you. What is this supposed to prove, exactly?

I thought these things worked off AC, and switched to batteries when AC
was lost.

HR.

Lacking sophisticated measurement equipment, a nightlight on the output of
the UPS will let you know whether the UPS is actually working. If the light
is on and relatively stable (in brightness), the UPS is working.

Charles Perry P.E.

 

[EEng]

I have one of those grey box UPS that is supposed to take over in case
of a power outage, connected to the VCR in the lving room. Recently, I
replaced the batteries due to problems when the power went out. (The UPS
did not keep power top the VCR.)

Since then, I've noticd a peculiarity which I don't fully understand.

If I shut off the circuit at the breaker panel which also feeds the UPS
& VCR, the resultant power outage - even if less than 15 minutes, and
with nothing being recordrd on the VCR - invariably leads to all power
being removed from the VCR, and the clock as well as all programming all
go to Electron Pergatory.

But if there is a power outage, lasting about the same time, the UPS
does lick in and saves the clock and will continue to record whatever is
being taped by the UPS.

As the Ancient Chinese Detective Harry Hoo used to say: "Two
Possibilities." Either my UPS wants to see the power drop off in a
semi-relaxed slope, and the abrupt slope of the panel breaker opening is
just too fast for the thing to respond; or is there another explanation?

HR.

A lot depends on the UPS that you use. Some are better than others..
Some, like switching supplies, require a minimum load to keep
operation in a power fault condition, some do not. Some have a slew
rate built in, some do not. A UPS does not "switch" to battery, it is
always on battery (keeps a trickle charge on the battery from the DC
Link. I suppose there may be some low cost cheapo UPSs that actually
do switch, but I'd advise against them for the very reason you had a
problem with utility loss v.s. tripped breaker. The ones that
actually do switch (very bad design) may depend on a slew factor,
provided by a discharge path from a charged RC at the input. If the
utility dies faster than the circuit can sense it, the UPS won't work.
Personally, you couldn't pay me to use a Slew Rated UPS. Very
undependable. You get what you pay for.

A UPS works by rectifiying the utility AC to provide a DC Link which
performs two functions. The DC Link provides a trickle charge to the
batteries and for the Inverters on the output which change it back to
AC. The resulting DC Link is typically 5-10% volts higher than is
necessary for the inverters, in order to do both functions. When the
input utility drops off, the DC Link continues at battery level, it
does not switch, providing sufficient DC from the batteries to the
Inverters. When utility power is restored, the rectified utility
resumes and the DC Link rises enough to do both functions once again.

In the type that require a minimum load for proper operation, there
will be a Current Transformer that provides Load Sensing. In the
absence of a load the Inverters are shut off to provide maximum charge
to the batteries (in an industrial UPS). In a simple store bought UPS
for computers, depending on the brand and model, this may not be true.

I have had great success with all Belkin UPSs. In applications where
a load may not be present at all times, I have added a small load bank
to guarantee a minimum current draw of 1A to ensure continued
operation for very light loads. Check with the manufacturer of your
UPS to determine whether a load is required for operation.

 

[Charles Perry]

A lot depends on the UPS that you use. Some are better than others..
Some, like switching supplies, require a minimum load to keep
operation in a power fault condition, some do not. Some have a slew
rate built in, some do not. A UPS does not "switch" to battery, it is
always on battery (keeps a trickle charge on the battery from the DC
Link. I suppose there may be some low cost cheapo UPSs that actually
do switch, but I'd advise against them for the very reason you had a
problem with utility loss v.s. tripped breaker. The ones that
actually do switch (very bad design) may depend on a slew factor,
provided by a discharge path from a charged RC at the input. If the
utility dies faster than the circuit can sense it, the UPS won't work.
Personally, you couldn't pay me to use a Slew Rated UPS. Very
undependable. You get what you pay for.

Nearly all small UPS (like those sold in office/computer supply stores) are
standye, meaning they switch to battery when the power goes off. The short
switching time does not effect computers since PC powe supplies will
actually ride through a few cycles without a UPS.

Charles Perry P.E.

 

[EEng]

Nearly all small UPS (like those sold in office/computer supply stores) are
standye, meaning they switch to battery when the power goes off. The short
switching time does not effect computers since PC powe supplies will
actually ride through a few cycles without a UPS.

Charles Perry P.E.

Right you are Charles, and the very reason I won't use them.
Obviously the OP is having a problem with one that doesn't switch fast
enough or his load hasn't the input capacitance to hold up the line
long enough for the UPS to kick in. I think its a piss poor design.
Belkin makes a low cost 1KVA industrial UPS with a Continuance DC link
(no switching) that performs beautifully, and is not load dependant.
Never had a problem. My UPS experience comes from servicing an 11
state region for Exide Electronics many years ago. We were installing
2.4MKVA UPSs on military bases. Beal AFB had the only 9.6MKVA UPS in
existence. I don't know if they've gotten larger since.

 

[Rowbotth]

Right you are Charles, and the very reason I won't use them.
Obviously the OP is having a problem with one that doesn't switch fast
enough or his load hasn't the input capacitance to hold up the line
long enough for the UPS to kick in. I think its a piss poor design.
Belkin makes a low cost 1KVA industrial UPS with a Continuance DC link
(no switching) that performs beautifully, and is not load dependant.
Never had a problem. My UPS experience comes from servicing an 11
state region for Exide Electronics many years ago. We were installing
2.4MKVA UPSs on military bases. Beal AFB had the only 9.6MKVA UPS in
existence. I don't know if they've gotten larger since.

It is hard to find small (say less than 1KVA) true double-conversion
UPS nowadays since 99% of them are of the "standby" types where
inverter only turns on when there is power outage. Short duration
(less than 1 cycle or about 17 milliseconds) break in power does not
cause problems for switchmode power supplies used in computers & most
modern electronic devices. With 97%+ energy efficiency and very low
cost ($50 to $300) it is hard to beat at the low end.

Many medium sized UPS's (from 1 to 3KVA) use "line-interactive" or
"ferroresonant" designs which can "fill-in" the missing 1 AC cycle and
also provide isolation/active voltage regulation all the time. Since
only a single AC to AC power conversion takes place, conversion
efficiency is around 90%.

Larger UPS's (3KVA and larger) mostly use true double-conversion
designs because breaks/makes of 30A + current can damage not only
connected equipment but also the UPS. Power conversion efficiency is
rather poor at 85% or so since AC to DC and DC to AC conversion must
always take place at full power.
Air exhaust fan must be on all the time (noisy) and they are rather
expensive machines, but this is the best design for larger UPS's.

These different topologies were developed over the past 30 years &
each found their sweet spot (optimum performance vs. cost is realized)
after millions of units were installed. One cannot compare
motorcycles, passenger car and bus engines with a broad, sweeping
statement. Each engine was optimized for the application at hand - so
were the UPS topologies chosen for the types of equipment to be
protected![/QUOTE]

OK, the comments about the security and reliability are understood, but
as I understood this, they are sort of being ignored. I'm wanting to
keep power to a VCR, not keep power to the NSA Computers. There is a
difference, you know.

As for why the thing seems to vomit when I open a breaker in the house
panel, here's a theory. (I know it sounds silly; then again they used
to make stuff that would switch remotely devices, and this system which
used a transformer as the demarcation...)

So is it possible that the UPS is looking for a circuit on the feeder -
like through a transformer winding - and if it doesn't see this, it
responds to an unknown input by staying off? That would fulfill the
Shrelock Holmes definition of "Truth", but I don't know if the UPS is
that sophisiticated.

(Reply to my email with any answers, please. Vacation! (How would
Charles Shultz do a dancing Snoopy here, anyway?)

HR.

 

[Gym Bob]

Your ni-cad batteries need exercise and they are improving as you cycle
them.

 

[Charles Perry]

[email protected] (Nam Paik) wrote:

As for why the thing seems to vomit when I open a breaker in the house
panel, here's a theory. (I know it sounds silly; then again they used
to make stuff that would switch remotely devices, and this system which
used a transformer as the demarcation...)

So is it possible that the UPS is looking for a circuit on the feeder -
like through a transformer winding - and if it doesn't see this, it
responds to an unknown input by staying off? That would fulfill the
Shrelock Holmes definition of "Truth", but I don't know if the UPS is
that sophisiticated.

Nope. The UPS is not looking for a circuit. Some high end ones will not
start up if they are plugged into an improperly wired plug. Your's is not
having that problem. Have you tried testing it by simply unplugging it from
the wall?

Charles Perry P.E.