Maker Pro
Maker Pro

UPS sine wave vs square wave output?

J

jsmith

Jan 1, 1970
0
Why such insistence on a UPS having a "pure" sinewave vs any other sort of
output wave form such as square wave?
 
D

Don Pearce

Jan 1, 1970
0
Why such insistence on a UPS having a "pure" sinewave vs any other sort of
output wave form such as square wave?
Because the power supplies in the equipment it is powering have all
been designed to accept a sine wave from the mains supply. There is no
guarantee how they would behave when presented with a square wave.

d

Pearce Consulting
http://www.pearce.uk.com
 
L

Larry Brasfield

Jan 1, 1970
0
jsmith said:
Why such insistence on a UPS having a "pure" sinewave vs any other sort of
output wave form such as square wave?


Many devices are over-stressed when powered from
square waves, or approximations thereof. For example,
a simple diode rectifier feeding a resevoir capacitor will
experience much higher RMS current when fed that way.
 
jsmith said:
Why such insistence on a UPS having a "pure" sinewave vs any other sort of
output wave form such as square wave?

Do a fourier transform on a square wave and you will find that it
represents the sum of the sine wave with same period, plus all the odd
hamonics of that sine wave, with amplitudes proportional to the
harmonic number - a 60Hz square wave is the sum of a 60Hz sine wave,
plus a 180Hz sine wave with one third of the amplitude, plus a 300Hz
sine wave with one fifth of the amplitude, and so on.

Put this into a transformer or a motor that is intended to be driven by
a 60Hz sine wave, and the higher harmonics excite extra eddy current in
the magnetic path, making the motor or the transformern run hotter than
it would have done if excited by a pure 60Hz sine wave.

The harmonics extend up to a frequency defined by the rise and fall
times of the square wave, which can be quite fast - fast enough to
interfere with local radio reception for a really cheap and nasty UPS.
 
C

Charles Edmondson

Jan 1, 1970
0
jsmith said:
Why such insistence on a UPS having a "pure" sinewave vs any other sort of
output wave form such as square wave?
Basically because there are things attached to them that assume that
they have sine waves, and behave very badly when when they get other
waveforms...

Motors, transformers, some electronics, esp. power supplies are designed
around low harmonic content sine waves. Give them bad waveforms, and
they loose efficiency, get warm, release magic smoke from components...
 
J

jsmith

Jan 1, 1970
0
-------------------------------------------------------> >
Because the power supplies in the equipment it is powering have all
been designed to accept a sine wave from the mains supply. There is no
guarantee how they would behave when presented with a square wave.
-----------------------------------------------------------Question:

How does one design a power supply to accept a sine wave but not a square
wave??
 
J

jsmith

Jan 1, 1970
0
-------------------------------------------------------->
Many devices are over-stressed when powered from
square waves, or approximations thereof. For example,
a simple diode rectifier feeding a resevoir capacitor will
experience much higher RMS current when fed that way.
-----------------------------------------------------------Question:
Wouldn't the filter capacitor enjoy receiving what amounts to alternating DC
in the form of a square wave?
 
L

Luhan Monat

Jan 1, 1970
0
jsmith said:
Why such insistence on a UPS having a "pure" sinewave vs any other sort of
output wave form such as square wave?
Probably the same marketing company that claims you need 'pure
rocky-mountain spring water and high country barley'.
 
L

Larry Brasfield

Jan 1, 1970
0
jsmith said:
-----------------------------------------------------------Question:
Wouldn't the filter capacitor enjoy receiving what amounts to alternating DC
in the form of a square wave?


That's a poor example, in the steady state (implied by
mentioning RMS). Consider instead the doubler used
in most off-line switching power supplies when strapped
to run off of 115 VAC. Your square waves will induce
much higher peak currents than the less steep wavefronts
of a sinusoid will.
 
L

Luhan Monat

Jan 1, 1970
0
Luhan said:
Probably the same marketing company that claims you need 'pure
rocky-mountain spring water and high country barley'.
But seriously folks, I ran a complete hardware/software engineering lab
out of a motorhome for several years. My Trace 810 inverter (modified
square wave) ran everything just fine. Thats not to say that some
electonic equipent may have problems - especially AM radios.
 
J

John Tserkezis

Jan 1, 1970
0
How does one design a power supply to accept a sine wave but not a square
wave??

You don't. Not intentionally anyway.

A square wave input of the same RMS value of an equivalent sine wave will
have a much higher crest factor. That is, the peak value is *much* higher in
relation to the RMS value, as compared to a sine wave.

Most power supplies simply bridge rectify the AC input, and feed that
somewhat pulsating DC into the regulator that does the bulk of the work.

With a sine input, that peak DC at the filters will be 1.4 times the sine RMS
value.
With a stepped square wave, or worse still, a pure square wave input of the
same RMS value as your sine input will have a peak much higher.

Thus, if your components are only expecting a peak of X volts, and you force
feed it something significantly higher than X, it'll blow.
 
J

jsmith

Jan 1, 1970
0
John Tserkezis said:
You don't. Not intentionally anyway.

A square wave input of the same RMS value of an equivalent sine wave will
have a much higher crest factor. That is, the peak value is *much* higher in
relation to the RMS value, as compared to a sine wave.

Most power supplies simply bridge rectify the AC input, and feed that
somewhat pulsating DC into the regulator that does the bulk of the work.

With a sine input, that peak DC at the filters will be 1.4 times the sine RMS
value.
With a stepped square wave, or worse still, a pure square wave input of the
same RMS value as your sine input will have a peak much higher.

Thus, if your components are only expecting a peak of X volts, and you force
feed it something significantly higher than X, it'll blow.
<http://counter.li.org>

So why not lower the peak voltage of the square wave to a safe level??
 
R

Roger Lascelles

Jan 1, 1970
0
jsmith said:
So why not lower the peak voltage of the square wave to a safe level??

Its not just the peak level which matters. Imagine a disharged capacitor
connected to a low impedance squarewave generator. The squarewave goes up
fast and the capacitor voltage must follow. This means a large current
flows through the diode and the capacitor. Now imagine a sinewave : the
capacitor has to follow the sinewave as it climbs up. The current pulse is
stretched out and not as large in amplitude.

Result : Transformer, diode and capacitor RMS current is less, which means
less heating of these parts.

Roger
 
T

Terry Given

Jan 1, 1970
0
Larry said:
That's a poor example, in the steady state (implied by
mentioning RMS). Consider instead the doubler used
in most off-line switching power supplies when strapped
to run off of 115 VAC. Your square waves will induce
much higher peak currents than the less steep wavefronts
of a sinusoid will.

Which is why the 900W OEM UPS' I worked on in MA had extremely long
switching times - the "square" waves had edges of around 500us - 1ms.
Think I=C*dV/dt. In our case the "problem" that steep edges caused was
the X capacitor in the downstream equipment, just like Larrys voltage
doubler.

Cheers
Terry
 
T

Terry Given

Jan 1, 1970
0
sort of



Do a fourier transform on a square wave and you will find that it
represents the sum of the sine wave with same period, plus all the odd
hamonics of that sine wave, with amplitudes proportional to the
harmonic number - a 60Hz square wave is the sum of a 60Hz sine wave,
plus a 180Hz sine wave with one third of the amplitude, plus a 300Hz
sine wave with one fifth of the amplitude, and so on.

Put this into a transformer or a motor that is intended to be driven by
a 60Hz sine wave, and the higher harmonics excite extra eddy current in
the magnetic path, making the motor or the transformern run hotter than
it would have done if excited by a pure 60Hz sine wave.

The harmonics extend up to a frequency defined by the rise and fall
times of the square wave, which can be quite fast - fast enough to
interfere with local radio reception for a really cheap and nasty UPS.

And the magnetic materials used in these motors, transformers etc. are
designed for 60Hz sinusoidal operation. Lamination thickness is chosen
based on 50/60Hz, so harmonics have much higher eddy current losses than
the fundamental. Hysteresis losses also skyrocket with frequency.

I once did an interesting job for a customer when I built AC drives. The
customer was a panelbeater (auto body shop? unsure of USian term) had a
large 400Hz angle grinder, all Al chassis. very rugged. They bought an
AC drive, and used that to run the angle grinder. The AC drive generated
a PWM voltage output (the load is usually an inductor, which integrates
the PWM voltage to get a nice sinusoidal (ish) current), the harmonics
of which made the grinder get too hot to hold after a few minutes use.
So we slapped some chokes on the output of the drive, and a few caps,
and voila - away went the heating problem. Well it did once we replaced
the 50Hz chokes (which caught fire) with something better designed for
the harmonics we jammed up it (much thinner lamination, grain-oriented
Si-steel set up such that the flux path followed the grain)

Cheers
Terry
 
J

John Tserkezis

Jan 1, 1970
0
jsmith said:
So why not lower the peak voltage of the square wave to a safe level??

That's not really possible.

If you reduce the on (peak) voltage, you'll have to increase the pulse width
to cater for the resulting lower RMS voltage.

That in itself isn't a problem. What *is* the problem is that by reducing
the peak, you're limiting your maximum power output.

A stepped square output is a reasonable comprimise, in that it's cheap to do,
and it brings the crest factor down to a reasonable value that won't kill most
(hopefully any) power supplies.

A sine inverter would remove all these issues. However, while it's
reasonably cheap to produce a good facsimile of a sine output, the filters to
smooth the output add to the weight and cost of the supply.

I'm still not convinced that it adds _that_ much to the cost of the inverter,
but mass production wins every time: square or stepped square inverters are
much more popular and plentiful than sine, thus their cost is artificially
brought down relative to sine inverters.
 
P

Pooh Bear

Jan 1, 1970
0
jsmith said:
-------------------------------------------------------> >
-----------------------------------------------------------Question:

How does one design a power supply to accept a sine wave but not a square
wave??

Start by considering the difference between the rms and peak voltage values
for each waveform.


Graham
 
T

Tam/WB2TT

Jan 1, 1970
0
jsmith said:
-------------------------------------------------------> >
-----------------------------------------------------------Question:

How does one design a power supply to accept a sine wave but not a square
wave??
John,

You can probably get a particular device to work with a square wave input.
For instance, if the load has a capacitor input filter that normally charges
up to near the peak of a sine wave, you could feed it a square wave of about
165 V peak. If you now fed this same square wave into a 120 V light bulb, it
would not last long. The manufacturer does not know what kind of load you
are going to connect to his UPS. Also, if the load uses power factor
correction, the square wave would probably mess it up.

Tam
 
F

Fabio G.

Jan 1, 1970
0
John Tserkezis said:
but mass production wins every time: square or stepped square inverters are
much more popular

However for medium/high power (>3-4 kVA) they are always sine wave
inverters (PWM bridge + LC filter)
 
G

Guy Macon

Jan 1, 1970
0
Terry said:
Which is why the 900W OEM UPS' I worked on in MA had extremely long
switching times - the "square" waves had edges of around 500us - 1ms.
Think I=C*dV/dt. In our case the "problem" that steep edges caused was
the X capacitor in the downstream equipment, just like Larrys voltage
doubler.

It also helps that many "square wave" UPSs don't just go from full
pluss to full minus, but instead go plus -> zero -> minus -> zero
-> plus -> and so forth. I seem to remember reading about a few
that would have five or seven voltage levels instead of three.
 
Top