On Wed, 21 Dec 2005 23:25:58 GMT, Rich Grise
wrote:

On Wed, 21 Dec 2005 14:09:10 -0500, Steve Spence wrote:

wrote:
George Ghio wrote:


Tell us why anyone would modify a sine wave.


It's called "engineering," George.

Nick


Really... Wouldn't they rather modify a square wave to approximate a
sine wave? What would be the point of modifying a sine wave, when a sine
wave (or close approximation) is the required result?

I only am posting this back to RAO as it's the only NG I receive out
of the list.


I hope you're not serious here.

They don't make a sine wave and modify it, they make a rectangular
wave and call it a "modified sine wave" because it passes enough tests
for harmonics and crap that it will run most stuff, and they can get
away with it. ;-)

Switching supplies are very efficient. Unfortunately things making sin
waves from DC are not. The faster you can turn the transistors on and
off the less heat and more power you can get out of a still smaller,
lighter, and *cheaper* power supply, or inverter.

Without digging too deep, a square wave contains an almost infinite
frequency range on the rise and fall times. Most likely many
thousands of times higher in frequency than the fundamental square
wave.

If you poke said square wave with all it's noise into an L/C filter
it'll round it off into more of an approximation of a sin wave. You
only have to get rid of enough of the spikes/harmonics to keep from
confusing what ever you plan on running off the thing.

Anybody wanna do an FFT of various duty-cycle waveforms, and give us real
THD information, and how that relates to power factor, and etc, and etc,
and etc?

You want power factor too?
Lordy, when I think of the caps we used to switch in on the power
mains at work depending on load. I doubt current and voltage were
ever in phase. Well, maybe with the exception of the time we fired
back up after doing PM on some switch gear and some one had forgotten
to remove the jumpers. Boy, but that was noisy! It actually bent the
cabinet doors into shallow U-shapes and those things were made of 1/8"
thick steel and about 7 or 8 feet tall.


The one inverter I've ever had my hands on the guts of made a waveform
like this:

---- ---- ----
| | | | | |
- - - - - -
| | | | | etc.
- ---- ----

And the regulator was just based on an ordinary rectifier - they didn't
care about RMS, or it was scaled to get "close enough".

They were just after "close enough". That's why the 400 watt PS in
one computer here weighs about a pound. The 500 watt in two others are
about 4# and the 630 is still heavier. BUT the 200 watt PS for my ham
rig weighs over 30#. I have a 600 watt (12 VDC @ 50 A) PS under the
desk that must weigh about 60 or 70#, None of the computer Power
supplies cost much over $50. I think the 400 was about $35. The big
one for the ham rig was several hundred dollars.

I have seen inverters that did a pretty good job on the output wave
form, but they were expensive and not nearly as efficient as the cheap
ones as a good portion of their work went into heat.

But I do wonder, what does the harmonic content really do when you vary
the duty cycle?

I would *guess* they would get pretty drastic at some portions
depending on just how hard the thing is working. There's not much
power in them so I'd expect to see them round off under load, but does
the inverter use dynamic filtering or hard filtering? Which is a good
way of saying I don't really know. :-))

Roger Halstead (K8RI & ARRL life member)
(N833R, S# CD-2 Worlds oldest Debonair)
www.rogerhalstead.com

Some years ago, in the USAF, I saw some pulses on a spectrum analyzer,
and they had some really pretty envelopes. :-)

Thanks,
Rich