Thread: Power Supples
View Single Post
  #10   Report Post  
Old April 9th 05, 11:18 PM
Telamon
 
Posts: n/a
Default

In article ,
Greg wrote:

Here's a line of power inverters marketed to photographers, but possibly of
interest for use with radios. They offer "pure sine wave" inverters as well
as cheaper switching power supplies.

What's the difference, in layman's terms, between sine wave and switching
power supplies? What are the pros & cons for practical applications?

I've always wondered if the switching power supplies, like the Radio Shack
models, that convert 120VAC to 13VDC, are appropriate for radios.


Those terms are for different things. Sine wave is the form or shape of
the voltage output from a supply. A switching supply is a circuit design
type that may or may not produce a sign wave.

The two main types of AC to DC supplies are linear and switching. A
linear supply will use a large and heavy, compared to a switching type,
transformer to step the voltage down and provide isolation from the AC
mains supply. The stepped down voltage is rectified, filtered and then
followed by a series linear voltage regulator. A switching supply has a
high voltage rectification bridge followed by the switching transistors
and transformer. The secondary side of the switcher is rectified
filtered and then a sample of the resulting DC voltage goes to a PWM
chip controller, which generates the the pulse that turns the front end
switching transistors on and off. The frequency of the switcher is
usually fixed for a design and the power transferred through the
transformer is controlled by the amount of time those front end
transistors are on, which is determined by the controller pulse width.
For isolation reasons the switching control pulse has to go through its
own small transformer from the secondary side back to the primary side.
Very small and cheap switchers use an Opto-isolator chip in place of the
small pulse control transformer.

What does all this information mean between linear and switching
supplies? Well, there are fewer opportunities for the simpler linear
power supply to generate radio interference and the switcher has many
more so the end result is that a linear supply will be quieter than a
switcher. The linear has its rectification and voltage regulation done
at lower voltages where the switcher does both at high voltage. Worse
the switcher is designed for fast on and off times for efficiency
reasons. High voltage and fast switching is a recipe for noise
generation. You could make the switcher as quiet as the linear but it
would cost a lot of money to make it.

Bottom line is the switcher will be more efficient, smaller and lighter
for the DC power it can deliver compared to a linear supply. The linear
supply will generate less conducted (EMI) and radiated (RFI) radio noise.

The noise may not bother cameras but can interfere with radio reception.

+++++++++++++++++++++++++++++++++++++++++

Now lets talk about AC to AC supplies, which is the sine wave question.

The linear way of controlling AC in and out in a linear fashion is a
transformer of some type. Some are designed to accept a small range of
input voltage variation and produce a fixed output voltage. These are
auto-transformers. No noise is generated with these units since there is
no switching. These are big, heavy and work for a limited input voltage
range. The wave shape in is the wave shape out.

Much more complex is the UPS or uninterruptible power supply. Where the
AC is converted to DC in order to charge the batteries. This AC to DC
could be linear or switching type. Now comes the "pure sine wave"
question. The DC needs to be converted back to AC. The simplest is a
square wave produced by high power transistors turned on and off at
60Hz. This is usually to raw for most electronic devices using AC power
so it is followed by some simple filtering. This then produces a
trapezoid waveform, which is the square wave with much slower rise and
fall time edges. Some electronics like computers will operate without a
problem with this type of waveform but many devices may have some
problems. For more sensitive electronics employing a higher rate of
switching than 60Hz can allow you construct the sign wave in a step like
form. The higher the switching frequency the smaller the steps and the
closer to a perfect sine wave you will produce. Draw a sign wave on a
piece of paper and then do the same thing using a stair case pattern
going up for the rise time and a stair case going down for the fall time
and that is what you would see with an oscilloscope.

There are obviously many opportunities for switching noise generation
and it would be a very expensive actively controlled AC to AC or UPS
type of unit to not generate plenty of radio noise.

I could go on but this is probably way more than you want to know
already.

--
Telamon
Ventura, California