View Single Post
  #40   Report Post  
Posted to uk.d-i-y
Peter Able[_2_] Peter Able[_2_] is offline
external usenet poster
 
Posts: 167
Default Full wave rectifier with a smoothing capacitor

On 09/04/2021 08:21, Pancho wrote:
On 08/04/2021 15:49, Peter Able wrote:


The OP mentions "LN reversal" so this thread is about mains. He also
mentions a "smoothing capacitor" - so it is AC Mains he is asking about.


Why does that mean AC Mains, as opposed to any AC rectification?



If it were DC mains or a battery he's far less likely to have mentioned
a smoothing capacitor - hence my guess that he's talking about AC Mains.
And I remember DC Mains !

"AC rectification" is a process, "AC mains" is an input to that process.
So they are different things - like petrol is to cars.

In this case AC mains is rectified, delivering only pulses of DC voltage
of one polarity. In the case of DC mains/batteries, the process of
rectification is virtually invisible, - DC in, DC out.

Half-wave rectification is accomplished using a single diode. The snag
is that one of the two mains input wires passes through the process. Get
that the wrong way around and what was thought to be the GROUND of the
TV chassis is actually at LINE voltage. Been a service engineer - got
the scars - still.

Bridge rectification is, effectively a switch being flipped 100 times a
second. Thus the output is always pulses of the same polarity - and
twice as many as for the half-wave. There's still the problem, though,
that neither of the outputs is safe to touch! Bridge rectification has
been known of for some time - but the cost of thermionic diodes was so
high, once upon a time, that it generally didn't catch on. These days
semiconductor bridges are so cheap - that's the way to go.

Isolating the Rectifier from the mains - i.e. making things safer - was
generally accepted to require an isolation transformer ahead of the
rectifier. At one time, these transformers were expensive. Now they
aren't - so some simple PSUs nowadays contain one. Some geezer came up
with the idea of having two identical output windings on the
transformer. You could then connect these windings in a way that one was
always delivering the mirror image of the other. This meant that when
one winding was offering a positive pulse the other would be offering a
negative - and vice versa. That cuts the number of diodes required from
4 to 2 in order to achieve full-wave rectification. That was a big
saving in terms of the cost thermionic diodes - but - cheap
semiconductor diodes have blown that option out of the water. No option
to have a DC input..... Until....

Modern power supplies take all of those ideas and shake them all up.
They start with a bridge rectifier, then a smoothing capacitor. The
resulting ripply DC is chopped into a transformer at about 100kHz by a
simple oscillator. The outputs of one or more of the post-transformer
rectifiers (usually half-waye) is fed back to the chopper to stabilise
the output voltages and this signal is the only thing that connects the
pre-tranformer circuitry to the post-transformer circuitry - usually,
nowadays, by means of an optical isolator, So the risk of electric
shock is gone. Open up an old PC PSU and you'll clearly see the Berlin
Wall nature of the isolation of one group of PCB tracks from the others !

E&OE !

PA