John Larkin wrote:
On Sat, 29 Sep 2012 19:34:02 -0400, Spehro Pefhany
wrote:
On Thu, 27 Sep 2012 12:44:00 -0700, the renowned Jim Thompson
wrote:
AC Cap Dropper DC Power Supply
Anyone know of a clever way to regulate other than the typical
brute-force zener clamp?
...Jim Thompson
If it's full-wave rectified you can turn a HV part 'off' when the
voltage exceeds your desired output voltage and lose the series cap.
Your output cap need only be as big as a full-wave rectified low
voltage supply would required.
The cap is better with transients most likely, but if it's behind a
bunch of electronics anyway, the above might be a good solution.
Best regards,
Spehro Pefhany
The series current-limiting cap dissipates no power, so it can be
shorted by a low voltage switch. If it drives, say, the input of a
bridge rectifier or a 2-diode restore thing, short the input of that.
It becomes a bang-bang regulator. If the controller is to be a custom
IC, a modest amount of complexity is free, so some intelligent
zero-crossing thing would be appropriate. But just dumb bang-bang
ought to work.
This is one variation:
https://dl.dropbox.com/u/53724080/Ci..._regulator.jpg
The first diode can be the fet's substrate diode, saving one part.
This is like an alternator regulator that shorts the alternator output
with a triac, to get regulation.
People often add a series resistor, for spike protection, and that
will dissipate some power, but no more than it would with a zener
shunt regulator.
that looks just like a charging circuit I tried once, only I didn't have
a ripple cap. All I had was a - feed back zener and a bjt with a
by pass diode, at the time. I only used a 24V AC source but the idea was
to allow it to have a constant short with out internal heating.
I wanted a pulse type of charging instead of constant DC.
10u
600V SI 30 Hz pulse out
||
-||-+--+------+----+-|-+-+---------------------+
|| | | +
AC line in | | z Max Voltage limit Diode
| | A
| | |
surge clamp z | |
A + HV NPN V Back flow stopper
-(diode too) | | -
5 watt + \| |
| |+--------+-------+
| | .-. +
| + | | ---
| | | | ---
| | '+' +
+------+-----------++------+
|
===
GND
(created by AACircuit v1.28.6 beta 04/19/05
www.tech-chat.de)
Even though that works, I wouldn't suggest using AC over 48 volts
unless you have it totally enclosed.
Never thought of trying a mosfet for that, I have seen some that have
zenering substrate diodes. Those can do the job of both the required
diode and the noise pulse that may exceed the fets voltage.
If I remember there was a little (-) voltage coming back through the
collector but it didn't seem to bother the rest of it.
Jamie