Date: Thu, 10 May 2012 09:29:14 -0400
From: Al Kozakiewicz <akozak@hourglass.com>
Subject: Re: [Amps] The power of an oversized electrolytic capacitor /
how much PS filter C is enough?
Although I understand the point being made, filter capacitors, no matter how
large, do not provide "regulation" any more than a big flywheel takes the place
of a governor on an engine. Regulation requires active circuitry and feedback.
OK, back to nuclear weapons design.
Al
AB2ZY
## really. Then how come my B+ doesn’t budge on ssb..and PO on the pep meter
has increased a bunch.
As far as how much uf required for a given amount of ripple... that’s easy.
psud will spit that value out for you.
The L4B uses 25 uf... and that equates to 3% ripple when sucking 800ma @
2500vdc under load. With 75 uf,
its down to 1%. And yes, my modified yaesu FT-1000 Mk-V will easily RX down
to 40 hz. I can hear any 120 hz
hum component easily on any rx signal.
## as far as diminishing returns goes...... well I have not found it yet !
this will also depend on the xfmr used. light weight
stuff..like used in a sb-220, 100-150 uf and you are maxed out. On a real
plate xfmr, 300 uf works wonders.
## On real big stuff... with real heavy loads.... a new problem arises....and
that’s V drop coming in off the street ! You don’t
have that problem with a 1.5 kw amp. But its unique and exclusive with 8-15
kw pep out amps. In these cases, huge amounts of
C will improve the dynamic regulation by a whole bunch. Its brute force
dynamic regulation..and it works. Then we also have the
v drop across the 50 ohm glitch resistor. On my buddys 6m amp, it does 5.1
kw output. B+ drops 150 vdc with a cxr. Now if the
50 ohm glitch is shorted out, the V drop is only 50 vdc with the same cxr !
With 2a of plate current, there is a 100 vdc drop across the
50 ohm glitch alone. there is nothing you can do about it, and Im not
prepared to use glitch values less than 50 ohms.
## so yes, high C does provide superb regulation...and by default, your ripple
is down to nothing. And your load line is not all over the map either,
hence the better imd. Moot point since most xcvrs have crap TX imd, unless
class A. PO does go up a whole bunch. A 8877 amp will easily
go up another 100w pep..which of course nobody will hear at the other end. On
the bigger stuff, po increases by several hundred watts.
Most of this started when I noticed my power master wattmeter display ..
dithering when a 700w or greater cxr was used..with my L4B.
The peak reading AS powermaster is so sensitive, it will easily show the 2-3%
ripple on the B+. I could also see it on the scope, plane as day.
Drove me nuts, since the xcvr alone is clean with a cxr, no ripple, dead flat
bar on the scope, and no ripple.
## On some of the other amps, the V sag was a bit much. A bigger xfmr is the
1st step..along with bigger ga wire to the panel, and low Z wiring
inside the supply, and low resistance contacts on all contactors, etc. Now if
ur still not happy with the bigger xfmr, then install the bigger caps.
But since big caps are a dime a doz these days, I chose not to mess with small
caps. I can buy 2000-3900 uf caps for the same or not much more $$.
For you folks that go on and on and on abt diminishing returns, why don’t you
toss the arrl book in the trash..and find out for yourselves.
later... Jim VE7RF
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|