-=> On 03-30-98  06:45, Ralph Mowery said to All,<=-
-=>"About Filter Cap values in...,"<=-
-=> From: Ralph Mowery 
-=> Reply-To: rmowery@ibm.net
 
Hi, Ralph;
 
 RM> The capacitors used as filters are not usually that critical.
 
True.  The only time the value is critical is when it's part of a
tuned circuit, such as an audio filter, and I have only seen that done
a couple of times.  (Yes, an audio LC tuned circuit with a very large C
and a fairly high DC voltage accross the C!)
 
 RM> You want to make sure the voltage rating is equal or greater than the
 RM> origional voltage.  I don't like to go over twice the voltage value as
 RM> sometimes the capacitors will not form correctly.  The plates of an
 RM> electrolitic will actually 'form' or take on differant properties as
 RM> voltage is applied to them.
 
Quite right.  When the applied voltage is different than what was applied
during manufacture, the dialectric, (aluminum oxide), will reform to the
new voltage over time, which in turn will change the capicity.  If the
applied voltage is lower than used during manufacture, the oxide will
get a bit thinner, causing the capacity to go up a little.  The opposite
is also true.  This effect will not change the operation of rectifier-
filter circuits signifigantly.
 
 RM> As to the value, in a pinch, you can go
 RM> somewhat lower in value, but it is best to go up.
 
Going more than 10-15% lower than the specified capicity will probably
result in insufficent filtering.  In the input filter, this will result
in lower DC output voltage, and will also likely result in hum in the
output if any of the sections are much less than specified, especially at
full load.
 
 RM> Again, I don't like
 RM> to go over two times the value, but it probably would not hure.  You
 RM> will get beter filtering, but it also puts more of a strain on the
 RM> other power supply components as the capacitor charges.  Usually not a
 RM> problem, but if you go way over, say 10 times the capacity, then you
 RM> could stress the diodes  and resistors in the supply and you might also
 RM> have to go to a slow blow fuse in the supply as the flter acts as a
 RM> short circuit tuil it is charged.
 
A very good point!  Most modern rectifiers using silicon diodes can handle
a little added charging surge.  Back in the days of selenium rectifiers,
which are quite sensitive to surges, a small resistor, usually around 100
ohms at a couple of watts, was placed between the rectifier and the input
filter to limit the charging surge.  This could also be applied to silicon
diode rectifiers.  Considering that the voltages are lower and the
currents are higher in transistorized equipment, I would recommend a
smaller value resistance, such as 2-10 ohms, 5 watt resistor for a typical
12V DC supply, depending on the supply's current rating.  But only make
this mod if you start blowing diodes after increasing the filter cap
values.  Other options would be to either use a series resistor with a
high negative temperature coefficient (sp?), which would allow the caps
to charge more slowly, or to use a resistor with a relay in a time delay
circuit - after a delay of several seconds, when the caps have reached
partial charge the relay pulls in and shorts out the series resistor.
But this gets into more major surgery than most people wouldn't want to
perform just because they had replaced the filters.
 
73 DE KB9QPM
   Ivy
 
 
 
... I wonder what happens if I touch these two wires toge>ZZZZZZAP!!!<
 
           --
 
           ÚÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ¿
           ³  SysOp: Ivy's WALL BBS - 6 Networks 920-457-9255  ³
           ³             IMHO: 5 WPM for EVERYONE.             ³
           ³           KB9QPM/AE as of 11/19/97 :->>           ³
           ³Happy New Year to All! May the band always be open!³
           ÀÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÙ
 
___ Blue Wave/QWK v2.20 [NR]
--- TriToss (tm) 1.03 - (Unregistered)
---------------