Those would be storage for the high voltage DC rail
running the output stage. I got a picture here, of
that section of circuit.
(
http://www.gaoyun.com.cn/en/news/226.html ) # Word salad
http://www.gaoyun.com.cn/data/upload/image/20160428/1461815827911068.jpg
You can see the bipolar supply there, one cap for the positive
rail, one cap for the negative rail. A bipolar supply being
used so the output voltage is pretty close to the center tap "zero"
of voltage when the amplifier is quiescent.
This is an unregulated design. If mains rises too high, it
endangers the life of the big blue capacitors (and of course,
everything downstream of that point). Amplifiers and receivers,
that's probably one of the things affecting their lifespan, is
their ability to swallow electrical insult from mains.
There will be a ratio, between the WVDC ("42 volts") of the
capacitor, and the actual filtered DC on it. Some of the
rating there is a "derating value". The very worst capacitors,
you might take the normal voltage and multiply by five to
come up with the WVDC needed. If you were to stick a
multimeter on the terminals of the capacitor, you might find
a voltage which isn't 42 volts, and does not appear to use
all of the voltage rating. But some room must be made for the
peaks of the rectified voltage from the bridge rectifier and
the capacitor voltage. And some room is also needed for
mains overshoot.
You can see in this example, the manufacturer used an 80V
capacitor, and the measured rail value is 32.5V. Part of the
headroom, is for the top of the rectified sine wave. While the
person here decided to use a 75V capacitor, be aware that
headroom was put there on purpose and it wasn't a "lark".
Maybe this amplifier will be a little less resistant to mains
overshoot, after the repair.
https://www.diyaudio.com/community/threads/amplifier-main-capacitor-replacement-voltage-question.132756/
If you are lucky, the capacitor uses screw terminals and no solder
is used. There may be a lock washer as well, to keep the screw
in place for fifty years. If so, this makes it easier to remove
them. The screw termination helps with the mass of the item and
the potential for strain to be placed on the joint to the foils.
As brands go, Elna is pretty good.
If the caps happen to be small enough to be solder
terminated, and they're a flush fit to the PCB, you'll
then be hoping the holes in the PCB are large enough
so the legs come out easily. Even using a vacuum desoldering
station at work, I've had trouble with that. If you pull too
hard, you will pull the "solder fillet" right out of the PCB.
Yes, I've done that. That's how you learn not to do that,
pull too hard.
Not all of the hum reduction in an amplifier comes from
those capacitors. The capacitors still have mains-related
ripple on the DC output voltage. The amplifier itself
has a PSRR (Power Supply Rejection Ratio) and this comes
from the closed loop feedback in the amplifier and the
balancing of the components in it. The amplifier in a sense,
"resists" the ripple and tries to make the output of the
amp, only follow the input, while ignoring the hum
on the V+ and V- rails.
It is fortunate, that higher values of capacitance, will
not "fit" into the footprint of the capacitor. Electrolytic
capacitors have "microfarads per cubic inch" relationship.
If the substitute capacitor is physically a lot smaller
than the original, that usually means it is a substandard
build. It means the ratings stamped on it might not be
correct. Because of the space limitations in the amp,
you can't replace a 10,000uF cap with a 50,000uF one.
And this is a good thing, because if you do move up too
much, you blow a diode in the bridge rectifier.
There is inrush current, and sometimes good designs, use additional
components to prevent the capacitor from causing the rectifiers
to blow. Like if the amp had a toroidal mains transformer,
there already needs to be an inrush limiter, just because
of the transformer type chosen (info from Hammond transformers
site).
You can replace a 10,000uF cap with a 12,000uF cap if
you want. But don't go nuts. The capacitors themselves
have some "tolerance" to the true value, and it's not
tight like 5% either. They're looser than that. Only
certain kinds of caps can manage to be +/- 5%. Some are
as loose as +20/-80 on percent. And chances are, if you tried
to put 15,000uF, the footprint of the cap might be the
next size up and it won't fit. Always check dimensions
(case size) before you buy, or you'll be sorry. Most of
the complaints I hear, are "dammit Paul, these don't fit" :-)
Yes, been there and done that.
Selecting capacitors is hard work, and one of the reasons
you don't want to be volunteering to replace them all that
often. The catalog has letter sizes, like the 10,000uF
might be a "C" casing, the 15,000uF might be a "D" casing.
Each casing has a drawing with dimensions (or a table of
dimensions) to help in your selection.
Paul