In your answer to Geff Waite, I was glad to see
your comment about the capacitor ripple current.
Too often it is ignored. Unfortunately, it isn’t quite as
simple as you said. Different manufacturers, different
series from the same manufacturer, and different cap
values have different ripple current ratings. A few of
the caps in the Mouser and Digi-Key catalogs have
ripple current ratings. Most manufacturers’ capacitor
datasheets give the ripple current ratings. The quickest
way to get the datasheets is to do an Internet search
of the manufacturer’s name and the capacitor series ID.
A quick look at the Mouser catalog gives some
examples. For Xicon snap-mount, series LS, a 25 volt
4,700 µF cap is rated at 2.31 amps ripple current,
while a 63 volt 4,700 µF cap is rated at 3. 49 amps, and
a 200 volt, 2,200 µF is rated at 5 amps! For nearly every
cap series of every manufacturer, the ripple current
goes up (and the ESR goes down) as the voltage
rating is increased. Because of this, I try to stay away
from low voltage caps, even in low voltage circuits.
(This difference may be primarily from the can size;
a larger can has better heat dissipation.)
For the 10 amp DC output of Mr. Waites’ power
supply, the ripple current would be 13 to 15 amps,
depending on the resistance of the transformer
windings and other factors (including the effective
series resistance of the AC power line). For any power
supply, if the caps seem to run warm, the ripple
current may be too high. The easiest cure is to parallel
more caps. The actual ripple current can be measured
fairly easily. Connect a 0.1 ohm or smaller, 1%, three
watt resistor (available from Mouser) in series with
the cap, measure the voltage across the resistor with
a true RMS meter, and calculate the current.
We might also mention the current rating of the
transformer. There has been some discussion of this
recently. For a bridge rectifier and 10 amps DC output,
the transformer should be rated at 16 to 18 amps
RMS (depending on which transformer manufacturer’s
guidelines you follow). And there’s always the
question, build or buy? Marlin P. Jones and Assoc.
( www.mpja.com) sells a 24 volt 12. 5 amp switching
power supply (no. 16489-PS) for $97.
— Bill Stiles
Thanks for the feedback, Bill. I had not noticed
that current rating goes up with voltage rating.
In response to your statement that “There is no
quality of sound advantage in hybrid digital; analog
FM stations already transmit more bandwidth than
most people can hear and with low distortion,” I
should like to invite you to my home in a northern
suburb of New York City where I, a classical musician,
can find virtually the only listenable FM by pointing
my 14-element beam at lower Connecticut. There I
will demonstrate what happens when I tune WSHU
at 91.1 and wait for my Sangean HDT-1 tuner to
recognize the signal as digital, at which point the
background noise drops out, the separation increases,
and all I hear is music.
I am now waiting for the one “classical” station
from New York City to go digital so I can receive it
without the multipath distortion which makes it
— Robert Voss
Not having had the experience of multipath
distortion with FM radio, I was not
aware that it was a problem. I get
excellent reception of classical
music from WCRB, Boston at 99.5,
50 miles away. I believe WCRB is
hybrid digital also; you should be
able to receive it.
AThe schematic (Figure 10) is
modified to charge 20 cells.
I added R9 and D5 because
33 volts exceeds the input
rating of the 78L05. I also added R10
to reduce the input to ground (ADJ)
voltage. R10 is not strictly needed
because the 317 is rated 40 volts
input, but it shows what needs to be
done if the voltage goes much higher.
I changed the voltage divider, R4 and
R5, to provide three volts max to the
A/D. The A/D voltage only has to
stay between 1V and 4V as the
batteries charge. NV