Re: B+ rectifier question

Hey guys,
I appreciate all the input. I replaced all the bumble bee capacitors.
I replaced all the electrolytic capacitors (cans if possible). I couldn't
find the can for the filament circuit so I had to go with 3 separate
capacitors for that circuit. I went with keeping the capacitor voltages the
same as the originals or at least close. I replaced the resistors in the
power supply as well as the filament bridge which was originally 4 separate
diodes mounted on a strip and I replaced the original 2 diodes for the B+.
I have to run the preamp on a variac at 110v because the filament voltage
tested at each tube is right at 12 volts. My goal is to get the resistance
added necessary to be able to plug the preamp into the wall outlet about
123V but I also don't want to add too much more heat to the innards. The
filament circuit pulls a lot of current. I've got a 6 amp bridge for the
filament circuit and it runs warm. The resistors (2-7.5 ohm 5 watt) run so
hot you can even touch them. But they were that way before the new parts.
The preamp at this point sounds great. I've used it running from the variac
for a number of months now with no problems and the PT runs just a little
warm. I have a number of resistors to be able to experiment. It's just
getting the thing unhooked and out of my cabinet. I know I sound lazy. I
guess I just don't like the idea of getting it out again. It was a lot of
work at time to (rebuild it). You guys know what I mean? I have the
schematic so I know the recommended voltages. I know just about enough
about electronics to be dangerous. If you go to www.nosvalves.com/C22.htm
you can see the 2 grey B+ diodes at the left of the pictures. I still have
them but they don't have stacks per se like I've seen some other selenium
rectifiers have. Do you recommend for the B+ using resistance before the
diodes or after? Also the 2 HT wires Red and Red/Yellow; do I need to
install a resistor between the capacitor and the Red/Yellow wire; the
voltage is running a little high, or will reducing the B+ voltage take care
of that?

Thanks again,
Edward Morris

"Ian Iveson" <IanIveson.home@xxxxxxxxxxxxxxxx> wrote in message
news:fcw1l.11578$QG4.10939@xxxxxxxxxxxxxxxx
Peter Wieck wrote:

Just been trying to find out about selenium rectifiers. It
seems they can have several cells in series, and each cell
can vary in several ways. The voltage drop can be
measured,
or could have been estimated from an appropriate data
***
which is unlikely to be available now. If you can measure
the drop across and current through each one, you can get
an
idea of what resistor you would need in series with each,
using ohm's law. The waveforms aren't sinusoidal, though,
so
you would need to measure the two final voltages and
adjust
to suit, depending on how exact they need to be. You would
also need to be careful with power ratings, for the same
reason.

A much easier way is to measure the resistance and
open-circuit voltage of the transformer secondaries, and
then simulate the circuit with SPICE. Especially if you
don't happen to have a stock of real resistors to try.

If you can measure but are unable to simulate, someone
here
might do that for you.

Or perhaps someone has already made a near-enough guess?

There is a question of strategy. Would it be better to add
resistance to the heater circuit after the feed to the HT,
and to add no resistance to that feed, thus raising the
floor of the HT circuit. Then add resistance to the HT
after
the doubler?

It would be a bit interesting to find the optimum
arrangement.

***
WOW ------ WAY, way to complicated.

Selenium rectifiers have a much higher voltage drop than
silicon
diodes. In *some* (but not all) circuits where the down-line
voltage
is critical, that drop must be replaced if the selenium
diodes are
replaced with silicon.

So far, so good.

Not all selenium rectifiers are created equal. Some are
single-
element, some are stacks of 2 - 20-or-more depending on the
current
they are to carry. The specific SIZE of the diode does not
matter, the
NUMBER OF ELEMENTS (stack) does matter relative to this
question.

Rule-of-thumb, all other things being equal (Ed's ballpark)
is 10 ohms
per element. After which additional resistance *may* be
desired to
further reduce B+ (or C- if a bias circuit) to make up for
higher line
(wallplate) voltage common in some areas these days.
Naturally, this
is to be verified in the field by actual measurement.

Ed was requesting ball-park numbers for a single-element
system. 10 -
20 ohms covers the from home plate to far-away center field.
Close
enough.

Sometimes a small brass hammer in the hands of a Vermont
Blacksmith is
all that is necessary.

***

The situation is no less complicated than I wrote. You may
believe that the complications don't matter, and Edward may
not care.

We don't have ball-parks here in England, but anyway, if
Edward asks what's the right ball-park voltage for his
heaters, would 35V be in the right field? How close is close
enough for Edward? We don't even know the voltage ratings of
his capacitors.

If the voltage drop across a selenium diode is truly "much"
greater than that of a silicon diode, and particularly if it
is a true diode drop rather than the result of inherent
series resistance, then the resistance necessary to restore
it will depend on the current through the diode. So if, say,
the current is 1A, then 10 ohms will result in a 10V
reduction, whereas if the current is 10mA, then 10 ohms will
result in a drop of only 100mV, each assuming DC or
sinusoidal current.

Now, the circuit in question includes a relatively
high-current heater supply, and a low-current HT supply, and
an even-lower-current link between the two. Further, the
heater supply uses a bridge rectifier, whereas the other two
are half-wave. Further still, the high-current supply is
relatively low voltage, so the greater voltage difference
will be a much more significant proportion of the supply
voltage. What is more, we don't know the specifications of
Edwards selenium diodes, although the knowledgeable may be
able to divine them from the voltages.

If 10-20 ohms is a good guess for one, it is unlikely to be
a good guess for the others, surely?

Included in your answer is the assumption that Edward
doesn't care too much about the consequent voltages. Maybe
you're right, and maybe they will be close enough for
reasonable operation.

In which case, considering 20 ohms will have so little
effect in the two lower-current situations, he could leave
them out and concentrate on getting the heater voltage
right.

What power rating should the added resistor(s) be, there,
and exactly where in the circuit should he put it, or them,
do you think?

A simulation would not be complicated at all, BTW, unless an
exact mean power dissipation is required because that's
awkward to do.

Perhaps Edward will report the results of his experiments.
I'm quite interested.


I wonder how you should dispose of the old rectifiers?
Authorities here are, quite rightly I suppose, getting
very
touchy about what goes in the bin. I noticed, passing a so
called "health food" shop the other day, that selenium is
essential for fitness, so perhaps you could grind them up
and eat them?

***
In our township, one may drop "hazardous waste" at the
township
transfer station for separate disposal without additional
charge. In
general, I collect this sort of stuff in an empty (clean)
paint can -
when it gets nearly full (perhaps once per year), I drop it
off at the
transfer station about 2 miles away. From there it is taken
to a
certified hazardous-waste landfill for proper disposal.

***

Probably the same here. I haven't managed to fill my jar
yet...just one dead valve and one extremely dead tantalum
capacitor. If there were a damaged selenium diode in there I
would want to get rid of it sooner.

Ian



.

Loading