[HBR] about resistors

[Walt Hutchens]

Chris wondered:

> I didn't communicate very well, in that I was also
> (or mostly) interested in the wattage of the resistors.

As a rule of thumb, except for power stages and circuits (the audio output
tube in a receiver and HV rectifier and associated circuits, you can use 1/2
watt.   That'll do for nearly all plate, grid, cathode resistors.   With a
power output stage (say a 6AQ5 or 6L6) it's wise to do the calculations.
Ditto for voltage dividers:  For example you are using 250V on the plates of
a receiver but want only 100V on the screen grid of the IF stages; screen
current is 4 mA so if you just use a dropping resistor you'll need 37.5k
(use 39k) and 150 x .004 = 0.6W;  use 1W for that job.

Modern carbon film resistors hold their value VERY accurately over decades.
If you are using vintage resistors -- the carbon composition type -- check
the resistance before installing as it may be far from the marked value AND
be extra conservative about ratings as they drift (in value) even more
rapidly if hot.

As a general practice, when a resistor handles power  -- say the dropping
resistor that feeds a VR tube for regulated HV -- a wirewound resistor is
preferred over carbon types.   OR (with modern resistors) the metal film
units.  Both of these types of resistor function well at high temperatures.

Generally these issues are greatly simplified by following the schematic and
parts list of an established design.   The designer will specify the sizes,
voltage ratings (and so on) for the parts.   "All resistors 1/2W unless
otherwise specified" but then in the parts list you might find '150 ohm 2W
wirewound' for a cathode resistor.

Even when you strike out on your own you can get help:  Using a 6AQ5 with
250V on the plate as an output stage isn't a new idea and you probably can
find an existing design with parts specs for such a stage.

Capacitor voltage ratings are a little trickier and here it's really best to
go with what your W6TC (or other) circuit calls for.   Electrolytic caps
need to be rated for more than the highest voltage they'll ever see;  that's
usually taken as 1.4 x the AC voltage to the rectifier since in nearly all
circuits it is possible for rectified voltage to appear on the filter caps
before (or without) a load to pull it down to the normal output.   125V line
operated HV,  peak voltage would be 175V:  Really should use 200V or 250
electrolytics for the filter.    The 160V rated caps will GENERALLY do fine
-- that number was chosen to handle AC line rectifier circuits -- but it's
not conservative.

However you don't want to use too high a voltage rating:   If you go more
than double the normal operating voltage -- say over 250V in the example
above -- an electrolytic cap may not function properly as it depends on the
applied voltage to form the dielectric.

Okay, you have a local oscillator circuit, nice hi-Q coil (W6TC set!) and
you need a fixed capacitor in parallel with the tuning cap.  You have some
of the 50V ceramic caps in the proper value: Is this a good part for this
job?   

Probably not.  Even receiving oscillator tank circuits have fairly high
currents flowing between the coil and the capacitor(s) across that coil;
that little dot of a cap may be handling 100 mA or more of RF and it will
drift like crazy.   Tank circuit capacitors for normal vacuum tube
oscillators should be the largest physical size that's convenient AND
designed for stable capacitance as temperature changes.  Silver mica (the
electrodes are plated onto the mica dielectric) and the larger zero
temperature coefficient ceramics are the best choices.

Here, however, you may not always get help from the parts list provided by
the vintage circuit designer because those itty-bitty ceramic caps were not
yet on the market:  Pretty much any cap that W6TC could buy was big enough
to work okay in a receiver oscillator.

This problem snarls out loud if building a vintage transmitter VFO:  Power
tube oscillators run even higher tank circuit currents and the capacitors in
these circuits must be chosen accordingly.   Any number of ham designers
have been amazed at the improvement in VFO stability when a single cap was
replaced by (say) four of 1/4 the size in parallel ... by sharing the
current, they suffered less heating.

Walt Hutchens
KJ4KV