[HBR] Transformer

[Walt Hutchens]

> In the back of the latest Antique Electronic Supply catalog, there's a
> replacement output xfmr, made per request of the CCA guys.  It's 600 ohms in /
> 8 ohms out / 12 watts.  In case you need one.

This transformer is intended for the Collins military receivers -- the R-390
would be an example -- that have a 'line' audio output at 600 ohms and thus
need another transformer to get to speaker impedance.   As an output
transformer in an HBR-set something more like 3000 or 5000 ohms to a speaker
value (4 or 8 ...) would be appropriate.   The AES transformer # T-P31 would
do (and at a very good price, as these things go), also their # P-T125ASE
or  P-T125A.

Output transformers are also easily improvised if needed.   Filament or
other small power-type transformers are fine if you use the right turns
ratio (square root of the ratio of impedances -- a 6.3 volt filament
transformer is perfect for 3200 in, 4 ohms out) and realize that the allowed
audio power level is a fraction of the rating for power use.   Ideally you
take the core apart and re-stack it with an air gap (layer of cardboard
between the I's and the E's) to reduce saturation but at the usual small
receiver output power levels -- 100 mW or less!  -- you'll never notice the
difference.

Remember that in communications service the transformer won't see
frequencies below about 300 cps.   A filament transformer that runs at its
rated level in 60 CPS power service with severe core saturation (because who
cares about a bit of waveform distortion on the filaments) will do fine at
1/10 the power and nothing below the usual comm voice frequencies.

(If you're into 'hi fi' audio -- lots of low end and such -- you have to
take a more serious approach to this: Push-pull output, the next size or two
sizes larger transformer AND one designed for audio service, next size
larger tubes and a music-grade (rather than comm grade) speaker.    But only
one station in 100 will be able to transmit any difference you can hear,
except, of course, that thunderstorm noise will be able to make the plates
jump in your kitchen.)

Impedances need not be too closely matched:   Over the range of (say) 3000
to 5000 in, 4 ohms out, the practical effect will be a difference in volume.

The Radio Shack power transformers are ideal in improvised audio output
applications because they're bobbin wound and thus easy to take apart and
reassemble.   Also you can do custom winding on them.   ('Easy' is relative
-- they are varnished and you have to break that loose.   An old pocket
knife and a scrap of hardwood as a mallet work pretty well.)

The use of 'high fi' circuitry -- start with a bit of extra gain in the
audio stages and reduce it to the level you actually need through the use of
negative feedback from the voice coil winding to the cathode of the 1st
audio -- ought to be universal.    A voltage gain of 10 from input to 8 ohm
voice coil is about right so if you use a pentode-triode or triode-pentode
combination for audio you have plenty.   You can hear the difference
instantly: Such receivers just SOUND better, speech is easier to understand,
and so on.  The cost is under $1 in extra parts and two minutes during the
assembly process to determine which way to connect the feedback.

The high-fi boys take this much farther, using specialized (very expensive)
transformers and much more gain inside the feedback loop.    However doing
this requires both the expensive parts -- AES has LOTS of this stuff -- and
a serious design in order to prevent instability.  (The designs can be found
in the RCA tube handbook circuit diagram pages.)  The big payoff in
communications service comes from the first few dB of feedback, usually
available with no more than a couple of extra 1/2 watt resistors.

No other receiver design tweak that I can think of has so large a payoff for
so little input.   

Walt 
KJ4KV