Homebrew Breadboard Receiver
2-tube Regenerative Shortwave Receiver
This
was a new idea for a project. I had access to some of Bill
Turner's tennis-ball type of glass globes for replacement tubes and
wanted to do a breadboard project that would show them off. I
settled on a basic 2-tube regenerative circuit for simplicity and
although its an anachronism for this style of construction I went with
shortwave coverage.
I first built up the set as test using
01As to see how it worked.
It
worked well enough to go forward. Smooth tuning,
smooth regenning, was copying a few stations on 31M this afternoon at
2-3PM so its working perfectly ok in that mode.
But with the 5676 fake tubes it misbehaved terribly. Too much gain
and although it worked it was full of howls and squawks. I cut back the
tickler winding
(sacrificial coil) and that helped some, tried the typical bypassing of
things, the fringe-howl resistor on the AFT, etc, etc. Everything
seemed to have some effect...which sometimes is a bad sign! Very picky
about grid leak value.
I eventually settled on 18 volts for the detector plate and that seemed
to calm things.
The audio stage runs at around 67
volts.
The above photo is my plug-in coil
assembly made with tip jacks and plugs. The coil became a point
of interest. My original coil was wound with #16 copper and I
wanted to use some pretty "German silver", aka "nickel-silver"
wire. It did not occur to me at the time that the German silver
has very poor conductivity and it was pointed out on one of the boards
that it could be a bad idea. I tested the coils for Q (copper vs
NiAg) and indeed the copper rated about 2-3 times better!
Nonetheless, the large air-core coils (these are 3-1/2" diameter) with
the lesser wire are still superior to typical plug-in coil forms wound
with #22 copper, for example. Actual testing on the air revealed
very little difference. Too close to call actually. The
NiAg coil DOES require more regeneration than the copper but it was
still within range of my tickler cap. Its also more tolerable of
higher detector voltage without causing instability in the
circuit. So clearly there's some loss involved but nothing that
cannot be compensated for in the tweaking of the circuit. I tried
very hard to find some difference in selectivity in actual operation
and there was no difference to be found. I suspect the total
loading of the circuit tends to even out the differences in the Q of
the coil.
.
A bonus of the large coils is that the
radio works pretty darn well with NO antenna
connected! That big coil does a respectable job as a 'loop' antenna and
can serve well for casual listening.
At first the audio output was
disappointing but I changed to a better set of phones and relegated the
old Cannonball phones that I had been using to the junkbox.
Interstage xfmr is an old MusicMaster type and it sounds good and is
plenty efficient.
The tuning caps were a lucky find on
ebay. I got 7 of these for 20 bucks because nobody else bid on
them, most likely because of the odd shaft and the fact that they were
in terribly ugly condition. The cap has a one plate trimmer built
in which was ideal for my use. I removed plates from the main
section to lower the capacitance for shortwave tuning. They were completely disassembled and
cleaned up. Very good quality caps.
The National B-type vernier was ideal
to use with these caps. The smaller, inner vernier shaft pokes
through where the National dial normally employs a decorative
'hubcap'. So the net result incorporates the very nice National
vernier tuning plus the added one-plate trimmer for fine tuning.
That makes for a very comfortable tuning arrangement. The
regeneration throttle cap has the same thing.
By the way, the breadboard layout didn't cause any problems at
shortwave frequencies. I tested up to 20 Mc/s but by then hand
effects were starting to show up and the tuning range was simply too
tight.
The next phase of
this project involves a matching 2-stage breadboard audio amplifier.
(That page is still under construction)
...
10 Aug 2008
