RadioBanter - View Single Post

Michael Black[_2_] · November 11th 11, 05:14 PM posted to rec.audio.tubes,rec.radio.shortwave

On Thu, 10 Nov 2011, wrote:

With the survivalist market as well as the DIYers who would build a
kit I have given thought to the idea of building a new tube shortwave
receiver as a usable, practical set.

But since you never specify why it should use tubes, you sound like a
kook.

I assume you are thinking tubes so they won't be damaged by EMP. But you
have to consider if that's a real reality, or some fantasy. There are
loads of reasons why someone might want to be prepared, without coming
close to a nuclear blast. But those other reasons might much rather have
a battery operated radio rather than the high current drain of tubes.

I can stockpile batteries for a solid state shortwave receiver, I can keep
some larger batteries on hand as an external supply, I could run a low
current receiver off a solar panel, there are lots of options. But once
you start drawing current to heat those tube filaments, you are really
stuck. Yes, you can use an inverter off a car battery, but then have to
keep charging the battery. Note that in the old days, running tube
equipment in the car, you mostly had the car running, so there was current
coming from the alternator, rather than just relying on the battery.

That means no regens, no DC bull****, and no plug in coils. It must
have production grade RF and IF coils, a bandswitch, and require
alignment. If sold as a kit the builder will need a RF generator and a
scope (or a spec an or CSM with a track gen).

Why? You really havent' specified what you want, you are then jumping
into fine details. For emergencies, it may be a really useful choice. But
you are being wishy washy in your criteria, so who knows. A regen is
lousy for regular reception at this point in time.

No, you don't want a direct conversion receiver, since those are best for
CW and SSB, not great for straight AM (which presumably is your target).
But once can get pretty fancy with DC receivers, even including proper
reception of AM. It will get more complicated, but proper design requires
looking at multiple possibilities, and since every design will be a
tradeoff, you need to take off your blinders and look at possibilities
before deciding something is more suitable.

Note that the regen is "direct conversion", at least once you kick it into
oscillation. And there were various designs of "direct conversion" in the
earlier days of radio, though not called "direct conversion". Even in
1961, there was a tube based direct conversion receiver in QST.

The early wave of direct conversion solid state receivers often
compromised. They'd be direct conversion on one band (or maybe not at
all) and then a converter ahead of it, which made it a superheterodyne
receiver, albeit with no IF selectivity. There are some points in that
favor.

Indeed, many a good receiver was made with a single conversion receiver
tuning a fixed band, and then converters ahead of it (lots of homebrew
receivers, but also classics like the Collins receivers). That meant the
local oscillator could run at a low and fixed frequency, rather than a
wide segment (traditional single conversion to 455KHz receivers had about
a 2:1 tuning range on each band), so you can have good calibration, and
good tuning, the oscillator running at a low frequency and not needing to
be switched in frequency from band to band (problems in that alone). The
problem was that it meant a crystal for every segment you wanted to tune
(got around initially by choosing which segments, nobody says you have to
have all 30Mhz of the shortwave band), though later synthesizers fixed
that. Of course, there was also the Wadley loop that sort of synthesized
the first oscillator, at the cost of an extra mixer and complicated
circuitry.

For that matter, one popular method of getting a shortwaver receiver was
to get a car radio (they often had better selectivity, and better image
rejection along with better sensitivity, plus good tuning) and put a
converter or converters ahead of it, getting double conversion. Leave the
bulk of the construction to the car radio manufacturers and just build the
converter, a relatively simple task. This is now harder if you can't find
a car radio with analog tuning, since the 10KHz steps of a synthesized car
radio is not the 5KHz that shortwave broadcasters use (and even 5KHz is
too wide for the ham bands).

Note also that in the thirties there were the "supergainers", regen
receivers with converters ahead of them (or looked at differently,
superhets with regen receivers as the IF), a fusion that provided some
advantages. Even in the solid state era you'd see those in the ham
magazines, sometimes people even putting crystal filters before the regen
detector.

No plug in coils? Then again you haven't stated your prime criteria (no
plug in coils in not criteria, it's the result of some criteria you
haven't specified. In the old days, the bandswitch often was a key
problem in a multiband radio. It had to switch LC circuits at
increasingly high frequencies. The switch often got in the way, and
physical layout was determined by the bandswitch (though some companies
bult the bandswitch for the receiver, so the layout could be better).
Coils have a simplicity, though of course that doesn't include fast
bandswitching. The HRO used plug in coils right up till the point of
solid state, and many thought that line was a great receiver. All those
recievers with converters ahead of them meant one could plug in a
converter per band, rather than switch LC circuits. More expensive, but
if you use transistors the solid state devices dont' add much to the cost,
unlike tubes that were costly and bulky.

Your fantasy designing has overlooked the home builder's need to align the
receiver. Fixated on the way things used to be, you havent' considered
that if you spend the money differently, it may make user alignment
simpler.

It should use off the shelf parts even if those shelves are bare, as
it is better to copy an existing item than design from scratch. I
would clone the Eddystone dial mechanism and the bandswitch and coils
from some Hallicrafters or Hammarlund set, they could be sold as
desperately needed replacement spares for the old sets too. I would
use a seeing eye tube mounted in a hole in the dial as opposed to a
meter movement, again, getting a run of new tubes made is possible if
you are buying several thousand. There are some surplus that could be
used if really needed too.

YOu want to construct a fifty year old receiver. There are virtually no
off the shelf parts left for those. There aren't shelfs or local radio
stores to sell those parts.

You havne't made the tradeoff between old and new. That crummy Eton hand
held shortwave receiver I got at a garage sale in September for 2.00 is no
better than the junk solid state Hallicrafters S-120A receiver I spent $80
for in the summer of 1971. But, it uses an IC to provide a frequency
counter, which means one can actually have good frequency readout, without
all kinds of expensive dials and calibration. The IC is dirt cheap, the
Eddystone dial if it was still being made today would be terribly
expensive. The dial seems simpler, but this is a case of complication
making the overall design far simpler. Oddly, that hand held Eton radio
does take more advantage of having a frequency counter on board, they
break up the tuning into smaller segments (since they don't have to
calibrate a dial, or provide space for a whole bunch of bands, why not?)
which means the simple tuning pot is not too obnoxious.

I would use a separate power supply and speaker for several reasons.
I would have the radio take in B+ and heater voltage and put out 600
ohm +4 audio. A regular supply could be used at home or car battery
and a switchmode brick for B+. A headphone jack would be supplied off
this tube.

A solid state receiver would run off a battery, and the built in supply
wouldn't be a burden when unused. If you're really stuck with tubes, why
not get original, wind your own transformer, then have the AC coming from
the wall turned into DC and an oscillator that feeds the transformer.
Running at a higher frequency, the transformer can be smaller. But, done
right, you can have another oscillator that runs off 12vdc, and that feeds
a diffeent winding of the transformer, so you've got your dual mode power
supply without making two supplies. You haven't throught this through, you
haven't done nearly enough wide thinking. You are just trying to
duplicate the past, without any great reason for it.

The set should cover 500 kHz to 30 MHz, AM, SSB and CW, with a
product detector of course. A 455 kHz IF is needed so as to use common
mechanical or crystal filters, which are optional. There should also
be a 455 kHz IF out for an external synchronous detector.

Your bias is there. For at least fifty years one could get good
selectivity in the shortwave frequencies, initially down around 2Mhz but
then 9MHz became kind of standard. Even tube receivers were built with
such "high" IFs. Right away you get rid of the problem of image
rejection, with such a high IF the front end selectivity is lessened
(especially for the higher bands). Alignment is simplified, indeed a
single conversion receiver with a 9Mhz IF no longer has to gang the local
oscillator tuning with the front end tuning, there's no longer a problem
of an image 910KHz away (I've seen reviews for single conversion 455KHz
low end receivers from the old days, I remember one said "we couldnt' tell
which was the image and which was the real signal, they were of equal
strength"). You lose some segment around the IF frequency, but chosen
properly you won't miss much. Of course, 9MHz IF filters are more
expensive than 455KHz ceramic filters (which is what many receivers use),
which can then be a problem if you want multiple bandwidths. One way to
get around that is to go with a high IF and then a low IF, though not
without tradeoffs. Note that mechanical filters are not inexpensive, that
is only the case if one finds one on the used or surplus market. Crystal
filters at 455KHz were commonly single crystals, a good peak but not so
great skirt selectivity, and they'd provide multiple bandwidths by loading
the crystal down. One might as well go with a higher IF and make ladder
crystal filters, one for SSB and another for AM.

Or put a phasing system at 9MHz, with a relatively wide (and cheap)
crystal filter ahead of it (10.7MHz may then be better, it's not a common
frequency for narrow bandwidth filters but is common for narrow FM
bandwidth filters). A good phasing system will knock out the unwanted
sideband, and audio selectivity will be effective. Done properly, it can
be synchronized to the incoming signal for AM reception.

Or use a high IF and then 455KHz. This has an advantage that you can add
some tuning to the second conversion oscillator, making it fine tuning.
Some portable shortwave receivers used this scheme before the move to
higher first IFs. Indeed, with a synthesized oscillator (well not if you
still want tubes), broad steps make it simpler to design and build, and
having a fine tuning on the second conversion oscillator then fills in
between steps.

Go back and figure out your design criteria. Who will want this? What's
the point? A practical receiver for now is different from the nostalgia
of the old days. Simple to build may mean a simple receiver, or it may
mean adding complication in order for the end builder to have little
problem assemblying it or aligning it. Adding extra stages may add cost,
but may some other point simpler.

And don't assume single conversion to 455KHz receivers were the ultimate
in design. They weren't, they were tradeoffs and some got around problems
by making them more complicated, others lived with the problems.

Michael

Any other comments?