[HBR] Another one-week (okay, probably not!) HBR -- Part 2

[Walt Hutchens]

The metalwork and the 'first mounting of parts' (as Heathkit would
say) are done. Pictures are at:

http://picasaweb.google.com/timbreblue/Walt#5254833299643465026
http://picasaweb.google.com/timbreblue/Walt#5254833820737451154
http://picasaweb.google.com/timbreblue/Walt#5254833430924088786
http://picasaweb.google.com/timbreblue/Walt#5254833472562276914
http://picasaweb.google.com/timbreblue/Walt#5254833653676698194

In the front view, the controls along the bottom are (L to R) BFO
tuning, ON/OFF switch, AUDIO GAIN, TUNING, and CAL (upper)/AGC
THRESHOLD (lower). The big hole on the left is for a 3" speaker and to
the right of the S-meter, switches for STDBY/OPERATE, BFO ON/OFF, and
CAL ON/OFF. Holes under the handles are for latches to hold the set in
the (to be built) cabinet.

Hinged top AND a drawer under ... now THAT will be a project.

The shield on top of the chassis is the upper half of a 3" x 4" x 5"
minibox.

>From the rear, the antenna coil is in the shield on the left, the RF
stage is the tube just to the right of there, and the mixer grid coil
is next toward the panel. The mixer is made up of the triode halves of
the two 19JN8's for which the pentode sections are the RF and
oscillator stages; the oscillator tube is behind the antenna coil in
the shield and behind that, the 1st IFT. There is a shield plate under
the chassis separating the antenna coil and RF stage from the
oscillator/mixer tube and mixer grid coil.

The oscillator coil is forward of the shield, next to the 1st IF
stage. The coil is sub-mounted to reduce heating from the chassis and
allow air flow around the coil -- there will be a way for air to come
in under the chassis. The two stages of IF run to the right from the
oscillator coil, ending at the 1st audio and plate detector tube. The
2nd audio and 3500 kcs marker oscillator stage is to the rear from the
detector tube. The 9-pin socket nearest the panel at the right is the
BFO and bias rectifier; the empty octal socket is for the 117Z6
rectifier.

The mounting of the dial and tuning capacitor is never easy but it is
especially tedious for the Eddystone 898 because exact alignment has
to be obtained with the capacitor mounted above the chassis. Laying
this out is a challenge. The shield holes on the chassis are slightly
elongated left/right to allow lateral adjustment; the holes in the
mounting ears on the tuning cap are elongated (by the mfr.) to permit
up/down tweaking.

The most obvious weakness of this layout is that the leads from the
oscillator tube to the coil are about 5" long. Long tuned circuit
leads are always an issue with plug in coil designs but I could have
done better by moving the first IFT out of the above-chassis shield
and moving the oscillator (and 1/2 mixer) tube closer to the coil.

It'll be a few days before there's anything else to report.

Peter said:
> I've been looking at the Star SR-700A receiver schematic

Thanks for that; I didn't know of this very refined design. I'll watch
for one on Ebay ... would be interesting to put one on the air!

Two striking features: The use of cascaded 55 kcs IF filter coils that
are detuned by switching capacitors in order to vary the selectivity
without moving the carrier edge of the passband. THAT is the way to do
adjustable selectivity if you're going to use a low IF to get narrow
bandpass; you could even apply it to use of the 85 kcs command set
IFTs.

(But the 'without moving the carrier edge' trick is only useful if you
also arrange for both LSB and USB to be LSB at this point; most
probably by switching the final local oscillator between above/below
the previous IF to select the sideband.)

And the Vackar local oscillator. While there is no magic 'stable
circuit,' the Vackar gives the best control over coupling between the
tube and tuned circuit of any of them, thus simplifying the job of
minimizing the effect of tube characteristics on output frequency when
using high gain (modern) tubes.

In a well-designed Vackar circuit at moderate frequencies, the only
drift you have to worry about is in the tank circuit parts: plugging
in a cold tube gives you the same frequency as the warmed-up one did.
So if the coil and capacitors are protected from temperature changes
and/or compensated, you get a very stable oscillator.

One improvement to the circuit shown is to tap the coil at the point
of zero RF voltage and feed the plate voltage through a resistor at
that point with no bypass cap. This eliminates the plate choke which
is effectively part of the tank circuit and thus a source of drift.

Push-pull oscillators also offer great advantages, starting with twice
the gain for the same (effective) tube capacitance.  However they're
best used with a push-pull mixer.

In the tickler feedback circuit of the W6TC designs, coupling to the
plate (actually, screen) circuit can be varied by adjusting the number
of turns and spacing of the plate coil, but the full grid capacitance
appears across the tank.

Using the screen of a pentode as the oscillator plate (as in the ECO
circuit) gives you a low gain oscillator tube anyhow, so the amount of
decoupling that can be done is probably quite limited. The ECO gives
up this advantage in exchange for the simplicity of a built-in buffer
and (in the W6TC designs) 2nd harmonic generator.

Since the major source of drift in these designs is likely to be the
plug-in coils, it's probably not worthwhile to spend a lot of time on
the oscillator circuit.  Good coils and a chassis layout that keeps
the oscillator coil as cool as possible are likely to be the most
useful focus.

Finally, I did find out why I can't post to this list using my regular
e-mail program. QTH.NET assumes that anyone using The BAT (my mailer)
is a spammer and rejects all such postings with a meaningless message.
I can post normally by using Outlook Express, or probably (with a bit
more work) by zapping the code that labels the mail files, as I'm sure
the serious spammers do.

Go figure.

Walt
KJ4KV