From: "Geoffrey G. Rochat"
Subject: Re: Digital divide by ten, 1949 style
Date: Sat, 21 Sep 2002 02:08:49 -0400
Organization: Sprint Advanced Network Services
References: <3D84D4B3.50B511EF@webaccess.net> <email@example.com> <3D893F55.1712189F@mfi.net>
NNTP-Posting-Date: Sat, 21 Sep 2002 05:53:07 +0000 (UTC)
X-Newsreader: Microsoft Outlook Express 4.72.3110.1
>That's a whole *color* TV, with fewer tubes than the average B&W TV of
Awww, heck! I've got a color TV in my den, and it ain't got but one
tube - the picture tube. That's a *lot* fewer tubes than a B&W set from
the 50s. It don't seem to have any transistors, neither, 'cept
for that hozirontal output thing. Just a bunch of multi-legged plastic
>On paper, it'd be possible to build an entire FM set with just 2
>compactrons and 1 or 2 minis. One mini would be RF amp, then the next
>would be osc/mixer, then a single compactron as a 2 stage IF, then
>another as a detector and output. Use a sillicon diode voltage doubler
>supply, and bingo. Granted it'd be a crappy set, but hey.
The following fellow "hot rods" classic All-American 5 AM radios, and
among other things has squeezed an AA5 into three Compactrons without
changing the circuit topology:
He also does interesting things with pencil tubes, space-charge tubes
with +12V on the plate, etc.
Get enough selectivity with a VHF-flavored AM receiver, perhaps by
making the IF regenerative, and you can slope-detect the FM. That would
make a very simple, minimal FM radio. Very crappy, but simple. I used
to spend many hours playing Spacewar on the MIT Electronic Research
Society's PDP-7 (our beloved Bertha) in Building 20, accompanied by
music from a station toward the high end of the FM band, received via
slope detection on an AM VHF receiver intended as ground support for
radiosondes. One had to fiddle with the receiver some to get it to
work, and keeping a station tuned in was a distraction, but given that
the 2 inch speaker on the thing was only intended to monitor the beeps
from radiosondes achieving high-fidelity reception with this rig was not
>> I've always been interested in triple triode tubes such as the
>> 6U10; most of a tube op-amp in one bottle.
>Triple triodes lend themselves to chroma amp circuits, or stuff like
>that. Op amps can easily happen as 2 12AX7s or such.
Typically the diff amp first stage of a tube op-amp uses a passive
resistor as the cathode current sink of a dual-triode differential pair.
The second dual-triode provides either a low-impedance differential
cathode-follower output or another gain stage followed by a single-ended
cathode-follower output (as in the once-widely-used Philbrick K2-W),
depending on what's wired where. Putting another triode in the first
stage as an active cathode current sink would give a much greater CMRR,
and reduce the negative supply voltage required for that sink. A
pentode would do even better. The trick is managing the
heater-to-cathode voltage to stay within limits, particularly if it's
all in one bottle.
Anybody interested in the use of tubes in analog computation would do
well to grab hold of a copy of "Electronic Analog Computers" by Korn and
Korn (McGraw-Hill, 1952, 2nd Edition 1956). The hoops one has to jump
through to stabilize precision DC circuits using inherently-drifty
devices such as tubes make for interesting, and sometimes fearful,
>ever made a quad triode, though there were quad diodes of various types
>(6JU8 was seen in lots of color TVs), and dual triode dual diode
>Towards the end, they got really funky with some tubes, too - dual
>plates, dual control grids, etc.
I've heard tell that some of the dual control pentodes, such as the
6GX6, were intended for use in a quadrature-grid FM detector circuit
designed to circumvent RCA's patent on the ratio discriminator. I
imagine they'd make fine symmetrical NAND gates (to bring this thread
around to something approximating computers) if somebody had a hankering
to play with them.