max@eros.uucp (Max Hauser) (09/17/87)
This follows up my posting of the questions alone, six weeks ago.
Answers to questions to identify old-fart electronics hackers. (General,
pre-microprocessor electronics.) I welcome comments, especially by mail.
I apologize in advance if there are any gaffes, but please be careful
about attacking these answers (some of which are deliberately tricky).
Send mail and I'll post a retraction if I goofed! Also, if there is heavy
demand, I can of course post references for most of this.
For clarity, context identifies registered trademarks.
1. Assuming that you are acquainted with the "cascode" configuration,
do you know where the term came from?
A cascode amplifier entails one device "stacked" on another
(as in drain-to-source, or collector-to-emitter), to decouple the
output circuitry from the input (as to reduce AC feedback, increase
output impedance, or increase output breakdown voltage). The term
comes from vaccuum tubes: a cascode was one amplifying stage
CAScaded into the cathODE of another.
2. What is a reflex amplifier?
A circuit that simultaneously amplifies two or more independent
signals, in separate frequency bands. Commonly used in RF receivers
and TVs to amplify both a radio and an audio signal, when active
components were expensive (tubes and early transistors).
3. What is the basic principle of a superhet receiver? of a regenerative
receiver? The difference between a mixer and a converter (in RF)?
Superheterodyne: intermodulate an incoming RF signal with a
tunable local oscillator to yield a frequency-translated signal at an
"intermediate" frequency (IF); this can then be filtered for selectivity
and sensitivity by a *fixed* tuned-amplifier chain (see #15).
Regenerative: positive feedback in a simple amplifier stage
increases gain.
A mixer combines two signals to produce beat frequencies; it
approximates an ideal of multiplying the voltage waveforms. A
converter is a self-oscillating mixer that generates its own
reference frequency. Five-grid tubes (heptodes) were sometimes used;
more recently, multiple-gate MOSFETs.
4. Who developed the op amp, and when?
The late George A. Philbrick, in the late 1940s. People even
proposed calling them "Philbricks."
5. What is the "purple plague"?
One of a class of parasitic alloys of gold and aluminum that can
form on IC wire bonds and impair reliability. It was a major
chip reliability problem in bipolar-technology parts circa 1970.
6. What is a class-C amplifier and where is it typically used?
An amplifying device operating in class C (as opposed to A, AB, B,
D, E, F, S, X, etc.) conducts current for less than 50% of a sinusoidal
cycle. Normally used in tuned power amplifiers, where a resonant
load circuit maintains output-waveform integrity despite the
distorted current waveform in the active device. Many transmitters,
such as citizens'-band, employ class-C power amplifiers. (Because
this mode is inherently nonlinear, it is not suitable as a *linear*
amplifier for amplifying amplitude-modulated signals. Nevertheless
it is common in AM transmitters as a carrier power amplifier
to which modulation is applied by varying the power supply.)
The late 1970's saw a vogue in theoretical circles for naming
new classes of amplifiers (roughly the electronics equivalent of
discovering a new star). Many of the later classes are slight
variations on one another; their principal constructive impact
may have been more in building personal publication lists than
in advancing technology.
7. Can you describe a tunnel diode? a unijunction transistor? an SCS?
(What is the basic principle of each and what are they used for.)
The tunnel (or Esaki) diode (the 1N3712 series was famous --
introduced, I believe, by GE): a low-voltage negative-resistance
device based on a heavily-doped PN junction with a very thin
depletion region. Although it never fulfilled initial high hopes as an
important switching element, there is still nothing handier if you need
to make a low-power VHF transmitter the size of a pencil eraser. The
tunnel diode, like the transistor, is one of those devices that
(a) occasioned a Nobel prize and (b) had actually existed much earlier
than generally thought, but was believed useless by earlier discoverers.
Unijunction transistors are three-terminal (base-1, base-2, emitter)
negative-resistance devices where the emitter conducts current in a range
of voltages set by the bias on the two bases. Commonly used as a threshold
sensor or relaxation oscillator. The Semiconductor Controlled Switch is a
four-layer (PNPN) device like the Semiconductor Controlled Rectifier (SCR)
but with its gate electrode at the N mid-layer instead of the P, and
optimized for small-signal rather than power applications. The SCS is now
often called an SCR.
8. Traditional op-amp ICs were designed whenever possible to run
on +- 15 volts. Why that voltage?
To permit a clean output swing of +- 10V, a convenient scale for
their original service as analog-computer computation elements.
Vaccuum-tube op amps set the trend earlier with +- 100V output swings.
9. What do the following acronyms stand for: PDP, VAX, ASCII, EBCDIC,
PRV, BFO, RTTY, CW, VSB, VOR, Conelrad?
Programmed Data Processor (of course!); Virtual Address eXtension;
American Standard Code for Information Interchange; Extended Binary-
Coded Decimal Information Code; Peak Return Voltage; Beat-Frequency
Oscillator; RadioTeleTYpe; Continuous-Wave; Vestigial-SideBand; VHF
Omnidirectional Range; (Plan for) CONtrol of ELectromagnetic RADiation.
10. What was revolutionary about the 741 op amp?
It was the first major IC op amp fully self-contained; it required
no external components for AC stability. As a result of this
convenience, the industry has been saddled with its shortcomings,
and those of its successors, ever since.
11. Can you specifically describe the commercial semiconductor devices
known by the following terms: 1N34, CK722, 2N107, 2N998, 2N1304, 2N2222,
2N3055, U222, uL900, uL958, uA703, uA709, SN7300 series, SUHL, CCSL,
HTL, ECCSL, Utilogic, COSMOS, Intel 1101, Intel 1702.
High points: 1N34, the ubiquitous glass-packaged germanium signal
diode; 2N998, TO-18-packaged Darlington transistor pair; U222, early
power RF junction FET, handy as hell but expensive; uL900, Fairchild
RTL buffer chip (very SSI logic); uL958, Fairchild early decade-counter
chip (one of the original MSI circuits); 703, Fairchild RF amplifier chip;
SN7300 series, TI DTL logic family, since eclipsed by the 7400 TTL family;
SUHL, Sylvania Universal High-Level Logic, and CCSL, Fairchild's
Compatible Current-Sinking Logic DTL-TTL family, both contenders in late
1960s for the industry-standard status that finally fell to TI's 5400/7400
series; ECCSL, Emitter-Coupled Current-Steered Logic, an ECL family
from (?) RCA; Utilogic, Signetics's offbeat SSI-MSI logic family with
a mix of circuit types and native gates of all four types; COSMOS, RCA's
trademark for COmplementary-Symmetry MOS (CMOS) logic; 1101, original
MOS static RAM from Intel (256 bits, high-threshold PMOS technology).
12. Can you identify the following trademarks: Nixie, Pixie, Numitron?
The Nixie, as discussed on this newsgroup, is a neon glow tube with
glowing cathodes shaped like characters. The Pixie, its predecessor
and a much better article of old-fartness, had a disk anode visible
from the front with holes cut out in the shape of characters; glowing
cathodes behind them illuminated the desired character. Both are
Burroughs trademarks. Numitron was RCA's trademark for its hot-wire
seven-segment displays, popular circa 1970 but largely supplanted by
LEDs and other more-recently-perfected technologies.
13. What is a thyratron? a magic-eye tube? a compactron?
Thyratrons are hot-cathode gas-filled tubes that latch in response
to a gate signal, predecessors of PNPN devices. Used mainly in switching
and power-control applications. Perhaps they will be useful again in
switching (class-D) audio amplifiers for die-hard vaccuum-tube freaks.
Some vaccuum-tube and hybrid digital computers, such as the Sperry (?)
SS-90, used arrays of thyratrons as registers. A magic-eye tube, for the
benefit of anyone missing the earlier discussion on the net, is an
end-viewed fluorescent analog display used often as a level indicator.
Compactrons are a class of late miniature tubes, common through the
early 1970s, with lots of independent devices within each bottle, and
large pin counts.
14. What magnitudes of voltage(s) are required for operating the
following devices: neon bulb; Xenon flashtube; Geiger-Mueller tube?
Neon bulb, around 100V; Xenon flashtude, 200-300V for small ones, and
a trigger pulse of a few kV; G-M tube, 400-600V.
15. What IF frequencies are commonly used in the US for AM broadcast,
FM broadcast, TV sound, TV picture?
455 kHz; 10.7 MHz; 4.5 MHz; 45 MHz. See #3 for definition of IF.
16. What is a Hartly oscillator? a Colpitts oscillator? a Pierce
oscillator? a Wien-bridge oscillator? a blocking oscillator?
Hartley, LC sinusoidal oscillator with tapped coil; Colpitts, LC
sinusoidal oscillator with capacitive voltage divider; Pierce,
a crystal oscillator; and the Wien-bridge type, an RC sinusoidal
oscillator. The blocking oscillator is an RL oscillator typically
used for timing at sub-megahertz rates and usually realized with a
transformer and a single active device (transistor or tube).
17. What voltage corresponds to zero dBm in a 600-ohm circuit?
(Quick, without calculating).
0.775 V (the voltage corresponding to 1 mW in 600 ohms). This is
a standard reference value for 600-ohm professional audio circuits.
18. What is an Eccles-Jordan circuit? a polyflop? a switch-tail ring
counter? a Johnson counter? an AC-coupled flip-flop?
The Eccles-Jordan multivibrator is the original flip-flop, built
with vaccuum tubes in the (?) 1930s. It is exactly the kind of device
to which the IEEE _Spectrum_ devotes those nostalgic, elegiac
articles for which it is lately so well known. A polyflop is a
generalization of the flip-flop, with more than two outputs,
only one of which is "on" at any time. It can be constructed by
cross-connecting NAND or NOR gates with as many inputs as gates.
A switch-tail ring counter or Johnson counter, two names for the
same thing, is counter formed of a shift register with its output
inverted and fed to its input. In practice, it is usually desirable
to gate all of the outputs back to the input so that the circuit
can recover if it enters an undesired state. AC-coupled flip-flops
are edge-triggered flip-flops whose clock input is capacitively
coupled. Common in two-transistor discrete FF's, of the Eccles-Jordan
type, this fell out of fashion and then re-emerged in some
bipolar-process IC FF's.
19. In the context of filters, what is a biquad? [Note: most engineers
get this wrong.]
A biquad is any block that realizes a biquadratic transfer function
(i.e., numerator and denominator are quadratic polynomials in s
[for continuous time] or in z to the -1 [for discrete time]). Many
analog people think it means a specific op-amp-based circuit
(the state-variable biquad) that is actually only one implementation
(and somewhat obsolete, as a finicky aside; it was displaced in the
mid-1970s by the NIC biquad, which needs fewer parts to achieve the
same properties).
20. What does it mean when a resistor is marked with bands
red-yellow-green-silver? a capacitor with letters "104K"?
2400000 ohms, 10% tolerance; 100000pF (0.1 uF), 10% tolerance
21. If an aluminum electrolytic capacitor is rated for a "working
voltage" of 10 volts, in what range of voltages is it designed to
operate?
Not more, but more importantly not much less, than 10 volts. Five
volts, probably OK; two volts, it may not form the dielectric layer
reliably.
22. Roughly, what are the main advantages and disadvantages of the
following capacitor dielectrics: ceramic, polystyrene, polycarbonate?
I posted separately on this earlier, but briefly, ceramic caps are
cheap, compact, wideband and not very precise; polystyrene are cheap,
stable, have very high leakage resistance, and have the larger inductance
characteristic of wound-film rather than sandwich construction; polycarb
are similar to polystyrene but extremely value-stable and expensive.
23. What is a "2 1/2 D" core memory plane?
I'll leave details the 1960s computer hackers. It's a magnetic-core
array of a particular wiring configuration. If the term is familiar at
all, then you remember computers from the Good Old Days (1620, 1108,...)
24. What is a couplate? a micromodule? the difference between
thin-film and thick-film hybrid circuits?
Thin-film hybrid circuits contain components (resistors, capacitors,
wiring) deposited with a screening process onto a ceramic substrate.
In thick-film hybrids, the components are discrete but leadless and
they are attached to a ceramic substrate after separate fabrication.
In both cases, transistor and IC dice also get attached after separate
fabrication. These circuits are called "hybrid" because they mix
discrete-component and integrated-circuit technologies.
Couplates and micromodules are two examples of thin-film hybrids:
couplates (Sprague TM?) were RC interstage coupling networks in
capacitor-like multilead packages; micromodules (IBM term) were digital
logic circuits, to be installed on small PC cards, and heavily used in
the 360 (?) series of mainframe computers.
25. Do you remember the commotion caused by Ovonic devices in 1969?
Ovonic devices are amorphous (non-crystalline) semiconductor elements
introduced by S. Ovshinsky of Energy Conversion Devices. They arrived
in style at a big press conference and were touted as, roughly, the next
big breakthrough in semiconductors. That the enthusiasm died down soon
after was less the fault of the devices, I think, than of the
sensationalized reporting that followed the initial announcement.
26. (Practical commercial engineering experience) Can you explain
the following business acronyms: FOB, CIF, NRE, ASP.
Free On Board (shipping reference point);
Cost, Insurance and Freight (NOT Caltech Intermediate Form);
Non-Recoverable Engineering (fees or costs);
Average Selling Price (of a product line). At one well-known analog IC
firm I worked for, a big push was to keep the ASP over one dollar.
As a final comment, this quiz arose in an effort to identify electronics
hackers who had pre-microprocessor experience. I've shown it to a number
of EECS undergrads here at UC Berkeley. It is unusual for them to be
able to answer even ONE of these questions. Not only is contemporary
engineering training almost unrelated, but also the college admission
process (at this school at least) strongly favors people with academic
rather than practical backgrounds. -- MH