November 1962 Electronics World
Table of Contents
Wax nostalgic about and learn from the history of early electronics. See articles
from
Electronics World, published May 1959
- December 1971. All copyrights hereby acknowledged.
|
A few years ago I was
in a second-hand shop in Erie, Pennsylvania, and happened to spot a Hewlett-Packard
model
HP 5212A Electronic Counter stashed in a cardboard box with a bunch of other
electronic stuff. It was a little dirty, but otherwise appeared to be in pretty
good condition. I took it to the counter and asked the lady what she'd take for
it, and we agreed on $15, provided when I plugged it in the front panel display
would light up and no smoke came from the chassis. It did and it didn't, respectively.
Once at home, I fired it up and ran some functional tests on it, and all seemed
to be working properly. After performing some major clean-up to nearly like-new
condition, I decided it should go to someone who could put it to good use, so it
went up for sale here on RF Cafe. Believe it or not, the best offer received was
$125 (+ shipping). It deserved more respect than that, but the guy was a collector
of vintage test equipment, so at least it went to a loving home. This 1962 "The
Counter as a Test Instrument" article in Electronics World magazine article
shows both the HP 5212A (300 kHz) and the HP 5243L (500 MHz)
electronic counters.
The Counter as a Test Instrument
By Walter H. Bucksbaum
In addition to counting pulses, counters measure frequency and its periodic drift,
phase difference, and the time interval between events.
The simplest way to determine the frequency of a signal would be to use an accurately
calibrated radio receiver and then read the frequency directly from the receiver's
dial. Another method would be to display the signal on an oscilloscope. If the sweep
frequency is known, it is possible to count the number of cycles displayed and calculate
the frequency. Still another way is to beat the known-frequency output of an oscillator
against the unknown signal and determine the latter by the zero-beat technique.
A much more precise and faster way of measuring frequency is with a digital counter.
The unknown frequency is applied to the counter which displays it directly in Arabic
numerals in cycles, kilocycles, or megacycles. In addition, time intervals can be
measured directly and with great accuracy. The counter is extremely helpful where
repeated measurements and accuracies up to five places are required. Such advantages
are not usually possible with other instruments or methods. For these reasons, counters
are widely used in industrial and military applications where precision measurements
are important. Counters are often used when servicing computers and certain radar
circuits.
The counter consists of four sections as shown in Fig. 1.
Fig. 1 - Block diagram of a counter shows its basic sections.
Fig. 2 - Opened gate sends 18 one-μsec. pulses to the counter.
Fig. 3 - Hewlett-Packard 5243L measures frequency to 500 mc.
Fig. 4 - Method of measuring oscillator frequency stability.
Fig. 5 - Technique for measuring two-signal phase difference.
The binary-counter section is a series of flip-flops connected so that two changes
in one stage cause one change in the following stage. The second portion, the clock-pulse
generator, is a well-stabilized, crystal-controlled oscillator whose output is the
time reference for the instrument. The third portion consists of gates (similar
to those used in keyed a.g.c. circuits) in which one signal gates or keys another
one. The fourth part is a read-out device. This may be a series of neon indicators
wired into the binary counter to indicate the number registered there. Or, it might
be a series of "Nixie" indicator tubes driven by a binary-decimal matrix that changes
the binary number of the binary counter into a decimal for display in Arabic numerals.
A typical counter may contain a number of binary counters, auxiliary storage-shift
registers, and circuits that generate multiples and sub-multiples of the clock frequencies.
The functions that can be performed by the counter depend on how the clock, binary
counter, and gates are connected. To measure the unknown time interval between two
pulses (for example 18 μsec. in Fig. 2) 1-μsec. clock pulses are fed into
the binary counter through a gate that is turned on by the first pulse and turned
off by the next one 18 μsec. later. The number in the binary counter, therefore,
is the number of 1-μsec. clock pulses between the two input pulses. Since 1-μsec.
clock pulses were used, the read-out will be in microseconds.
If each cycle of the unknown signal is counted by the binary counter and the
gate is turned on for exactly one second the display would be the input frequency
in cps.
Further applications of the counter are discussed below. In all cases the precision
of the measurements depends on the accuracy of the clock-pulse generator. For this
reason the clock oscillator usually is controlled by a crystal housed in a temperature-stabilized
oven. It is necessary to allow at least half an hour warm-up time before making
any measurements. Aside from the clock, all other circuits are digital, which means
that they are either on or off, require no adjustment, and cannot contribute significantly
to inaccuracy.
How They Are Used
As we mentioned, counters are widely used for frequency and time measurements.
They have one great advantage over oscilloscopes in these applications in that they
provide a direct indication that does not have to be interpreted. It is also possible
to connect a digital printer to the counter to obtain a printed record of periodic
readings. In checking the frequency stability of an oscillator, for example, the
counter and printer would be connected as shown in Fig. 4 to automatically print
a record of frequency drift every ten seconds. In some newer models, the printer
is part of the counter. If a counter is connected to a radiation detector, the radiation
level in counts-per-minute can be monitored.
It should be obvious why counters are used widely in servicing computers and
data processing devices. In these applications they can check the operation of sections
of the computer, measure time intervals between gating and switching functions,
and verify the computer's own counting operations. By using a preset count arrangement,
the counter can work with other devices to provide an output whenever a predetermined
number of events has occurred. For example a counter might be connected to a photocell
to count the number of objects passing a point. After a predetermined count is reached,
the counter can send a pulse to an actuator which will separate or pack the first
batch. The next object would start the count over again.
It is possible to measure the phase difference between two signals or simply
indicate the time period between them as shown in Fig. 5. One signal is connected
to the "start" input of the gating circuit to turn it on. Clock pulses will now
go to the binary counter. The second signal is connected to the "stop" input of
the gate to turn it off. The number displayed represents the time period between
the two signals.
In each case a direct read-out in microseconds or milliseconds is possible. A
further refinement available in practically all counters is count averaging. In
this mode of operation, the measurement cycles, usually 10 or 100, are added and
the total is divided by the number of cycles.
Counters are available with different frequency and time-interval ranges. In
their early stages of development, counters with a 100-kc. basic clock-pulse frequency
were used for frequency multiplying up to 1 mc., which meant that the most accurate
measurement that could be made was within ±1 μsec. Today counters with
clock-pulse frequencies up to 100 mc. permit measurements to within 0.01 μsec.
The frequency and time range of the counter are but two of its important performance
characteristics. Oscillator stability, which assures the accuracy of all measurements,
is usually given in parts-per-million per week and is generally better than 1 p.p.m./week.
Another important characteristic is the input sensitivity and impedance; this determines
what kind of signals can be measured. Typical values are 0.1 volt r.m.s. at an input
impedance of 1 megohm, shunted by about 50 μμf. That is the minimum-level input
signal. However, a gain control for each input is usually available to permit the
measurement of voltages up to about 300 volts. The counter will also have a scale-factor
switch that automatically locates the decimal point in the read-out. A switch permits
selection of the various functions such as frequency or period measurements, phase
difference, or averaging. It's possible to make a single count, requiring manual
reset for the next.
Posted November 9, 2022
|