|
November 1947 Radio-Craft
 [Table of Contents]
Wax nostalgic about and learn from the history of early electronics.
See articles from Radio-Craft,
published 1929 - 1953. All copyrights are hereby acknowledged.
|
It has been three or four decades
since I have seen anything about a
Lecher Line, the last time
in memory being in a college lab. It might have been a physics lab, but most probably
an EE lab. We used one to measure wavelengths of signals from an RF generator. The
apparatus looked sort of like the one in the
Wikipedia link, only just
a little more modern (but not much more, being typical school equipment). This new
patents report from a 1947 issue of Radio-Craft magazine has a waveguide version
of a Lecher Line that supposedly was able to do more precise measurements of very
short wavelengths by providing for detecting the internal wave over multiple wavelengths
instead of just a single half wavelength. It was developed at Bell Telephone Laboratories,
so undoubtedly it was a quality instrument. The other patent covered was an AM-FM
detector circuit which boasts a novel method of switching between AM and FM detection
by a simple change in tube plate bias. Links to the patents are provided below.
New Radio-Electronic Patents
By I. Queen
Precise Lecher Measurements
Patent
 Ultra-High
Frequency Wave Meter Glenn R. Frantz, Pt. Washington, N. Y.
Allen F. Pomeroy, Bernardsville, N. J. (assigned to Bell Telephone Laboratories,
Inc.) Patent No.
2,419,208
Low wavelengths are conveniently measured by Lecher Wires. The r.f. current is
introduced along a transmission line which is shorted at some point. A detector
is connected across the line and its distance from the short is varied until it
shows a voltage node. The detector is moved further along the line until another
node is indicated. The distance between consecutive nodes equals 1/2 wavelength.
For very short wavelengths, a wave guide is used instead of a transmission line.
The detector is coupled to a probe inserted into the guide.
The relative error of a measurement of length increases at short lengths. This
invention reduces the error by measuring over a distance of several wavelengths.
As shown, 2 reflectors are used to short-circuit the guide. The distance between
them is equal to any number of half-wavelengths at the average or median wavelength
which it is proposed to measure. Each reflector is equipped with a handle so that
it may be inserted or withdrawn from the guide. The probe is movable over a limited
distance, the position being measured by calibrated scales.
To make a measurement of wavelength, v.h.f. energy is introduced into the guide,
and the reflector A is inserted. If the wavelength equals the average or medium
value for which the equipment is calibrated, the probe will pick up no voltage at
C, exactly 4 half-wavelengths from A. At this point the scales indicate the average
wavelength of the equipment. If the wavelength is higher, the probe must be moved
back, say to D, to obtain a null. The length CD is 4 limes the change of wavelength.
The new wavelength is read on the upper scale.
For still higher precision, A is withdrawn and B inserted. Now the probe must
be moved back to E for a minimum pickup. CE equals 11 times the change in wavelength
because this reflector is 11 half-wavelengths from C. The lower scale is now observed
for actual wavelength. It is clear that this scale will have more widely spaced
and readable calibrations due to the fact that the actual wavelength change has
been multiplied by 11 instead of 4.
Still greater precision is obtainable by placing a reflector still further from
the probe.
FM-AM Detector Patent
 Carrier Wave Detector Circuit
Frederick C. Everitt, Brecksville, Ohio (assigned to Radio Corp. of America)
Patent No.
2,422,087
Both FM and AM have advantages of their own and several manufacturers are now
selling receivers which can be switched to pick up either type of broadcast. Each
requires a different band width and a different intermediate frequency. so there
must be two separate i.f. channels. This patent discloses a single detector stage
which can be used on both, however.
Output from the two i.f. channels are combined and connected across two i.f.
transformers in series. One is tuned to the 4.3 mc FM channel and the other
to the 455 kc AM channel. Each transformer has negligible impedance at the
frequency of the other, so it is not necessary to switch or short one out while
the other is effective.
When the 2 switches are in the FM position (as shown) the plate voltage is dropped
to about 25 by the plate resistor, and at the same time an R-C network is placed
in the grid circuit. Therefore the triode acts as a limiter. When the switches are
thrown to AM, the grid network is shorted out and the plate voltage is returned
to normal (250 volts). The tube is then a class-A amplifier.
The detector circuit is rather unconventional. The cathode coil has an inductance
of about 200 µh and a natural frequency of 4.2 mc. Its reactance varies with
frequency when FM broadcasts are being picked up. The deviations from the center
frequency of 4.3 mc. are thus translated into amplitude changes of voltage
on the cathode. Since the two diode plates are normally at ground potential, the
potential (with respect to the cathode) changes in the same way. Currents therefore
flow through the diode resistors. One diode is used as a detector, the other as
a.v.c. supply.
An important advantage of this system is that there is no loading of the i.f.
transformer secondaries when AM broadcasts are received. The grid circuit does not
carry current. This gives better selectivity and sensitivity. On the other hand,
the triode produces no gain since its load is in the cathode circuit.
November 13, 2020 Update:
The following note was received from RF Cafe visitor B.B.
"Hi Kirt, Re your article on the Lecher line - I used one in a college RF lab
also. All the lab students seemed to have trouble getting anything close to the
predicted results when measuring along the line when it was loaded with anything
other than the nominal line impedance (600 ohm?), or something that put a reflection
on the line, stubs, whatever. Looking at the setup, the sources for the lines were
donated 50 ohm signal generators. These were attached directly to the line with
no matching network. Not even a balun.
Therein was the source of the problem - the initial reflection caused by non-nominal
loads was being re-reflected by the mismatched source impedance and summing with
the initial reflection at the measurement point (and being reflected again...).
I talked to the prof about it, even gave him a paper I wrote on it with calculations.
(He was all about the math). As expected, he never did anything about it. Wouldn't
it have been nice for the students if their measured results had come close to the
predicted values?"
Posted November 12, 2020
|
