How often have we all mistaken
Barkhausen oscillations? Yeah, it's embarrassing, but we've all
done it. I can't tell you how many times as a kid I saw the tell-tale effects on
our old black and white TV and said, "Mom, can you remind Dad to do something about
those dang Barkhausen oscillations when he gets home from the newspaper office?"
If you believe that line of bull hockey, I've got some waterfront property in the
Sahara Desert to sell you. The only thing close to "Barkhausen" I might have known
back then was the name of a German beer house on
(for which I own the entire DVD
set). Anyway, this article, written in the days of over-the-air television broadcasts,
presents a solution to the annoying "spook" effect caused by poor oscillator circuit
shielding. It is a propos for posting on October 31st (Halloween).
The Spook - Another Weird Effect to Haunt TV
Fig. 1 - Hand-made r.f. chokes used for minimizing spook
By M. B. Knight
Tube Department, Radio Corporation of America
A simple solution to the problem of electromagnetic radiation from. horizontal
deflection circuits of set.
A number of interesting phenomena, sometimes not anticipated in fundamental studies,
have cropped up in the electronic art. These phenomena often are identified by colorful
names which are intriguing in themselves. The "spook," an interference effect in
television receivers, may be destined to take its place in the language of the trade
with such terms as "ghost," "barks," "motorboat," "birdies," "snow," and "jitter."
Description of Spook
The spook originates as electromagnetic radiation from the horizontal deflection
circuits of television receivers and is picked up by the sensitive r.f. or i.f.
circuits of the receivers. Like any other signal in the r.f. or i.f. circuits, the
spook signal is amplified, detected, and applied to the grid or cathode circuit
of the kinescope. When seen in the picture, it appears as a narrow vertical band
very close to the left-hand edge of the raster and resembles the more familiar interference
from Barkhausen oscillations in the horizontal-output tube. If the television signal
is weak compared to the spook signal, the line is black and has ragged edges as
shown in Fig. 2. If the television signal is of normal strength, the line is
not black but has within its margins crawling diagonal lines which are caused by
heterodyning between the television signal and the spook signal. The appearance
of the spook when the television signal is of normal strength is shown in Fig. 3.
Despite the similarity to Barkhausen oscillation, several distinguishing features
establish the spook as a separate effect: (1) The line is always in the same place
in the raster, very near the left-hand edge. (2) The interference, if picked up
in the r.f. circuits, is always strongest on the lower-frequency channels, whereas
Barkhausen may be more pronounced on either the low or high-frequency channels.
(3) Experiments show that the radiation does not come from the horizontal-output
tube, nor do the usual cures for Barkhausen oscillation, such as magnets, have any
effect on it.
Actually, the spook seldom has a serious degrading effect upon receiver performance.
Because of its location on the raster, it is usually within the blanking period
and it is almost always off the kinescope screen because receivers are normally
adjusted to have a good margin of deflection width. The more common degrading influence
of the spook is that it upsets the receiver synchronizing circuits. Because of the
action of the detector circuit, the spook shows up in the video circuits as a pulse
in the "black" direction, similar to the sync pulses. If it is of sufficient amplitude,
the spook pulse will pass through the sync amplifier along with the regular sync
pulses and may impair the operation of the horizontal a.f.c. circuits.
Occasionally, two receivers are close enough together so that one picks up the
spook interference from the other. The resulting picture disturbance is quite objectionable;
if the receivers are tuned to different stations (the line-scanning frequencies
of which may differ slightly), the spook line may move back and forth across the
Discovery of Spook
Fig. 2 - Appearance of "spook" interference with weak signal.
Fig. 3 - "Spook" interference with signal of normal strength.
To the best of our knowledge, the spook phenomenon passed unnoticed, or at least
uncommented upon, for three or four years of commercial receiver production. There
are several reasons for this delay. First, because of the usual practice of scanning
beyond the kinescope mask, the spook line is rarely seen. Second, because the intensity
of the spook radiation is a function of the deflection power, it has become more
evident as larger kinescopes, having larger deflection angles and accelerating voltages,
have come into popular use. Third, the older deflection circuits were susceptible
to Barkhausen oscillations, and the spook, even if observed, could easily be dismissed
The advent of modern high-efficiency deflection circuits, however, called attention
to the spook as a unique effect. The writer's first encounter with the spook occurred
about two years ago during the development of the RCA- 223T1 horizontal-deflection
transformer. The narrow vertical band appeared to be due to Barkhausen oscillation;
further investigation, however, showed that cause to be unlikely because careful
measurements established that the plate voltage of the horizontal-output tube was
not negative at any time during the scanning cycle. Other engineers observed the
effect at about the same time and established that it was not Barkhausen. No parasitic
oscillation could be found and the mysterious nature of the effect caused it to
be dubbed the "spook." The name seemed apt and has persisted. The effect was distressing
to receiver designers, mainly because of its elusive nature, and efforts were made
by the writer to locate the cause. After some investigation, a reasonable explanation
was found, and methods of minimizing the interference were easily devised.
How Spook Is Generated
In the investigation of the nature and cause of the spook, a separate television
receiver was used to search for the most prominent source of radiation. Although
some radiation could be detected from most parts of the deflection circuit, the
damper tube and its leads produced the strongest radiation.
Scrutiny of the current waveforms in the horizontal deflection circuit showed
that the spook line appears at the same instant that the damper tube begins conduction,
approximately one microsecond after the completion of retrace. The damper tube plate
current rises from zero to its maximum value of 350 to 400 milliamperes very rapidly.
The rise time of the current was not measured precisely, but avail-\able equipment
indicated that it was 0.1 microsecond or less. At any rate, it was apparent that
the electromagnetic fields associated with such a rapid change of current and voltage
must contain many high-frequency harmonics. It could be expected, therefore, that
the high-frequency harmonics could be radiated to the signal circuits of the receiver.
This theory was checked by exploring the radiation spectrum with a communications
receiver. The receiver was tuned from about 300 kilocycles to 18 megacycles and
a signal was detected at every harmonic of 15,750 cycles. A more significant observation
was that no other signal was detected. The intensity of the harmonics diminished
steadily as the receiver was tuned to higher frequencies. In addition, spook interference
was found to be most severe on television Channel 2 and was successively less severe
on higher-frequency channels. If the harmonics were being radiated as a result of
the rapid plate-current change in the damper tube, it would be assumed that high-frequency
harmonics would be of less amplitude than low-frequency harmonics.
Small r.f. chokes were placed in the leads to the damper tube at the socket and
the radiation was reduced considerably. The residual radiation came almost entirely
from the internal structure of the tube itself. Further tests confirmed that the
interference originated with the rapid change in the damper-tube plate current.
Minimizing Spook Interference
The rapid rise of plate current in the damper tube is inherent in the proper
operation of deflection circuits. Practical means for slowing down the increase
in current are not at hand, and it is not expected, therefore, that the spook can
be eliminated entirely. It is possible, however, to reduce considerably the detrimental
One approach to the problem of reducing spook interference is to minimize the
susceptibility of the r.f. and i.f. circuits to the radiation by physically separating
the r.f. and Lf. circuits from the deflection circuits. This separation is chiefly
a chassis design problem; good chassis layout in this respect is normal in commercial
receivers. The technician is more concerned with the installation of the receiver;
he should be sure that the antenna lead-in is dressed away from the deflection circuits.
Not much can be done along this line if an in-cabinet antenna is used.
A second approach to this problem is to minimize the radiation from the deflection
circuits. Because the damper tube and its leads can be thought of as a transmitting
antenna which radiates the spook, a logical approach is to make the transmitting
antenna as small as possible and to provide a shield between this "antenna" and
the receiver r.f. and i.f. circuits. It was mentioned before that insertion of r.f.
chokes in the leads to the damper tube limited the "antenna" to the tube structure
itself. The value of the chokes is not critical, but must be large enough to be
effective in the television band without being so large that ringing is caused in
the deflection circuit. Chokes having inductance values between 1 microhenry and
5 microhenrys are suitable and are available commercially. Chokes for the plate
and cathode circuits can be made by winding approximately 30 turns of AWG #28 enamel
or Formex wire on a one-watt resistor. Ordinarily, it is not important to insert
chokes in the heater circuit. If heater chokes are used, however, wire at least
as large as AWG #20 should be used to carry the heater current. The stiffness of
this size wire makes a coil form unnecessary; a coil of approximately 20 turns about
3/8 inch in diameter is adequate. Fig. 1 illustrates typical hand-made chokes.
After chokes have been placed in the damper tube leads, it is desirable to shield
the tube from the receiver r.f. and i.f. circuits. The high-voltage enclosures used
in most receivers provide adequate shielding. The shield should be inspected to
see that it is grounded at as many points as possible. If there are any large holes
in the enclosure, they may be covered with ordinary copper wire screen to improve
the effectiveness of the shield. Capacitive coupling between the damper tube and
any leads which come out of the high-voltage enclosure should be minimized by careful
lead dress. A close-fitting shield around the tube, however, is neither necessary
nor desirable because of the resultant increase in bulb temperature.
In the commercially popular auto-transformer and direct-drive types of deflection
circuits, experience has indicated that spook interference can be greatly reduced
by the addition of only one r.f. choke. In such circuits, most of the radiation
usually comes from the "B+" lead which is connected to the plate of the damper tube.
(The cathode lead is quite well shielded by the high-voltage enclosure.) The r.f.
choke, therefore, should be placed in the plate lead of the damper tube at the socket.
The addition of a condenser of approximately 100 μμfd. between the chassis
and the "B+" side of the choke may give further improvement.
The techniques suggested for reducing spook interference are not all-inclusive.
Each type of deflection circuit and each mechanical layout requires individual attention.
It is expected, however, that an understanding of the source of the interference
will be of help to the troubleshooter when spook problems appear.
Posted October 31, 2022
(updated from original post