肛门处不舒服:G3YNH info: Variable inductors used in HF ant...

TX to Ae
Ch.3 Contents
-
-

Variable Inductors
D. W. Knight
Variometers
A variometer is a pair of coils arranged in such a way that the axis of one can be rotated relative to the axis of the other. The coils are usually wired in series, and rotating one of the coils alters the mutual inductance. When the magnetic fields produced by the coils are in full or partial opposition, inductance cancellation occurs (mutual inductance is negative). When the fields add, the inductance increases (mutual inductance is positive). The mutual inductance is zero when the axes of the two coils are at right angles.
The principal disadvantage of using a variometer as a variable inductance is that the conductor length remains constant as the inductance is altered. This means that the RF resistance is unnecessarily high for small inductances; i.e., the Q is poor at the low inductance end of the range. The fixed conductor length also gives rise to a lower than necessary self-resonance frequency (SRF) for a small inductance setting, i.e. the useful frequency range is limited. Open-circuit self-resonance occurs when the total conductor length is λ/2 (estimates using c as the propagation velocity are usually accurate to within a few %). A variometer should not be used at frequencies approaching its open-circuit SRF.
Due to the electrical shortcomings discussed above, the use of a variometer as a wide-range variable inductor is not good practice. An acceptable application is supplemental fine inductance adjustment, provided that the variometer inductance is a small part of the total. Variometers are also useful (with the coils wired separately) in applications which require variable coupling between different parts of a network (e.g., for injecting balancing signals in some types of measuring bridge).

ITT-Lorenz variometer (1937). As used in the German WW2 FuG10 aircraft transmitter.
6 to 38μH (4 to 17μH with dust-iron core removed).
For more information, see: RAD COM, Nov 1987. Tech Topics, p837.

LF Variometer. Used for reactive loading of a 500KHz maritime antenna. The two sperical coils are wound with litzendraht and connected in series. The coil formers are moulded plastic, possibly polystyrene. The SRF of this unit lies in the 1.5 - 2.1 MHz region, being at its maximum when the axes of the two coils are at 90° (measured using the scattering method described inTA3.3). Separate resonances were also found for the individual coils, in the 3.1 - 4.4MHz and 4.9 - 6.4MHz regions.
Roller Inductors
The roller inductor (known in the vernacular as a "roller coaster") has the advantage that the conductor length reduces as the inductance is reduced. This keeps the RF resistance of the coil at its practical minimum and allows reasonably high Q to be maintained at the low end of the inductance range.
The big disadvantage of the roller inductor lies in the fact that the unused turns are magnetically coupled to the turns in use. If N is the number of unused turns, and r is the radius of the coil, the length of the unused winding is approximately 2πrN. If the unused winding is left open-circuit, the device becomes a Tesla coil when the unused winding length is λ/2 at the frequency of operation. The resulting corona discharge is likely to etch the wire and blacken or damage the coil former.
One solution to the corona discharge problem is to short the unused winding. This however, gives rise to a large circulating current which reduces the inductance substantially. It also factors the resistive losses of the unused winding into the overall losses, and reduces the Q considerably.


Motorised roller inductor from Marconi AD307 HF aviation transmitter (1958).
This is the PA anode π-network inductor. Unused turns are shorted. The transmitter was designed to be connected to an end-fed wire antenna of prescribed length. Fixed value mica capacitors were selected according to transmission band by means of a rotating turret. Motor operation was controlled according to antenna power factor by means of a phase bridge at the output terminal.
In this example, the original 400Hz three-phase motor has been replaced by a DC motor.

Motorised variable auto transformer from Collins 180L-3A automatic 2-25MHz antenna tuner (1957). Total inductance 20μH.
A roller inductor can be used as an auto transformer by applying an input across the ends of the coil and taking the output from the rolling contact. In this example, the primary connection is made halfway down the coil, to give continuous variation from step-down to step-up

Roller inductor from British Navy antenna tuner (Second World War) . Marked AP.55333. External spring-loaded phosphor-bronze wiping contact on silver-plated sliding-bar. Unused turns are shorted to the wiping contact.

Roller inductor of unknown origin. External spring-loaded phosphor-bronze wiping contact on silver-plated sliding-bar.
Helipot Inductors.
Some variable inductors have a stationary coil and an internal wiper arm, i.e., they use the same mechanical arrangement as a helical potentiometer. These devices do not appear to have a proper name. They are called "roller inductors", even though they are not (except in the sense that a rolling contact may be used for the wiper). To avoid confusion, they will be referred to here as "helipot inductors".
Helipot inductors have the advantage that they do not require brush contacts for the coil end connections. Hence the ESR should be slightly lower (or at least more consistent) than that of a roller inductor of comparable geometry. Another advantage is the absence of a ceramic (or worse still, phenolic-paper composite) coil former.

MFJ "roller inductor" as used in theMFJ989C T-network antenna tuner.
In the HF T-network application, the wiper contact and the end of the unused turns are both grounded. An additional shorting contact called the "Self-Resonance Killer?" (MFJ Registered Trademark) is applied when less than half of the coil is in use. The self-resonance killer (upper centre in picture above) is operated by fibreglass cams which pass between the turns of the coil and engage with the roller contact.
Coil parameters: Mean diam: 53.3mm; length: 139.5mm; wire diam: 2mm; conductor length: 7.2m; N=43 turns; Max inductance (LF, nominal): 30μH.
MFJ variable inductors may be purchased new fromMFJ enterprises.
Binary Switched Inductors
When two or more inductors are placed in series and arranged in such a way that there is minimal mutual inductance between the separate coils, the inductances simply add. By shorting out individual coils by means of relays, a stepwise variable inductance can be created. By choosing the component inductances according to a binary sequence, the relay drivers may be connected directly to a data register and operated by a microcontroller using the simplest possible program.
The disadvantage of the switched inductor chain is that the degree of mutual inductance used in producing a given inductance is sub-optimal in terms of Q and SRF when more than one coil is in circuit, i.e., the same amount of inductance could have been produced with a shorter length of wire. The conductor length drawback is however, nowhere near as severe as with a variometer, because (in impedance matching applications at least) the larger inductors are not used at the high frequency end of the operating range.

Binary switched inductor chain of theSGC 230 automatic antenna matching unit.
Inductors are 32, 16, 8, 4. 2, 1, 0.5, and 0.25μH, giving an inductance range 0 - 63.75μH in 0.25μH steps. The largest coils are wound on stacked Amidon (Micrometals) T-157-2 carbonyl iron toroids. The closed magnetic circuit in these cases prevents coupling to other coils. Most of the smaller inductors are arranged so that adjacent coils have their axes approximately at right angles.

Racal VRA549c, automatic AMU for 500W HF radio transmitters. Pi/L network using binary-switched inductors and capacitors. Switching is by means of vacuum relays.
Coil and Capstan
The coil and shorting drum is the 'no compromises' solution to the problem of designing a variable inductor. For any setting, the inductance always has the shortest possible conductor length and hence the highest possible Q and SRF within the constraints of the chosen geometry.
The downside is complexity. The rotation of the two cylinders must be synchronised using gears, and the conductor must be tensioned (usually by means of a clock-spring inside the shorting drum). An elaborate end-stop system is needed, since the device will fail if either drum is repeatedly unwound as far as the end connection. Notice also that, in order to avoid failure due to metal-fatigue, the drum diameters must be chosen according to the conductor thickness and hardness in such a way that the conductor experiences minimal plastic deformation (i.e. the winding and unwinding process must depend principally on the spring-like nature, rather than the plasticity, of the metal strip). Hence the conductor base (beneath the silver plating) is made from thin brass or hard copper.

Variable Coil & Bar Assembly, from Redifon HF aviation transmitter (Ca. early 1960s).
Machined ceramic former. Flat silver-plated conductor with silver-plated shorting bar. Stacked Nylon cams in the background are the end-stop arrangement.
NATO stock #: 5820-99-633-6198

Motorised Coil and Capstan fromCollins 180L-3A 2-25MHz antenna tuner (1957).
Aluminium shorting drum. Flat silver-plated conductor.
Text and photogtaphs ? D W Knight 2007-2009
David Knight asserts the right to be recognised as the author of this work.

TX to Ae
Ch.3 Contents
-
-