Rob's web

Home - Techniek - Electronica - Radiotechniek - Radio amateur bladen - QST - Flat lines and loading

There seems to be a common notion that the way a line couples to the transmitter gives an indication of whether or not the line is fiat. To be specific, the case we have in mind is the one where the line is connected to a swinging link coupled to the final tank coil, no auxiliary feeder-tuning apparatus being used. The assumption is that if the transmitter can be loaded to normal input with fairly loose coupling, the line is matched to the antenna and everything is lovely, but if the amplifier won't load then those malicious standing waves are getting in their underhand work.

The facts of the matter may be - and usually are - just the opposite. A line that is really flat looks like a pure resistance to the transmitter, a resistance having a value that - depending on the type of line - may be between 50 and 600 ohms. It is not so easy to put power into a resistance in that range of values with just a loosely-coupled coil, particularly if that coil has only two or three turns and the tank coil itself has a great many. Loose coupling is possible only between two circuits having relatively high Q. In the swinging-link case, the Q of the secondary or load circuit is practically negligible, and the power is transferred by simple transformer action. To get anywhere on that basis, the coupling between the two coils has to be tight, and in many cases the number of turns on the link is not sufficient to couple into the line even though the two coils are as close together as possible.

The doubting Thomases can prove this to themselves by a simple test with a dummy antenna. The Ohmite dummies provide a means for getting reasonably close to a purely resistive load of known value, and the two sizes - 70 and 600 ohms - will simulate two popular line impedances. Two of the 600-ohm units can be connected in parallel to provide a 300-ohm load. If the transmitter power output is too great for the capacity of the dummy antenna, the final-amplifier plate voltage should be reduced to a value that will keep the output under the 100-watt rating of the dummy when the plate current is reduced in the same proportion. For example, suppose the amplifier normally operates at 1500 volts and 300 ma. and the power output is 300 watts. Dropping the plate voltage to 750 and the plate current to 150 (thereby maintaining the same ratio of plate voltage to plate current) will keep the output below 100 watts for testing - and, if the amplifier is actually operating Class C, the same load that draws 150 ma. at 750 volts will draw 300 ma. at 1500 volts with no change in coupling.

Our experience has been that with the ordinary commercially-built tank coils and links it may be possible to load the amplifier properly with a 70-ohm dummy at 28 Mc. and perhaps also at 14 Mc., but at the tightest possible coupling it is not possible to make the amplifier draw enough plate current at 7 and 3.5 Mc. It is usually impossible to get sufficient coupling with a 600-ohm dummy at any frequency. Conditions are most favorable when the tube or tubes used in the amplifier operate at high plate voltage and relatively low plate current. This is because manufactured tank coils generally have inductance values that will give a good tank Q only when the plate-voltage plate-current ratio is high; the coils are too large for tubes that operate at relatively low voltage and high current. Insufficient Q in the tank circuit makes it difficult to couple to a load, particularly when the load is not a resonant circuit.

To get sufficient coupling at the lower frequencies in a test of this sort, it is almost invariably necessary to forget about the swinging link and do one of two things: either wind a new link right over the tank coil, adjusting the number of turns until the proper loading is obtained, or tap the dummy directly on the tank, adjusting the spread between taps until the loading is right. It is not uncommon to find that a great many more coupling turns are needed than the link provides; in fact, the 600-ohm load often requires just about half as many turns as there are in the tank coil itself. The experience of trying to make a dummy antenna "load up" the transmitter should be quite illuminating to those who take it for granted that a swinging link should couple into any "flat" line.

If a dummy antenna won't load the transmitter, how can a line of the same resistance be expected to do it? The answer is, of course, that it can't. If the line does load the amplifier and the dummy doesn't, it is certain evidence that the line is not flat. Before using loading as a criterion of line performance, find out first what sort of coupling is required to get full loading with a dummy antenna of the same resistance as the line impedance. Then if the line loads the transmitter in the same way as the dummy, there is at least a chance that the line is flat. It is by no means a certainty, because it is readily possible that the line length and termination are such as to present a good load at the transmitter with a far-from-negligible standing-wave ratio. Lacking a good means for measuring standing waves, perhaps the best stunt for checking is the old one: Add about an eighth wavelength of line and see if the loading changes; if it does not, add another eighth wavelength and check again. If the loading is still the same, your worries about standing waves are over. But if either test causes the loading to change or makes it necessary to retune the final amplifier for resonance, the line is not flat even if the amplifier can be fully loaded with the swinging link in the next county.

It is an unfortunate fact that one swinging link cannot serve for coupling to a nonresonant line on all bands from 3.5 to 28 Mc. If it has enough turns to do the job at 3.5 Mc. it has far too much inductance to be useful at 28 Mc., and if the turns are kept down so the link reactance will be small at 28 Mc. it is impossible to get sufficient coupling at the low-frequency end of the scale. The solution for all-band work - and for any single band where coupling via the link alone is insufficient - is to use an antenna coupler that provides a means for matching to the line.