Q of a Coil With Some Unused Turns
by Xtal Staff
In designing a coil-based antenna tuner and/or an LC-tank for a broadcast band crystal set, coil inductance is always a concern, given the breadth of the band. Dialing the bottom of the band, ~ 500 kHz, calls for a large inductance. Tuning to the top of the band, ~ 1,600 kHz, works best – generally- with a smaller inductance. Over the years, a number of multi-coil combinations have been used to improve reception across the whole band, as compared to sticking with a single coil, traditionally 250 uH. All schemes must use some sort of coil-segment switching. One method is to simply switch in one coil for the bottom half of the band and another for the top half. A second method used is to combine two coils in series or parallel, taking into account “hot” and “ground” ends of each. A third method – outlined here and intended for intermediate grade sets – is to use just one coil but leave a portion of it unused when tuning the top-portion of the band. We can justify this scheme, compared with method one, if the selectivity in the top-portion of the band does not suffer much.
For example, consider a single coil made up of roughly 60 turns of #26 hookup wire, wound on a 3.5-inch ABS form, with a tap at 40 turns. Let’s call the 40 turns L1 and the additional 20 turns L2, as shown in Figure 1B. There are two ways we can leave L2 our of the circuit schematic wise: leave the other end of L2 open, as in circuit B, or short both ends of L2, as in circuit C. Either way, the combination is a tank circuit made up of L1 and C1, with L2 “left out” schematically but not fully physically. This is our coil for the top portion of the band. We know, of course, that L2 is still in the “physical space.” of L1, contributing capacitance to the combination.
The questions that arise are: How do the Qs of circuits B and C compare over frequency, and, how do they compare with the Q of L1 itself, without L2 present at all. We can theorize about these questions, but the quickest way to an answer is to assemble the various coil combinations and measure their Q.
Using a B&K 4017B RF signal generator, with digital readout, a Tekronix 2445 scope, with 10pf/10meg probe, and a 365-variable cap, the Q of each circuit arrangement was measured, using the half-power bandwidth method. Results are shown in Figure 2. The Q of circuit A, with L1 alone and circuit B, with L2 attached just to L1, are roughly the same. A local station provided some interference at 1320 kHz. The slight dip in Q at 1,000 kHz for circuit B is unexplained, “likely operator head-space.” The Q of circuit C, with L1 and L2, with L2 shorted, is clearly lower all across the band. Hence, it appears that circuit B is preferred if one coil with one tap is to be used to configure bottom and top of band coils in a simple set.