Good old Doug DeMaw, W1FB (rest his recently departed soul!) came up with the "SPC" antenna tuner a while back, in response (apparently) to the "Ultimate Transmatch" that had been circulating for a while before that. The version I'm reading from my 1988 handbook claims better matching range with the same value components.

All I know is it has worked well for me in two versions. Both are operated as "pseudo-balanced" -- that is, a balun placed BEFORE the tuner circuit. It seems to work better with the balun than without, though the gurus like W7EL say lately (in some postings on that this is not a true balanced situation. More arguments below on this.

Basic schematic from my ancient 1988 ARRL Handbook (except I add the balun at the input):

  |      C1  B  |
Balun        B  = C2a      ant
  |       L1 B  |
The nominal values from the article are: C1, 200 pF; C2, 200 pF per section; L1, 28 microhenries, roller inductor.

Big SPCMy QRO version (left) has moderate-power components I've had on hand for years, with a "choke balun" on the input of about 25 feet of RG-58 coax wound on a 5 inch PVC pipe. C1 is about 500 pF, C2 is 200 pF per section, and the roller inductor is a GE component from a hamfest. After some years of just having the components loose and clip-leaded together, I finally mounted them on a piece of wood in November 1997, connected with straps of brass stock from the local hobby shop.

Little SPCThe QRP version (right) uses smaller variable caps with narrower plate spacing -- about 350 pF for C1, 150 pF per section at C2. The inductor is a T-106-2 toroid with lots of turns of #18 wire, with wire taps soldered on; a clip lead makes the necessary inductance adjustment. The input balun is wound on a BN-43-3312 Amidon two-hole ferrite core, 3 turns each for primary and secondary. This handles the 50 watts of my Ten-Tec Argosy pretty well -- no arcing, and the balun seems to stay cool. The slide switch at left permits bypassing the balun. Between the capacitors is a kludged RF ammeter - a small ferrite toroid with a number of wire turns, a diode/capacitor to rectify, and a meter to read the current.

The big one matches just fine down to 160 meters with several turns to spare. The little one is just fine too, and in a pinch serves as a suitable preselector for my DX440 receiver -- as you may know, such small shortwave portables need some preselection when used at night with an external antenna or they are just useless due to the intermodulation of the wide-open front end.

The original article cautions you to use tuner settings that use the maximum possible capacitance for the greatest bandwidth and lowest loss. And sure enough, there is quite a range of inductance that you can use as a starting point for a given frequency, the two C's being subsequently twiddled for a match. The lower the C (at higher L settings), the sharper the tuning feels, and sure enough, the bandwidth for a given SWR range is narrower. Going from one end of a band to another, you usually just have to retune L or C2 to fine-tune your match.

On reflection, you have an interesting practical laboratory exercise in "bandwidth versus loaded Q." What's happening is that when you use the tuning options that make for higher C, you are actually loading the tuned circuit more heavily, thus decreasing the loaded Q. Just read those sections on bandpass filters in ARRL's "Solid State Design" or other references, and it will make sense.

I find this and the Z-Match tuner circuits to be useful because of their additional selectivity. Being a few miles from a 50 kW AM broadcaster, and a few other lower-power AM stations that seem to ride in on my power lines, there are some circumstances where this helps clean up problems. Sure, the simple L network is simplest, and has the broadest SWR bandwidth. But the SPC seems to go down to 160 more easily than an L network with the same C1 and L as this tuner (the big one at least) uses.

<>In closing, just one more true balanced tuner to mention... Rick Measures, AG6K, published "A Balanced Balanced Antenna Tuner" in Feb 1990 QST - an updated online version from the author is at  I played with a version of this, and it worked well, but it doesn't serve to add any selectivity to your system (the only reason I have not used it lately). It is a truly symmetrical network, as opposed to the above network which simply floats with respect to ground.

In Oct. 2004 QST, Jack Belrose, VE2CV, again proposes simply putting that balun ahead of a floating (but unbalanced) network, in this case an L network. His view is that a true balanced situation results, and one set of measurements backs this up.

Back to main page