From owner-qrp-l@Lehigh.EDU Fri Jan 16 01:22:05 1998 Received: from fidoii.CC.lehigh.EDU (fidoii.CC.lehigh.EDU [128.180.1.4]) by oucsace.cs.ohiou.edu (8.8.5/8.8.5) with ESMTP id BAA04575 for ; Fri, 16 Jan 1998 01:22:04 -0500 (EST) Received: from Lehigh.EDU ([127.0.0.1]) by fidoii.cc.Lehigh.EDU with SMTP id <12710-38740>; Fri, 16 Jan 1998 01:21:43 -0500 Received: from nss4.cc.Lehigh.EDU ([128.180.1.13]) by fidoii.cc.Lehigh.EDU with ESMTP id <12665-41302>; Fri, 16 Jan 1998 01:20:24 -0500 Received: from sunburst.usd.edu (sunburst.usd.edu [192.55.228.48]) by nss4.cc.Lehigh.EDU (8.8.8/8.8.5) with SMTP id BAA31878 for ; Fri, 16 Jan 1998 01:20:08 -0500 Received: from localhost (aweiss@localhost) by sunburst.usd.edu (8.6.12/8.6.12) with SMTP id AAA19745; Fri, 16 Jan 1998 00:20:10 -0600 Message-Id: Date: Fri, 16 Jan 1998 00:20:06 -0600 (CST) Reply-To: aweiss@sunflowr.usd.edu Sender: owner-qrp-l@Lehigh.EDU Precedence: bulk From: Adrian Weiss To: "Low Power Amateur Radio Discussion" Subject: Re: Bright Idea Award #1; RG174U; Balun loss; c.w. tone In-Reply-To: <199801151538.KAA26389@mailhost.infi.net> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-To: Bob Kellogg X-Cc: Low Power Amateur Radio Discussion X-Authentication-Warning: sunburst.usd.edu: aweiss owned process doing -bs X-Sender: aweiss@sunburst X-Listprocessor-Version: 8.1 beta -- ListProcessor(tm) by CREN Status: RO Hi Bob & gang: On Thu, 15 Jan 1998, Bob Kellogg wrote: > Subject: Re: Bright Idea Award #1; RG174U; Balun loss; c.w. tone > > Adrian, > > I've been doing quite a bit of tuner testing using the method described by > Frank Witt in his 1995 QST articles. So, naturally, I have some comments > about your balun loss example. > > You said: > > For measurements, a calibrated wattmeter monitored power fed into the > > tuners, and a simple field-strength meter connected to a pick-up wire > > outside monitored the field strength. Power into the tuners was kept > > constant. A reading was taken with tuner A peaked, then a reading for > > tuner B. Then the power to tuner A was backed off until it reached the > > f.s. reading for tuner B. Then the difference in dB was calculated. > > The results were interesting: loss in tuner B attributable to the balun, > > with a 1:1 SWR obtained on all bands, was: > > 80m = 2.dB; 40m = 1dB; 20m = 1.7dB; 15m = 6.6dB > > I was unable to reach a low SWR with tuner A on 40m, so the 1dB would > have > > been slight worse since this represents output from tuner A at a 2:1 SWR. > > IMHO this is a valid test, and certainly does indicate relative efficiency. > It is a good way to get a rough idea of tuner performance. There is one > limitation, and that is that for each band the tuner is being tested for > just one impedance situation -- the one presented by your antenna. If this > is the antenna you intend to use, of course, that's fine. I'm not sure the > test would be a valid general comparison between the two tuners. It was > certainly a valid test with your antenna. > You are perfectly correct Bob. I was only interested in my system, not in measuring tuners in general. And of course, one cannot generalise to all situations in which these two tuners would be options. But I suspect that the balun loss will always be there but with differing magnitude depending, as you note, upon the impedance presented by the system > > Now, the choice on 15m is obvious. It illustrates the kind of loss that > > can occur in the balun assisted tuner. I won't speculate on why it > > occurred at this magnitude on 15m. I didn't check 10m. > > Someone else might make the identical test with their antenna and get > considerably different results. Instead of 15M being the poorest band, it > might be 40M, for example. I'd judge that the reason test results were so > bad for you on 15M is that your antenna presented an impedance that was > well out of the range for your balun on that band. I absolutely agree with you except that I would bet a couple of 2n2222's that the balun-tuner will be the loser in any situation where it is compared to a decent balanced tuner with the right combination of taps. Many years ago when I was using an 8JK system, George Bonadio W2WLR (ham radio, 1970, pp 28-31) published a piece about conjugate matching tuners. His main point: in order to eliminate loss from circulating currents in the tuner, one had to feed the line across a lumped L/C circuit designed to handle the specific complex impedeance presented at the end of the line. So, I did what he advised and it worked quite well, but I compared his tuner to nothing else. So, who knows? I think maybe I'll just whip up a lumped L/C unit here and include it in the tests. Boy -- isn't QRP fun? What QRO guy is going to go thru all this trouble to see if he is losing 52% in the tuner? WE QRP'rs have tghe best of all possible worlds -- except that it includes QRO's QRM'ing us. > > > At any rate, the choice is between ease of bandswitching with the > > balun-assisted, or the relative gain that occurs along with the messing > > around adjusting taps with tuner A. I go for the gain myself. > > Yep. > > I guess what I was saying about tuner testing is obvious. A simple test is > to use fixed power, then hook a wattmeter, the tuner and a dummy load in > series. Measure the power. Then put the wattmeter on the dummy load side > of the tuner and measure the power again. The difference in power reading > indicates tuner/balun loss. This method has the same limitation as above, > but worse. The measurements are being made at 50 ohms impedance. (why use > a tuner at all if the antenna system presents 50 ohm impedance?) OTOH, If > there are losses indicated with a 50 ohm impedance, we can expect much > worse with the typical antenna! True -- but I suppose, in theory, a poorly designed tuner could exhibit loss in the 50-50-Ohm situation, so, a perfect 50-Ohm line termination would still suffer. But who still believes that such a thing exists? Incidentally, I might note in passing that, when a 50-Ohm coax is terminated in a dipole, it is automatically a 1.4:1 SWR situation with a complex impedance. Raising and lowering the antenna or trimming the xoac can get you down to 1:1, but this is illusory. Moxon shows a variant of the usual calaculated graph of dipole impedance vs height but puts in a dotted line under 1/4-wave height to show what actually happens in real antennas. Due to ground-coupling and power absorption in the dirt, the actual measurable impedance of a low dipole stays fairly constant around 45-50-Ohms although it should be dropping with the antenna height. Also, while in passing, I ought to comment about a misconception about inserting a balun between a length of balanced line and a coax run to the shack. I think this arrangement came about from the frequently seen G5RV, where, given the specific length of the legs of the antenna, and given length of balanced line drops from the antenna, is connected to a run of 50-Ohm coax of any length, which then completes the run to the shack. Now, the length of the balanced line, given the impedance of the antenna, is selected to provide an impedance transformation down to the vicinity of 50-Ohms. Use any other length of line with that antenna, and it won't be a 50-Ohm termination. In reality, the balanced line dropping from the G5RV is an impedance transformer doubling as the feedline. It is an integral part of the antenna system, just as an extra element on a yagi is. The coax is actually feeding a system terminated in the balun. However, and this is the main point, it only works when (1)when the antenna feedpoint impedanceis known and (2) the length of balanced line which will transform that to 50-Ohms is calculated (or determined with a Smith Chart) . If these two conditions aren't met, then the (usually random, selected for some reason other than electrical length) run of balanced line terminates in an unknown impedance which the balun then transforms down to another unknown impedance to which is connected coax with a Zc = 50-Ohms. I'll bet 2 dozen 2n2222's that the coax will be running a high SWR -- usually. Now, where does a balun belong in the antenna system if used at all? Pick a point where something unbalanced has to be connected to something balanced -- a dipole is a balanced antenna, coax is unbalanced line. Of course, for a half-wave dipole, a 1:1 balun is required. The insertion point for a balun thus is at the antenna terminals. Why even take such a step? To eliminate radiation from the coax which will severely distort the radiation pattern. Back in the April 1980 QST, Bruce Eggars WA9NEW published the results of his testing of tghe effect of a balun at the feedpoint in the r.f. anechoic chamber (non-reflective, hence only the radiation pattern from the antenna is measured, as if in free space) at U. N.Carolina. All one has to do is glance at the two cases to see that hooking the coax directly to a balanced dipole produces the same kind of radiation pattern that is traced by those whirly-giggy July 4th fire-works. Insertion of the balun produces a not-quite perfect, but nearly so figure eight. Interesting that we never even think of putting a balun where it belongs! Thanks for your comments Bob. Incidentally, I was going to try Witt's measurement technique, but didn't have time to build the geometric resistance boxes. Obviously you did and have played with them. I'd encourage you to write up your results for here or the QQ and QRPp, or all. I for one am dying to find out what you came up with. Anybody else out there interested? 73, Ade W0RSP Oh yes, doing these posting in the wee hours isn't the greatest idea -- I forgot to put in my promised comment on RG174U. Simple: it's a tradeoff. I used RG174U out camping because of its convenience -- did a tiny surface-mount tuner which went along with the Viking-5 (see HISTORY OF QRP) with a toroid inductance and trimmer capacitors, all fine-tuned with the antenna up at home. Worked well. But even though the run of RG174U was only 35-ft, I wonder if now I wouldn't opt out for a 35-ft hank of small twinlead plus a tuner. The longer I do this QRP thing, the mor I value a dB or two. > > 72/73 and CUL, > Bob Kellogg, AE4IC, Greensboro, NC > Prolably, but not nececelery. -- Benny Hill > > >