Antenna Matching Extreme

[flintstone mop]

I always take my hat of to the AM guys that continue to keep the transmitters on the air with reasonable audio.
I found this article interesting about how important it is to maintain a good match of the AM Bcast antenna array even at the band edges of the assigned frequency. Not like we do 1.5:1 at 1.885 mhz. That magic low SWR
Here is a link to a problem an engineer had trying to get a Harris MW 50 to operate and modulate properly.

http://mighty1090kaay.blogspot.com/2010/12/hollis-breland-comments-re-kaay-harris.html

Fred

[W8IXY]

I had some similar experience at WWWE (now WTAM) in Cleveland with their Harris MW-50 50kw transmitter.  When I took over as chief engineer in late 1987, the Harris MW-50 was in horrible condition, dirty, modifications, you name it.  I knew the transmitter sounded awful.  We measured the THD at 11% and the IM at 55% (!).  The modulation was almost 10db asymmetrical (40% positive and 100% negative on a sine wave).  Over the next several months we completely rebuilt the input audio PDM circuitry, restored the original circuitry to the PA components, and installed a Harris mod to reduce power supply bounce that increased the IM.  When we finished, at 50kw carrier, at 95% modulation, our total THD was 0.9%, IM was less than 3%, and frequency response was +/- 0.25 db from 30 to 12,000 Hz.  That was about as good as an MW-50 could sound.  The antenna was a half wave non-directional vertical and was quite broadbanded.

Like the article said, EVERY ADJUSTMENT on that transmitter affected the audio performance, from PA tuning and loading, to drive levels, and the "third harmonic resonators" (look up that technique if you have never heard of it).  Those third harmonic resonators "squared off" the PA RF waveform and allowed a substantial increase in PA efficiency.  But those resonators drastically affected the audio performance.

So, the words of advice are that EVERY SINGLE PARAMETER of a modulated transmitter stage will affect the audio performance of that transmitter.  Just a simple "dip and load" or "tune for maximum" won't guarantee that your transmitter is running at its best performance level, whether its a modulated stage or a linear.

73
Ted W8IXY

[ke7trp]

A few years ago I made up about 20 different Antenna tuner designs here while watching a spec AN and an analyzer. I made them up on a wooden board.  I noticed why the tuner was running warm in that it had a very high Q. My Wide AM audio(20KC+) at the time was up against that sharp cut off of the tuner. I swept the tuner and found it was very sharp and the sidebands where up against a very high swr.  The sidebands where at a bad match.  No matter what I did with the low and high end of the audio, the sidebands where attenuated in the form of heat.

I think I made a post about tuners being hifi.  A few passed this off as nonesense. I adjusted the amount of capacitor to inductor to open up the Q and the heat went away.  The tuner could now pass the 20+ KC I was transmitting.  My on air sound was much much better.

Now, The transmitter runs 7KC wide with filters. The lesson was one I enjoyed learning about.  Side band match...  A good one.

Thanks for posting Fred.

C

[flintstone mop]

I think I better check what going on here too. That might explain why some Ham stations have very nice smooth high-end....ya know voice freqs and sibilance,,without the fake FM sound.

I might pass this on to Allan Weiner, WBCQ...Might explain what I'm hearing lately. Very sensitive adjustments on the MW-50

Fred

[kb3ouk]

Not only with the MW-50, but wonder if there might be something going on with the antenna system throwing it off. I have to wonder if the winter weather isn't causing part of the problem. I've noticed that WBCQ seems to sound slightly different in the winter than in the summer.

[Steve - K4HX]

Q=fc/BW

Your antenna system (including the tuner) would need to have a Q of 190 to have a 20 kHz bandwidth at 3.885 MHz. The Q of a dipole would be something like 12-15.

[kb3ouk]

Going by that formula, an antenna with a Q of 12 at 3.885 Mhz would have a bandwidth of 320 khz, which seems wide enough, which fits in with what Clark said about lowering the Q of his tuner improved the audio since the bandwidth would be larger.

[Steve - K4HX]

The point is, it's not likely you could have a Q as high at 190 with any reasonably sized antenna, most especially a dipole. Could a link coupled tuner be set up improperly to produce a Q of 190? Maybe. If you had the tuner set up for parallel feed with a very low-Z input or vice-versa. The likelihood would be greater on 160, especially if one were tuning up an 80M dipole on that band.

That said, even with impedances with resistive components as low as 10 or as high as 2500 Ohms, I don't see the Q being much higher than 50.

[kb3ouk]

And you really wouldn't want to go any higher than a Q of 190 if yo could even get it that high, since the higher you go, the narrower the bandwidth gets.

[ke7trp]

Open wire line and link coupled balanced tuner.  Never used a coax dipole.  Thought about it over the years

[Steve - K4HX]

The current in the tuner would be HUGE - like 40-50 Amps or more. The thing would melt pretty quickly.

[WA3VJB]

I have found that the use of 10ga wire instead of smaller stuff improves the dipole bandwidth.

Cage-type doublets would improve things still further.



The AM Broadcast community had to confront antenna bandwidth when stations considered whether to include Hybrid Digital datastreams on their traditional analog signal.

The datastream required much greater antenna bandwidth (apx. 30Kc) because it covered a far larger chunk of spectrum than did the analog signal (apx. 18Kc).

Because of potential interference issues, the digital signal already was required to operate at reduced power compared to the primary signal, so if anyone wanted to listen to the hybrid digital overlay, their reception could have been hurt by significant signal attenuation blamed on a narrow RF response of the transmitting antenna.

Thankfully, although the allure has mostly passed of using a hybrid digital scheme on the Standard Broadcast band, stations that improved their antenna array and audio chain as part of the experiment can today transmit improved analog signals.

[flintstone mop]

I'll tread lightly on HD radio...tip-toe
IF we didn't have to be compatible all the time, the FCC cudda assigned the digital radio to another band.
HD radio on AM, still some on the air, is phenomenal!! AM is better suited, as the MF freq are not so finiky about multipath.
HD radio on FM is CD quality..more than I can say about XM radio and near-cd quality.
Sorry for side-stepping the topic.
Fred

[K6JEK]

Didn't extremely early sideband transmitters filter out the opposite sideband at the antenna?

[WA3VJB]

Didn't extremely early sideband transmitters filter out the opposite sideband at the antenna?

I wish that worked on receive when somebody moves in nearby.

[flintstone mop]

Didn't extremely early sideband transmitters filter out the opposite sideband at the antenna?

I wish that worked on receive when somebody moves in nearby.

Hey Paul
A software defined radio will do just that. I can make almost any QSO arm chair copy. Even the little softrock I.F. board will do that....sooooooooo nice

[Opcom]

Didn't extremely early sideband transmitters filter out the opposite sideband at the antenna?

I wish that worked on receive when somebody moves in nearby.

Use the USB/LSB switch on the receiver or some tuning is my only respite from such encroachments.

[Steve - K4HX]

I assume you are talking about the stuff done in the 20s and 30s. It was all LF, so the antenna BW there was probably quite small. BW = f/Q, so make f small and BW gets small.

That said, the SSB circuits I've seen in the Bell Tech Journal from the 20-30s all used filtering within the transmitter.

Didn't extremely early sideband transmitters filter out the opposite sideband at the antenna?

[kb3ouk]

Some of the first commercial use of SSB was for trans-Atlantic telephone service at LF, so it could've been possible but would've been more practical to do the filtering in the transmitter.

[K6JEK]

I remembered where I heard about it, in a Ham Nation episode on the history of SSB. A guy named John Carson developed SSB at Bell Labs in 1915. He eliminated the carrier with a balanced modulator but suppressed the opposite side band in the ACU, antenna coupling unit. This episode has some great old schematics. The part about John Carson is around 25 minutes in.

The idea of an antenna system with high enough Q to mess up an AM signal reminded me of this ninety-eight year old scheme which took advantage of it.

http://twit.tv/show/ham-nation/81

[KM1H]

Didn't extremely early sideband transmitters filter out the opposite sideband at the antenna?

Bell did that in the teens over baseband long lines and the same was used on transatlantic cable, it wasnt over antennas.

Check out the March QST for an article on a broadband 80-75M dipole.

[W7TFO]

Over the years I've worked on enough antenna coupling units at AM BC stations to relate to this topic.

I usually use a Delta Operating Impedance Bridge.  It goes into an RF circuit at power (up to 5 kW), and then reads directly the impedance and reactance from separate dials.

My Agilent network analyzer is useful as well, but everything must be off and quiet for it to stabilize, tough to achieve when connected to a 200' tower acting as a receive antenna.

A vertical tower usually has decent B/W for full fidelity, usually at a fairly low self-impedance.  Keep in mind, BC antennas are designed for one frequency, and never move about the dial.

When used in a 50-Ohm environment such as the input to the antenna tuner unit, one can easily determine the BW from the slope of tabulated readings taken at 5 kHz intervals.

It takes time and patience when dealing with a single radiator, and days on the job when immersed in a multi-tower directional .

Lost of tap sliding and cap shuffling.  I've seen coils wound with 1" silver-plated tubing burnt in two from trying to absorb RF meant to go into the antenna but was stopped in a incorrect Pi or T net, yet the input to the 'tuner' read as R50+J0 .

The end result is rid the system of the following: "The sidebands check in but they don't check out" .

Dennis

[K6JEK]

Didn't extremely early sideband transmitters filter out the opposite sideband at the antenna?

Bell did that in the teens over baseband long lines and the same was used on transatlantic cable, it wasnt over antennas.

Check out the March QST for an article on a broadband 80-75M dipole.

Not surprising, even though the patent shows an antenna:

[flintstone mop]

I remembered where I heard about it, in a Ham Nation episode on the history of SSB. A guy named John Carson developed SSB at Bell Labs in 1915. He eliminated the carrier with a balanced modulator but suppressed the opposite side band in the ACU, antenna coupling unit. This episode has some great old schematics. The part about the John Carson is around 25 minutes in.

The idea of an antenna system with high enough Q to mess up an AM signal reminded me of this ninety-eight  year old scheme which took advantage of it.

http://twit.tv/show/ham-nation/81

I think analog TV did the same thing to dump the other sideband in a dummy load. The copper piping needed to tune out the unwanted sideband musta been expensive. Glad that's over with.

[kb3ouk]

With an analog tv transmitter, the signal coming out of the transmitter before it gets passed through the sideband filter is DSB AM. The filter system cuts off a portion of the lower sideband so that the result is a vestigal sideband signal, in this case a complete upper sideband and part of the lower sideband.