History question, re: AM directional antennas

[w9wi]

Just posted this on another board & figured I'd ask here as well:

OK, so it's reasonably well-known in the industry that the first DA was installed at WSUN in Florida in 1927, to protect WTMJ-620 Milwaukee.

A number of history sites have mentioned WLW had to install a second tower in the mid-1930s to null its 500,000-watt signal towards Toronto, where CFRB - then on 690 - was complaining about adjacent-channel interference. The context suggests that DAs were still quite rare at the time.

I've never seen an early station listing - nothing earlier than 1966 - that indicates which stations used DAs. I see earlier listings that seem unlikely without the use of DAs - for example, St. Paul and Washington were both already 50,000 watts on 1500 in 1942. Even on regional channels, the listed 5kw night power of WISN-1150, Milwaukee, seems unlikely to have coexisted with KSAL Salina, Kansas and WAPO Chattanooga (both 1kw night power) without DAs.

So the question is...

When did directional antennas become commonplace among AM stations?

 

[LA_Guy]

I believe the first DA's were used to allow nighttime co-channel power increases at the 1250s in Tampa, FL and Pittsburgh, PA. This was sometime in the mid '20s.
However that said, I've also heard a rumor that around the same time that the 620 in Milwaukee, WI and another 620 were doing the same thing.

 

[TomT]

There were a number of pre-war directional installations. When I arrived at WSPD in the early eighties they were still using the original 1938~1939 installed phasor for that three-tower array. Built for 1340, it was re-tuned to 1370 in 1942 to comply with the NARBA treaty.

 

[rfry]

OK, so it's reasonably well-known in the industry that the first DA was installed at WSUN in Florida in 1927, to protect WTMJ-620 Milwaukee.

Dr. George H. Brown (RCA - Princeton) states in his autobiography and part of which I was that WFLA in Clearwater, FL started using a directional antenna system on 620 kHz in 1931, to protect WTMJ at night. He wrote that it was designed by Raymond Wilmotte, and was the first DA used by a commercial AM broadcast station in the US.

When did directional antennas become commonplace among AM stations?

Brown's book has an entire chapter on DAs he designed in the 1930s, including one for WLW in 1934 to reduce interference to a Canadian station on an adjacent channel when WLW was operating at 500 kW non-D.

Probably DAs could not be considered as commonplace in the 1930s, but they did exist then.

//

 

[radiosaur]

There were a number of DAs in the planning and building stages that were delayed by WW II. I would say the era of the DA started in the mid to late 30’s, paused between ‘41 and ‘45 and became commonplace after the war was over.

 

[DanStrassberg]

Dr. George H. Brown (RCA - Princeton) states in his autobiography and part of which I was that WFLA in Clearwater, FL started using a directional antenna system on 620 kHz in 1931, to protect WTMJ at night. He wrote that it was designed by Raymond Wilmotte, and was the first DA used by a commercial AM broadcast station in the US.

Brown's book has an entire chapter on DAs he designed in the 1930s, including one for WLW in 1934 to reduce interference to a Canadian station on an adjacent channel when WLW was operating at 500 kW non-D.

Probably DAs could not be considered as commonplace in the 1930s, but they did exist then.

As I understand it, the statement about the first AM DA in the US being at WFLA is only PARTIALLY correct. Back then, WFLA and WSUN St Petersburg shared time on 620 and used the same transmitter facility. Time-shares were very common among large-market (and a few not-so-large maket) AMs in the late '20s and early/mid '30s. As for WFLA then being licensed to Clearwater, rather than to neighboring Tampa, as is usually reported and as the station with the WFLA calls is today, someone much more familiar with the market and the stations will have to settle that one.

Anyhow, WTMJ protested the interference from the Tampa Bay-area stations and the FCC (I think it was still the FRC back then) threatened to restrict the Florida stations to daytime-only operation if they could not satisfy WTMJ's complaints some other way. Apparently, one or more MW DAs were already in use (probably only experimentally) in Europe and publication of technical articles about the technology provided the inspiration for the installation at WFLA/WSUN. The WFLA/WSUN phasor was apparently constructed using transmission lines rather than lumped circuit elements, which complicated tuning of the simple two-tower cardioid pattern. Reportedly, the second AM DA was built not long afterward, also in Tampa Bay not far from WFLA/WSUN, at what was then WDAE and is now WHNZ 1250. That system was the first (in the US at least) to use a lumped-element phasor. An interesting footnote to this story is that, for the last maybe seven years, the former WDAE on 1250 and the current WFLA on 970 have been diplexed from the same six-tower site, which was built by Clear Channel, new from the ground up for both stations.

BTW, I have seen a plat of the WLW site dating from the days of 500 kW DA-N. There were three towers, not two, in a triangular configuration. The main tower was, of course, WLW's famous Blaw-Knox diamond tower, which still stands. I believe that the two shorter self-supporters are long gone, but one of them might still be standing, and if so, it could possibly be used an an auxiliary.

 

[rfry]

The WFLA/WSUN phasor was apparently constructed using transmission lines rather than lumped circuit elements, which complicated tuning of the simple two-tower cardioid pattern.

According to a Radio Guide article about this first DA:

'The RF circuit was quite simple, with the transmission lines cut to quarter wavelength, and routed in a curved way to the towers. One tower was coupled from the transmitter with a coil to shift the signal -45 degrees; the other was coupled with a condensor, to shift the signal +45 degrees. The phase angle between the two towers was 90 degrees. The result was a tight cardioid pattern, with the null at 346 degrees, toward Milwaukee.'

So it appears that this system did have an elementary way of setting the relative phase between the two towers, whether or not the coil and "condensor" were easily adjustable.

Here is a link to the complete article...

http://www.oldradio.com/archives/stations/tsp/WSUN-WFLA.pdf

//

 

[nmoore6676]

Dr. George H. Brown (RCA - Princeton) states in his autobiography and part of which I was that WFLA in Clearwater, FL started using a directional antenna system on 620 kHz in 1931, to protect WTMJ at night. He wrote that it was designed by Raymond Wilmotte, and was the first DA used by a commercial AM broadcast station in the US.

Brown's book has an entire chapter on DAs he designed in the 1930s, including one for WLW in 1934 to reduce interference to a Canadian station on an adjacent channel when WLW was operating at 500 kW non-D.

Probably DAs could not be considered as commonplace in the 1930s, but they did exist then.

As I understand it, the statement about the first AM DA in the US being at WFLA is only PARTIALLY correct. Back then, WFLA and WSUN St Petersburg shared time on 620 and used the same transmitter facility. Time-shares were very common among large-market (and a few not-so-large maket) AMs in the late '20s and early/mid '30s. As for WFLA then being licensed to Clearwater, rather than to neighboring Tampa, as is usually reported and as the station with the WFLA calls is today, someone much more familiar with the market and the stations will have to settle that one.

Anyhow, WTMJ protested the interference from the Tampa Bay-area stations and the FCC (I think it was still the FRC back then) threatened to restrict the Florida stations to daytime-only operation if they could not satisfy WTMJ's complaints some other way. Apparently, one or more MW DAs were already in use (probably only experimentally) in Europe and publication of technical articles about the technology provided the inspiration for the installation at WFLA/WSUN. The WFLA/WSUN phasor was apparently constructed using transmission lines rather than lumped circuit elements, which complicated tuning of the simple two-tower cardioid pattern. Reportedly, the second AM DA was built not long afterward, also in Tampa Bay not far from WFLA/WSUN, at what was then WDAE and is now WHNZ 1250. That system was the first (in the US at least) to use a lumped-element phasor. An interesting footnote to this story is that, for the last maybe seven years, the former WDAE on 1250 and the current WFLA on 970 have been diplexed from the same six-tower site, which was built by Clear Channel, new from the ground up for both stations.

BTW, I have seen a plat of the WLW site dating from the days of 500 kW DA-N. There were three towers, not two, in a triangular configuration. The main tower was, of course, WLW's famous Blaw-Knox diamond tower, which still stands. I believe that the two shorter self-supporters are long gone, but one of them might still be standing, and if so, it could possibly be used an an auxiliary.

Neither of those "suppressor" towers still exist. I can remember on a couple of occasions when my dad would drive down route 42 for some reason going from Springfield to Cincinnati, this was before the freeways. Anyway I can recall looking up at the massive diamond tower with the WLW neon sign. In those days "Everybody's Farm" still existed and there was not much else around and no other towers for certain. The current auxiliary tower was built more recently, 1998, and it replaced an earlier one that I don't recall so it may have gone up after the 1950's which was the era when my dad took his back road route.

 

[Bob1370]

In answer to how soon DA systems became common, they were popping up all over the place within a few years after the WFLA/WSUN system became a topic of common knowledge in the business, meaning about 1936-37.

WWJ in Detroit, operating on 920 in the pre-NARBA days, installed a small second tower to allow it to hike night power from 1000 to 5000 watts around 1937. (That system has nothing to do, by the way, with the six tower monster CBS built in the downriver district just seven years ago that allowed them to bump up to 50,000 watts 24/7 on 950 and scorch the market with their signal.) A similar two tower DA was installed in the late 1930s at WBZ in Boston, and another at WEAF in New York, not to protect anyone but to null out open water and increase field strength in populated areas in their core markets--those two plants were fulltime DA-1s. WEAF successor WFAN now runs a single stick, but the WBZ system, although substantially rebuilt, is still in use today, nulling out the Atlantic Ocean and giving WBZ an effective power boost over virtually its entire market area 24/7. WOR in New York also went to a 50 kW DA-1 system in 1936-37 to concentrate its signal in what was then the populated area of its market. (Urban sprawl left some growing parts of north Jersey in the null after World War II but the two postwar rebuilds of their transmitter plant haven't changed the pattern much because they still get a stronger reach in the city, Long Island and the Hudson Valley than they'd get with a single stick).

After about 1938-39 a lot of regional and even 1-B clear channel stations started working on directional nighttime antennas, and in some cases round-the-clock DAs, to boost their signal strength within their market population cores. KSTP St. Paul and WJSV Washington began building 50 kW DA plants while they were still on 1460, WKBW Buffalo while it was still on 1480, but didn't finish before the NARBA channel changes kicked in. They just tweaked their phasors and antenna tuners to get similar or better signal on their new channels once they finally powered the new rigs up in 1941. They enabled KSTP and WJSV (later WTOP, now 3WT) to bump from 10,000 to 50,000 watts, and KB to jump from 5,000 to 50,000. Lots of Class III stations were able to increase from 1,000 to 5,000 watts fulltime, pulling in their patterns toward cochannel stations while increasing them in other directions where the bulk of their market populations as of 1940 were living. WBEN in Buffalo, to give one example, relocated its transmitter from a plot of land in southern Niagara County to Grand Island in the Niagara River in 1940 to allow it to go from 1000 to 5000 watts fulltime on 900; its channelmate WKY in Oklahoma City did likewise at the same time, and each of them created plants with broad round-the-clock coverage with a single null kicking in toward each other's city of license after local sunset to protect the other's nighttime prime market. The NARBA channel change in 1941 required slight retuning and re-phasing of the night pattern to move the signal to 930 (where each station still operates today)...no big deal, just a swap of crystals and a tweaking of the phasors and antenna tuning units, all preplanned for months and completed during an overnight silent period on the big changeover day.

Starting in 1941 the DA was enough of a proven technology that the FCC started looking at licensing new stations in underserved markets and accepting applications. Existing stations on graveyard channels, including the antecedent of the one I work for in Rochester, NY, started looking to the DA to expand their reach on a less congested channel. Some, including mine, got construction permits late in 1941 and early 1942--but those permits stayed on hold until after the war because of a freeze on availability of everything from transmitters and phasors to tower steel. Our late 1941 CP (which would move our predecessor station from 250 watts on 1240 to 5 kW-DA-N on 1370) wasn't built out until 1946. At least four other companies in our town applied for construction permits for licenses involving either fulltime or nighttime directional operation between 1941 and 1945. Three of them (one an existing station stuck at 1000 watts through the war, two others brand-new operations) got them...all of them were built out in 1947.

Now you know why the number of operating AM stations in America, stuck at about 800 nationwide before the war, nearly tripled to over 2000 by 1950, and rose again to 4000 by 1970. Growing population and urbanization of markets under-served by radio made it economically possible, and the directional antenna made it technically possible to put many of these stations on the air fulltime when, before the DA, they'd have been consigned to daytime-only operation and would have struggled to survive.

 

[DanStrassberg]

WOR in New York also went to a 50 kW DA-1 system in 1936-37 to concentrate its signal in what was then the populated area of its market. (Urban sprawl left some growing parts of north Jersey in the null after World War II but the two postwar rebuilds of their transmitter plant haven't changed the pattern much because they still get a stronger reach in the city, Long Island and the Hudson Valley than they'd get with a single stick).

WOR was constrained somewhat by co-channel CKVD in northern Quebec. To what extent, if any, co-channel stations in Los Angeles and Seattle entered into the WOR antenna design, I don't know. But back in the '40s, WOR used to boast that it served more people at night than any other US station. Implicit in that claim was the reason--WOR's pattern concentrated the signal in areas of high population density along the East Coast. The original DA in Carteret NJ sent signal maxima toward Philadelphia and Hartford, both about 90 miles aeay and approximately 180 degrees apart. That DA was a three-tower in-line that used as its center element a drop wire, suspended from a horizontal wire attached to the tops of the two self-supporting end towers. Was this a conversion from a nondirectional long-wire setup? I don't know but I know two people who are positive--one is positive that it did indeed begin life as a long-wire, the other is just as positive that it was designed from the git-go as a three-element DA.

WOR's 1960-vintage three-tower DA in (I think) Lyndhurst NJ differed considerably from the Carteret system. The Lyndhurst setup was asymmetrical (as is WOR's current array, which is only a couple or three years old). Instead of the relaxed figure-eight pattern produced by the Carteret array, the new setup produced a sizeable minor lobe directed toward Philadelphia to the southwest and an even stronger major lobe to the east. Daytime coverage of central and northern CT was sacrificed to obtain better in-market coverage of Long Island and southern CT. The new array, which suppresses radiation to the northwest even more than did the Lyndhurst array, was completed only months before the demise of CKVD, the station whose presence, along with the "ratchet rule," necessitated the signal reduction to the northwest. Of course, CKVD, though gone and unlikely ever to be replaced, is still notified to the US, so it must be protected as if it were still there.

 

[rfry]

WWJ in Detroit, operating on 920 in the pre-NARBA days, installed a small second tower to allow it to hike night power from 1000 to 5000 watts around 1937. (That system has nothing to do, by the way, with the six tower monster CBS built in the downriver district just seven years ago that allowed them to bump up to 50,000 watts 24/7 on 950 and scorch the market with their signal.)

Regarding the present, downriver DA-2 of WWJ -- it should be noted that the pop count now within WWJ's daytime 2 mV/m contour is much less than that of 50-kW, non-D WJR in Detroit, even in the azimuth segments where the ERP of WWJ exceeds that of WJR.

The reasons for this are related to the higher frequency used by WWJ, and the greater distance of the new WWJ site toward the sector of their high ERP, with respect to that of WJR.

Given the quite narrow day/night azimuth patterns of WWJ, 50 kW non-D WJR has a far greater pop count in its useful day/night coverage areas than does WWJ.

//

 

[DanStrassberg]

Regarding the present, downriver DA-2 of WWJ -- it should be noted that the pop count now within WWJ's daytime 2 mV/m contour is much less than that of 50-kW, non-D WJR in Detroit, even in the azimuth segments where the ERP of WWJ exceeds that of WJR.

The reasons for this are related to the higher frequency used by WWJ, and the greater distance of the new WWJ site toward the sector of their high ERP, with respect to that of WJR.

Given the quite narrow day/night azimuth patterns of WWJ, 50 kW non-D WJR has a far greater pop count in its useful day/night coverage areas than does WWJ.

Everything else being equal...frequency, power, antenna efficiency, soil conductivity... over any real ground (meaning any ground with finite--meaning non-infinite--conductivity) the nondirectional station will ALWAYS (meaning in EVERY case) include a greater area within ANY defined signal-strength contour than will the directional station. The reason is that if the conductivity is less than infinite, the signal strength drops off with distance faster than in direct proportion. That is. the signal strength will be proportional 1/d^x, where d is distance and x is an exponent that is always greater than 1; that is, x increases as the soil conductivity decreases. Thus, to maximize the area included within a given contour you want to maximize the angle within the "beam" of the pattern. You can't make the beam any wider than 360 degrees, which is the nondirectional case. It also follows that the narrower you make the beam, assuming you keep the frequency, power, antenna efficiency, and soil conductivity constant, the smaller the area included within any given field-intensity contour. If the population is evenly distributed within the service area, the population served will be directly proportional to the area covered.

 

[rfry]

You can't make the beam any wider than 360 degrees, which is the nondirectional case. It also follows that the narrower you make the beam, assuming you keep the frequency, power, antenna efficiency, and soil conductivity constant, the smaller the area included within any given field-intensity contour.

Understood, but my point is that even at the peak of the WWJ daytime azimuth pattern where its ERP is around 492 kW, their 2 mV/m contour in that direction is located at about the same place as that of 50 kW WJR, for the reasons I gave. I thought most people would find this of interest, and somewhat unexpected.

//

 

[DanStrassberg]

Understood, but my point is that even at the peak of the WWJ daytime azimuth pattern where its ERP is around 492 kW, their 2 mV/m contour in that direction is located at about the same place as that of 50 kW WJR, for the reasons I gave. I thought most people would find this of interest, and somewhat unexpected.

Isn't a good part of the closeness of the 2 mV/m daytime contours to the north attributable to the fact that WWJ is more than 10 miles south of WJR? Also, I notice that you based your 492 kW number on an inverse-distance field of ~316 mV/m/kW @ 1 km, which is pretty typical for 1/4-wave towers, but four of WWJ's five daytime towers are 138 degrees and the #1 tower is 168 degrees, which explains why WWJ's standard-pattern RMS is more like 355 mV/m/kW. Using 355 would yield an equivalent power of something like 377 kW instead of 492. Meanwhile WJR's 195 degree stick produces an RMS field of more than 400 mV/m/kW @ 1 km. So WJR has not only the 760:950 frequency advantage, it is also, in effect, radiating 30% more power. In addition, if you used the Class B minimum efficiency of 281.7 mV/m/kW @ 1 km to calculate WWJ's peak power, you'd come out with ~617 kW. I guess 492 is just about half way between 377 and 617, so maybe that's a good number to use;>)

 

[rfry]

Isn't a good part of the closeness of the 2 mV/m daytime contours to the north attributable to the fact that WWJ is more than 10 miles south of WJR?

Yes, and I included that point in my original post comparing the two stations (2nd paragraph).

Also, I notice that you based your 492 kW number on an inverse-distance field of ~316 mV/m/kW @ 1 km, which is pretty typical for 1/4-wave towers, (etc)

What I did as a quick estimate was to convert the 6653 mV/m peak, augmented field of WWJ at 1 km (from the FCC website) to the ERP required to produce it with respect to the ~typical inverse distance field at 1 km for 1 kW radiated from an efficient, ~1/4-wave stick.

The math: ( 6653 / 300)^2 = 491.8, approx.

Meanwhile WJR's 195 degree stick produces an RMS field of more than 400 mV/m/kW @ 1 km. So WJR has not only the 760:950 frequency advantage, it is also, in effect, radiating 30% more power.

Your number of ~400 mV/m at 1 km for 1 kW radiated from WJR's 195-degree stick leads to an RMS of 2828 mV/m at 1 km for 50 kW of applied power, compared to WWJ's daytime RMS of 2513 mV/m for 50 kW of applied power to its array. This difference is mostly related to the fractional wavelength heights of the radiators involved.

And of course, even though the peak, daytime ERP of WWJ is many times greater than that of WJR, the 2 mV/m groundwave contours of both these stations in the direction of WWJ's peak ERP both fall at nearly the same physical location -- which demonstrates that raw ERP may not always "scorch the market" quite as impressively as first thought.

//

 

[semoochie]

I understood that at 500kw, WLW interfered with WOR and had to install a directional antenna. It was noted that the amount of signal in the null was 50kw! I never heard anything about Toronto.

 

[DanStrassberg]

I understood that at 500kw, WLW interfered with WOR and had to install a directional antenna. It was noted that the amount of signal in the null was 50kw! I never heard anything about Toronto.

There could easily have been interference to WOR but I have seen a crude polar plot of the pattern. The radiation minimum was toward the northeast--toward Toronto. Was the minimum equivalent to 50 kW? Possibly; I don't recall whether it was more or less than that. MW DA technology was in its infancy back when the WLW array was built. I think that if today's technology could have been used, a design with the same objectives would have resulted in a considerably different implementation.

 

[JohnnyElectron]

Speaking of 500KW, if I had my way, I'd petition the FCC to move 90% of the AM stations to a new FM digital band, simulcasting for 3 years, then the only stations left on AM for the most part would be 500KW - 750KW AM "Super Stations" with true clear channels once again, 20KHz spacing minimum, using single omni towers, analog audio bandwidth of 20KHz, and they could actually play music again! They would maintain analog so as to be useful for national emergencies and maximum compatability with the hundreds of millions of analog AM receivers out there. No DA required! I could even see SiriusXM putting some on the air as an advertising 'barker' channel rotating the playing of one of their channels everyday as an advertisement.

...End of dreamworld, back to reality....

 

[DanStrassberg]

Speaking of 500KW, if I had my way, I'd petition the FCC to move 90% of the AM stations to a new FM digital band, simulcasting for 3 years, then the only stations left on AM for the most part would be 500KW - 750KW AM "Super Stations" with true clear channels once again, 20KHz spacing minimum, using single omni towers, analog audio bandwidth of 20KHz, and they could actually play music again!

If the receiver manufacturers who want you to spend $200 for an HD radio would build high-quality analog AM/FM receivers whose circuit design, component cost, manufacturing cost, and performance justified a $200 selling price, AM would sound just great! However, for decades, the conventional wisdom has been that the public will not pay for such receivers. I don't get it. From what I've read, no manufacturer has yet shipped an HD radio that performs acceptably on the AM band. Why do the manufacturers think the public will spend $200 a copy for HD radios that simply don't perform acceptably on the AM band but won't spend the same amount (or somewhat less) for great sounding analog receivers with good sensitivity, selectivity, sound quality, interference rejection, and immunity from overloading on strong signals on both bands.