Antenna History

The following manuscript was found among the station records in 1994. Author thought to be Arthur M. Kay, W2AIP, SK. It was scanned, OCRd, and converted to HTML by Alan Crosswell, N2YGK. Supplemental content, including historic 1920 photographs was added in 2017.

Multi-wire Aerial Cage, c.1906

Early antenna history of the station reveals an experimental multi-wire aerial cage was strung approximately 90 feet in the air above the campus between chimneys of Havemeyer and Schermerhorn Halls, the cage being self-resonant between 200 and 300 meters and fed with single wire feeder at the Havemeyer end. The Columbia University Experimental Wireless Station (as inquiries reveal, the first of its kind at an American University) was transmitting with this antenna in l906. The station, then consisting of an experimental hi-power spark-gap transmitter located in the basement of Havemeyer Hall, was granted early recognition by the government as Experimental Station, Manhattan Island, and subsequently licensed to transmit in the second supervisory district of the Department of Commerce using the identifier "XM." (It is interesting to note that a similar experimental station was licensed in Queens at the same time under the call "XQ," later "2XQ," "W2XQR," and finally "WQXR.") Unfortunately, exact amateur antenna history at Columbia remains obscured until the period following World War II. The following represents a chronology in antennas used at W2AEE since that time.

[In the 1920 view from old Engineering (now Mathematics) building looking West toward Schermerhorn Hall, a possible antenna mast is visible on Schermerhorn along with guy-wires for it's partner on Havemeyer Hall. In the complementary view of Havemeyer, and another view across University Hall, we clearly see the matching mast. - Ed.]

Overhead Doublet, c.1924

Approx. 50 ft. wire cage assumed supported between 10 ft. vertical wooden supports at North and South cupolas, running lengthwise above Engineering roof, fed with single wire transmission line thru Engineering roof. Exact date of erection unknown, but probably occurred sometime after move of station to Engineering attic, c.1924, Cage later modified to center-fed 35 ft. doublet, using open-wire line. Served as principal post-war ten meter antenna ("two half-waves in phase"), giving fairly good results on that band. The doublet and its wooden supports were finally removed during the fall of 1952.

South Long Wire, 1951

Erected April, 1951. Approx. 160 ft. #10 copperweld wire from South cupola Engineering to Northeast chimney of Mines building, 120 ft. above Broadway, fed at Engineering end with approx. 75 ft. #10 copperweld wire running along roof to glass feedthru insulator near North cupola of Engineering. Feeder unbalanced to roof, supported by 6" porcelain standoff insulators of various types. Feeder tuned by pi-section coupler, rated at 1 kw, AM service. Antenna performed consistently well to South and West on 80 meters, but due to (unbalanced) feeding problems, no serious attempts were made for its usage on the higher frequency bands. Noticeably poor high-angle (local) radiation on 80 meters due to height of antenna and radiating feedline (presenting pattern almost that of a full-wave on that band, with resultant overhead null). South Long Wire modified to Mines Beam during fall of 1956.

North Long Wire, 1952 -

Erected May, 1952. Approx. 90 ft. #12 hard-drawn copper wire from North cupola Engineering to Southwest chimney of Havemeyer, 120 ft. above Broadway, about 90 ft. above the campus, fed at North cupola with approx. 10 ft. #10 soft-drawn copper wire to ceramic feedthru insulator thru roof. Standoff insulators on roof 5" porcelain "beehive" type. Overall performance of antenna considered poor due to similar problems faced by South Long Wire, and further hampered by fact of relative shortness on 80 meters. Too much r.f. in shack on 40 meters due to height of feedpoint above true ground (for any antenna of unbalanced varieties at W2AEE, for that matter on 40 meters). Also, usable pattern directed to Canadian North and out to sea responsible for relatively poor results. Feeder tuned by L-section coupler, rated at 1 kw CW service, 1/2 kw AM service. This antenna recommended to eventually come down.

80 Meter Dipole, 1952

Erected May, 1952. 135 ft. #14 copperweld running from South cupola Engineering to Southeast chimney of Havemever, approx. 100 ft. above the campus, fed at center with RG-59/U coax, feeder running from dipole to North cupola Engineering, thence down ventilating duct directly to fan room, Center joint of antenna made from a coaxial cable connector, one half or dipole attaching to outer (shield) part of connector, the other to center pin by means of egg-type insulator. Due to inherent unbalance and "non-flat untunability'. of the system, this antenna was a single-band affair. Pattern quite poor on 80 meters, with major radiation directed towards Manitoba and out to sea, Antenna also apparently too high for useful high-angle (local) radiation for low frequency work. This antenna taken down during fall of 1955 to be replaced with heavier wire and tuned feeders for multiband operation. Multi-Band #1

2 Meter Coaxial, 1952

Erected December, 1952. Antenna constructed from RG-59/U coax made into vertical dipole 38" overall length. Braid material of the coax is "sleeved" over the feedline for 19", the upper conductor being soldered to a 19" section or #14 hard-drawn copper wire at start of "sleeving." Antenna supported by 10 ft. wooden mast (used previously to support the Overhead Doublet) with ceramic standoffs at top and 38" down from top of mast. Feedline is a "continuation" of the RG-59/U coax and enters station via North cupola Engineering, where the supporting mast is located. The "sleeved-back" braid serves as the lower half of the dipole and at same time prevents feeder radiation. Mast supported by three guys, consisting of 3/16" aluminum wire (stranded), attached to eyebolts in roof (backed-up below roof) through "U"-bolts and turnbuckles (equipped with safety wires to prevent "thumbscrewing") to adjust proper tension. This antenna erected for 2 meter teletype reception from W2BFD/W2NSD RTTY exhibit at Vim's Times Square during Xmas 1952. Performance was excellent with many messages handled via the antenna. It was removed with removal of the VHF and RTTY equipment, the "coaxial" idea being adopted for the ensuing 10 meter coaxial erected in its stead in March, 1953.

10 Meter Coaxial, 1953

Erected March, 1953. Antenna constructed from RG-59/U coax made into vertical dipole 16 ft. overall length, made on exactly same principle as 2 meter coaxial, above. Braid material "sleeved" 8 ft. from center; 8 ft. #14 hard-drawn copper used as top section of dipole. Mast 16 ft. pine 2 x 4, to which are attached two 1 x 2 pieces each about 7 ft. long. Joint overlap is about 1 1/2 ft. Lower end of 1 x 2 section spaced by 2 x 4 pieces, Joints and spacers attached by 6" x 3/8" carriage bolts. Bottom of mast notched to fit "V" of roof peak. Mast supported by 6 guys arranged in two sets of three. Lower guys attached at about 8 ft. level, upper guys at about 18 ft. level. Guys 3/16" galvanized stranded steel (later replaced with double thickness "glasline") and broken-up about every 5 ft. with "egg" compression-type insulators. Joints in guys made as in 2 meter coaxial case, with upper & lower sides of eye bolts passing thru wood load bearing pieces to distribute load over greater area around eye bolts on roof. Antenna worked only fair on 10 meters, probably because no "reference plane" was established beneath it, thereby producing too high an angle of radiation. A simple ground plane placed about 16 ft. below the dipole center would have been desirable to establish a definite "reference" ground, as well as retard feedline radiation which was in evidence. This antenna replaced by H-F Vertical (installed on other, i.e., South, cupola) in September, 1955.

2 Meter Whip, 1953

Erected December, 1953. Antenna constructed on same principle as 2 meter coaxial, above, only using TV aluminum tubing for lower section and Gonset Communicator whip for upper dipole section, with coax fitting in center. Clamped to mast of 10 Meter Coaxial at about 6 ft. level, at North cupola of Engineering. Replaced with 2 meter beam in October, 1956. Performance fair.

6 Meter Quad, 1954

Constructed January, 1954. Basic quad, 53" per side, placed about wooden framework, no reflector or director. Constructed for propagation reliability tests made throughout sprig of 1954 and crew races held that May. Semi-portable in size, antenna was never placed in service and currently rests in South storeroom. May be used in field-strength measurements in tuning and calibration of other 6 meter antenna equipment, if desired.

Multi-band #1, 1955

Erected April, 1955. A modification of the 80 Meter Dipole erected May, 1952, using #12 copperweld wire and tuned feeders. Intended as the station's first antenna designed for true Multiband operation, but due to antenna tuning problems at the coupler, could be tuned for only 80 and 40 meters. 135 ft. dipole center-fed with 450-ohm open-wire line, series-tuned at coupler (later, parallel-tuned). Feeder enters North fan room of Engineering via two porcelain feedthru insulators located near North cupola. Compression-type "egg" insulators serve as center and end isolators. As in case of original 80 Meter dipole, height and directivity patterns are poorly suitable for favorable work on these bands, unless contact with Northern Manitoba is desired. Since actual performance on both 80 and 40 meters has been quite poor using the MB-1, it is suggested that this antenna be removed from its present location (South cupola Engineering to Southeast chimney of Havemeyer), and the possibility of another balanced, multi-band antenna to Low Library be investigated. If such an antenna to Low be feasible, it will greatly add to its benefit if other antennas (such as the MB-1) are not present to absorb and reradiate useful energy otherwise put to good use. The MB-1 has been a very "quiet" antenna as to noise, but then so have the signals emanating and received from it. It has not fulfilled its purpose.

High-Frequency Vertical, 1955

Erected September, 1955. A five-section Ward mobile trailer whip, trimmed to resonate at around 15 Mc., installed on tank mount and mounting bracket assembly affixed to south cupola of Engineering. Fed with a single-wire #12 copperweld along Engineering roof from porcelain feedthru near North cupola to base of whip, using approx. four 4" standoff insulators along roof. Base of whip screwed into tank mount thru special tapped adapter. Base approx. 3 or 4 ft. above peak of roof, feedline forming that amount of vertical radiator to bring resonant frequency to about 14 Mc. Base mounting assembly constructed at W2AEE using weather-treated 2 x 4's, heavy aluminum plate, and 3/8" machine bolts. Assembly universally adjustable. Whip sections unscrew for portable carrying, if desired, and mounting assembly also detachable. Feeder tuned with pi-section coupler for 20-15-11-10 meters, thereby making this a multi-band antenna. Whi;e perforiance is not too bad on 20 meters (and somewhat worse on 10, considering vertical polarization at this frequency), it is perhaps best at 15 meters, where a very close match is obtained. Low angle radiation could be considerably improved, however, by means of ground radials situated beneath the antenna, and the possible use of a balanced feed- line and tuning system, as the roof itself is a poor conductor at high freqs. All in all, for a vertical it is not a bad antenna; however, if and when, a rotary beam for these frequencies would prove a tremendous asset to station daytime and evening time operation as well. Until such a time, the H-F Vertical will have to continue doing its job as an omnidirectional short-wave antenna.

2 Meter Beam, 1955

Erected November, 1955. A standard wide-spaced 4-element yagi TV antenna, cut-down to 2 meter dimensions. Driven element is folded dipole, fed with RG-59/U coax thru a half-wave coax "balun" section. Beam erected on light-duty TV rotator atop 10' steel mast, guyyed to roof at North cupola of Engineering using old 10 Meter Coaxial antenna guy posts. Double thickness "Glasline" used to guy. Although a mismatch is indicated, the beam performs quite well, & good results have been had. Property of Floyd Hollister, K2DUV.

2 Meter CD Vertical, April 1956

Erected April, 1956. A commercial version of the earlier 2 Meter Whip erected in December, 1953. Cut to 146 Mc, designed (as were the others) for 73 ohm coax, it uses RG-59/U fed from North cupola of Engineering and is untuned. Antenna situated above 2 Meter Beam, and feedline parallels same to the station. Some form of "reference ground" would be desirable (in form of a ground plane, 1/2-wave below center point-of-feed) to prevent "antenna" currents from flowing back along outer coax conductor, as well as lower the angle of radiation. (The feedline is noticeably "unflat" with standing waves upon it, indicating some degree of mismatch, as well as the foregoing outer conductor radiation problem; however, overall performance is considered fair to good for the results obtained, but should probably be better for the excellent VHF location using vertical polarization.) The antenna takes its "CD" name due to its primary installation purposes, i.e., the necessity of having a good omnidirectional vertical on this band for the Civil Defense comittments of the station. If and when a more developed horizontally-polarized array is placed into service on 2 meters, the increased gain will probably more than exceed losses due to cross-polarization in working vertically-polarized stations. The only justification in retaining the CD Vertical in that case would be its omnidirectivity for Civil Defense purposes, the maximum realizible gain of the antenna being 0 db. It could probably be mounted in some other location, perhaps farther away from the 2 meter beam, if RG-11/U is substituted as the feedline.

Low Beam, May 1956

Erected May, 1956. An experimental NHF echelon beam designed far 160-80-40 (and possibly 20) meter operation. Two parallel, staggered wires in horizontal plane, running from SE corner Engineering roof to NM corner first parapet Low Library. Insulating sections arranged such that resonant transmitting sections lie in echelon. Balanced transmission line leaves SE corner Engineering South cupola, runs parallel to and between the non-resonant end sections, and feeds both resonant transmitting sections at Engineering end through balanced phasing sections. (A detailed description and theory of the antenna may be found in later pages of this section.These pages missing -ed.) For some carefully investigated but yet unknown reason this antenna failed to live up to its expected performance on the indicated bands. One possible reason is that the concentrated angle of radiation was just too low for the distances for which the MHF bands cover. However, this in itself would not supply the total answer. After a year's constant trial and frustration, the Low Bean was taken down and replaced by the current V Beam which gives far better performance all the way around. The failure of the Low Beam to "work out" will remain one of the antenna mysteries of W2AEE as well as its designer and disheartened operator-helpers.

Mines Beam, 1956

Erected September, 1956. An end-fed "8jk" beam for 75 meters. Two parallel, non-staggered wires in the horizontal plane, running from South end of Engineering roof to the two northern chimneys atop Mines building. Transmission line leaves Engineering North cupola, is supported atop and runs parallel to roof of Engineering to South cupola, and thence to point of balanced feed through balanced phasing sections. Elements 125 ft. long (active) fed in phase opposition. (A detailed description of the antenna may be found in later pages of this section.) The Mines Beam, using the old MB-1 antenna coupler, worked quite well for its intended purpose of a good low-angle radiator to the far west on 75 meters during low sunspot activity. However, for more usual and more practical purposes, it is strongly suggested the two elements be fed in phase (instead of phase opposition) to produce a much more desirable pattern all the way around for periods of normal and above-normal sunspot activity. As an "8JK" beam, the antenna is useful only an one band; As "two long wires in phase" the antenna should produce very desirable characteristics on all bands, 160 thru 10 meters, including optimum directivity to the southwest on the higher bands (and to northern Europe as well). Changing the phasing is a very simple matter (see diagrammissing - ed.) and should at least be tried. The station balanced tuner should cover 160 thru 40 meters, and an additional one should be built for 20 thru 10.

V Beam, 1957

Erected May, 1957. Two 133 ft. elements in horizontal quadrature, one going to cement block atop Northwest corner first parapet low Library, the other going to southeast chimney of Havemeyer, fed at their intersection with 137 ft. 450-ohm #18 copperweld open-wire line from south cupola Engineering, the feedline being supported by twin "Glasline" runners its entire aerial length. Material of elements and supporting-tie sections is #10 copperweld wire. End insulators are two pieces of "Glasline" approx. 2 ft. long between compression-type "egg" insulators. Tie bridles at far ends made from #10 copperweld. Tie bridle for feedline at South cupola made of twin thickness "Glasline" and egg insulators. Feedline supported above Engineering roof at both cupolas, entering North cupola through air duct, and thence to station antenna patch panel in North fan room. Center insulator at feedpoint consists of seven (count em!) thicknesses of "Glasline" at the antenna's weakest point. When in proper tension, the average height of the antenna is about 75 to 100 ft. above the campus. Elements fed in normal V-Beam fashion (i.e., in phase opposition), the resonant feedline tuned by the old MB-1 coupler at the station. Designed for 80 and 40 meter operation, the antenna has produced excellent results on those bands to date, as well as 20 meters too, although at admittedly reduced efficiency due to the wideness of the apex angle on that band. Designed as a high-angle radiator for 80 meters, and a medium-angle radiator for 40 meters, the antenna correlates predictions quite well. With proper feedline tuning, this antenna should have relatively good useful radiation for all bands (10 thru 160 meters). The posiblity of an exponentially-tapered feedline fed with wide-band balun coils for obtaining all-band operation is discussed below. As tensions will vary with season and weather, it is important they be adjusted accordingly and checked fairly often.