I have lived with the GAP Titan DX antenna at my QTH since summer of 1998. During that time I have observed much lively discussion regarding the performance of this antenna. Anecdotal claims have ranged from outrageously great to extremely poor. Since I had already designed monoband replacements for the Titan, I decided that it was an opportune time to make quantitative performance measurements.
General:
The GAP Titan DX is an unusual antenna, with optimistic claims made by the manufacturer. It is appx. 25 feet high, and split in the middle with an insulator. It is claimed to function as a shortened vertical dipole (or doublet if you prefer), without needing any ground radials. The Titan utilizes a complex arrangement of linear loading/tuning rods, an internal coaxial tuning stub, and a small counterpoise at its base. All of this allows it to operate with "acceptable SWR" on every band from 80 through 10 meters, including WARC bands, without any active tuning. The complexity of the loading and tuning makes this antenna difficult to model, and even harder to understand which parts are actually radiating for each frequency range. The manufacturer publishes little to aide in understanding its operation. As best as I can determine, on various bands it performs as everything from a full wave to a loaded 1/8 wave radiator. This makes actual performance tests of interest to me, for as we all know, the fact that it "tunes up" on all bands does not say anything about its pattern or gain.
Rather than compare the Titan to other commercial verticals, I decided to compare it to a series of monoband verticals. These reference antennas are easy to model, and thus provide an absolute frame of reference for the Titan. One may thus be able to deduce the actual pattern of the Titan, and make hypothetical comparisons of it to other vertical or horizontally polarized antennas, placed in other surroundings.
Due to the contentiousness of much of the GAP discussions, I feel compelled to state that I have no connection with the manufacturer, nor any position to defend either promoting or denigrating the antenna. My sole interest is in performing an interesting experiment, controlled to the best of my ability. Hopefully these tests will lay the foundation for more unbiased and rational antenna performance commentary by the Amateur Radio community, and lead the way for other hams to publish quantitative antenna test results, rather than anecdotal testimonials or condemnations.
General Methodology:
Siting:
All testing was performed at the AA3RL QTH in Crownsville, MD. The Titan and all reference antennas were mounted at an elevation of 5 feet. An Ameritron RCS-8V remote coax switch is used to alternate between the antennas.
Ground: The antennas were placed on a pier over brackish (mildly salty) water. The ground (water) has conductivity of appx. 1.6 S/m. and dielectric constant of appx. 80. Water depth is sufficient that it can be considered homogenous for the antennas at each frequency.
Transmission Line Loss: Transmission line losses are calculated based upon the length of the lines, type of cable used, and the SWR of each antenna. TLA was used to make the computations. Specific values are indicated for each test if relevant. When the losses are not equalized, an adjustment is made to the relative gain figures.
Mutual Coupling: Since there is limited mouting space available on the pier, mutual coupling and pattern distortion is of major concern. Once placement is established, the reference antenna is modeled along with the Titan using EZNEC. The Titan is simulated as a vertical dipole or monopole of similar height depending on the band. The feedline impedance of the inactive antenna is included in the model, and where necessary its length is designed to detune the unused antenna.
Signal Strength: The antenna with the stronger signal has an MFJ-762 Step Attenuator placed in series with its feedline. The manufacturer claims accuracy of 0.2 dB or 5% of total attenuation, whichever is greater. Due to cost concerns, I elected NOT to send the unit out for commercial calibration.
On the Air Procedures: An effort is made to conduct performance measurements over an extended period of time - at least several days. This will allow morning, afternoon, and night time tests. It will encompass periods of varying solar flux and propagation conditions. It will include sufficient sample size - at least 40 - 50 measurements, and will encompass many azimuth directions. The tests will be receiving only, and will last a minimum of 1 to 2 minutes each to average short term fluctuations. Rapidly fluctuating signals will not be used. Signal strength must be adequate to register on the S-meter, and thus provide a visual indicator of strength. The attenuator is used to equalize the signal strength viewed on the S meter, thus eliminating any dependence on the meters linearity, and thankfully avoiding any mention of "S-units" (whatever THEY are). The receiver used is a Kenwood TS850S, and all settings remain the same during the course of the tests: AGC - Fast, AT - OFF, RF Gain - constant. The feedlines are periodically routed to different RCS-8Vswitches to check for any variation. Most contacts are made on the CW bands, and a written log is kept of every contact, with the azimuth and distance calculated. Any correlation of relative gain with path length (distance between stations) is noted.
Performance Test Results:
Gain of GAP Tital DX compared to monoband reference antenna:
*** 80 meter results will be forthcoming.
Band: 10m15m20m40m80m
Titan DX, Average -11.0 -7.8 0 -7.6 n/a
Relative Gain, dB
Reference Ant 4.89 4.68 4.54 4.62 n/a
Gain, dBi
Reference Ant 10 10 10 10 n/a
Launch Angle, deg.
Test Details.
*** I would urge anyone seriously interested in this study to review the "Test Details" page so that they may evaluate my methodology, and judge for themselves the relevance of each test.
Click here for test details.
This page is published by Mike Banz, AA3RL
as a service to the Amateur Radio community.
Please distribute freely.
The author welcomes any questions, criticisms, or compliments via email.
email Mike with comments or questions.