Wullenweber CDAA on Galeta Island circa 1991

Wullenweber circularly-disposed antenna arrays (CDAAs) are large HF direction-finding antennas, sometimes colloquially referred to as "elephant cages", that were popular in the Cold War. They allowed the bearing of an incoming signal to be located to perhaps 0.5° (for the US Navy's AN/FRD-10). I've been fascinated by them ever since I saw the one on the Silver Strand south of San Diego, California. Dozens were built all over the world. Now nearly all have been demolished. The Wikipedia article for the US Air Force's AN/FLR-9 says "advances in technology have made the FLR-9 obsolete."

What advances in technology made Wullenweber CDAAs obsolete? What technologies are used for military HF direction finding now?

EDIT: I'm talking about fundamental advances in technology. I presume if the technology were still broadly useful, then the electronics would be modernized. To rephrase the question, why do modern militaries not use giant circular antenna arrays for HF direction finding any more? What do they use instead?


Two major reasons HF direction finding arrays like that aren't particularly useful any more (you won't find many fixed HF DF stations at all, even more modern ones):

  • As William said, HF is not used very frequently for military communications these days, it is primarily a backup to satellite systems for long distance communication, and what communications occur tend to be more mundane and of less tactical and strategic interest. You aren't likely to hear a carrier group in theater chatting on HF these days, they know that's easily intercepted and will use systems that require geographic proximity to detect (like satellites). You're much more likely to hear encrypted links between fixed stations (bases) or inconsequential communications between assets not of much interest. Some asset phoning home because of a maintenance issue, or a routine training net, etc, and you already know where those sorts of assets are: "At home".
  • Why use a massive, fixed array such as this, when an aircraft (there are several in the US inventory) can use synthetic aperture to do direction finding of similar or better accuracy, from a position much closer to the transmission in the first place, perhaps eliminating the uncertainty injected by the ionosphere, and making the intelligence far more actionable?

These days, direction finding work is of the most acute utility on the tactical level, not strategic. You've got satellite imagery and various long distance surveillance platforms for finding carrier groups in the middle of the ocean. Those assets can't help you find smaller elements such as a platoon or even just a lone operative, so you employ an asset that is operating on the level where the things they're going to look for are of most interest, primarily the local/regional scale, and a single aircraft may be able to cover an area the size of a small country for such work.


First, there are better DF antennas now and they take up way less real estate. Some of them are .. the Pusher and "L" and "T" arrays. The electronics for these is more sophisticated than the conventional "Elephant Cage" CDAA but they are faster if the electronics is designed and implemented correctly. The FCC in Laurel Maryland took their CDAA down and installed an "L" and "T" array (I can't remember which now) In addition, since the solution is an electronic one instead of mostly mechanical (manual goniometer run by a motor) it is far easier to remote ALL of the controls so that it can be operated from anywhere given the correct control protocols.

Second, the old HF BULLSEYE NET as we knew it has been disabled now with various sites being returned back to the host governments in most of the foreign countries.

I hope this has helped answer your question.


First of all: why is it surprising that a radio device built in the 1960s has become obsolete?

I don't know the specific reasons why this specific type of direction finder has become obsolete, but among these reasons might simply be:

  • cost of renovation > benefit
  • Shift in strategic demand for localization of these boring HF signals
  • instead of a few circular setups, a lot of smaller setups can actually do the same or a lot better. We can now phase-synchronize distributed receiver systems through e.g. GPS. That wasn't possible in the 60s!
  • distributed smaller multi-antenna systems allow for actual localization instead of just direction finding; the localization happens inherently and with less error instantly, instead of detecting multiple directions of arrival on multiple old systems, and then crossing lines on a map, manually, after communicating these bearings.
  • terrible, terrible receiver bandwidth, noise, power consumption, reliability, operation cost. (best case: noisy germanium transistors. More likely in the early 60s: tube amps that eat kW of power for such a big receiver array). Noise Figure 7dB for a low-bandwidth receiver? That's not even remotely good nowadays.
  • Operation manual suggest these amplifiers are lock-in amplifiers, and thus would be unsuitable for everything but AM and low-modulation index FM, and thus especially unsuitable for digital communications, probably.
  • terrible antenna setup for systems where computational power allows to get higher resolution with fewer antennas
  • antennas optimized for narrowband reception. Modern RF localization systems can track and find a lot of signals at once, spread over a large bandwidth, with a single set of antennas, through DSP
  • Have you seen how large these are? They're positively a big waste of space.
  • Resolution of this system is very bad (15dB directional gain of the beamforming system) compared to modern 3-antenna setups, which are smaller, cheaper. Manual says "nominal resolution 4°". That is seriously sub-par for anything built since the 90s with that many antennas to increase resolution.
  • analog beamforming depends on physical properties of a system that are hard to maintain over time, and costly to re-calibrate
  • availability of mobile systems that are better
  • seriously, these are from the early/mid sixties. That's 50 years in the past. Even with ever so slight improvements in technology, it's not very likely that these would still be on par with modern technology. We have better
    • cables,
    • lacquer,
    • power supplies,
    • displays,
    • operation monitoring,
    • building materials,
    • thermal isolation,
    • cooling,
    • antenna materials,
    • antenna simulation and
    • measurement tools,
    • higher construction placement accuracy

      just to name a few things where there has been gradual, but not necessarily revolutionary improvements with respect to construction of antenna array systems in the last, I repeat, 50 years.

It's really no surprise that the military has adopted different systems since then.

Generally, however, CDAAs aren't obsolete – circular antenna sets still have nice properties when paired with the appropriate DSP algorithms. Thus, they are still in active usage in SIGINT; it's really just that 1960's analog systems are really not relevant anymore.

  • $\begingroup$ The question is about the antenna arrays, not the original 1960's analog electronics that were replaced long ago. $\endgroup$ – rclocher3 Nov 22 '16 at 19:39
  • $\begingroup$ Well, if the current manual PDF is any solace, all the performance metrics I used in my answer are currently accurate. $\endgroup$ – Marcus Müller Nov 22 '16 at 19:56
  • $\begingroup$ My question is about all Wullenweber CDAAs, including the Russian Krags, the British Pushers (AN/FRD-13), and the German array at Bramstedtlund, not just the AN/FRD-10s and the AN/FRN-9s. $\endgroup$ – rclocher3 Nov 22 '16 at 20:12
  • $\begingroup$ well, as said, there's still CDAAs active as SIGINT stations. I don't think they're Wullenwevers, though, simply because, as I tried to explain, that design has become obsolete through technical progress of the last decades. $\endgroup$ – Marcus Müller Nov 22 '16 at 20:21

Having worked as an intercept operator at these sites in the 70's I can say that there just isn't anything to intercept on the HF bands anymore.

  • $\begingroup$ Ok, having experience with military RF systems in the last decade, I can say he's exactly right. See my post above. $\endgroup$ – Hamsterdave Jun 4 '17 at 20:37
  • $\begingroup$ I have spoken with the Naval Security folks who run the elephant cage at Elmendorff and they are quite operational, they assure me. $\endgroup$ – SDsolar Jun 7 '17 at 4:43
  • $\begingroup$ @SDsolar Apparently, its days are numbered: sandiegouniontribune.com/military/… $\endgroup$ – Mike Waters Jun 19 '17 at 0:24
  • $\begingroup$ That's a shame. Thanks for sharing that, @Mike. I remember flying around with my buddy practicing approaches and departures into and out of Merrill Field downtown and it was just always there. We used it as a VFR navaid when we were out over the water. $\endgroup$ – SDsolar Jun 19 '17 at 4:25

Wullenwebers have been replaced by the FCC HFDF ("huff-duff") system. Several HFDF installations are located in the USA along both the north and south borders and the east and west coasts, plus Alaska, Hawaii, and other locations such as some US possessions. Most of them are unmanned.

Google Maps has a listing of their locations. You can zoom into each location and view the antenna supports, but it's tough to make out all the wires so as to determine exactly what the antennas are. In spite of that, it appears most of those sites use a single V-beam, judging from the support structures (which is all I could make out from the satellite images at maximum zoom a few years back). A few apparently used the older Wullenweber arrays. You can certainly see enough to realize how simple these small HFDF antennas are compared to a Wullenweber!

The FCC HFDF installations all tie in to a manned but automated central control headquarters. When someone transmits, their precise location is quickly recorded and with far greater accuracy than expensive Wullenweber arrays could.

And the FCC uses it, according to this ARRL page about it. For example, the FCC wrote a letter to a wayward ham that "...you were heard by staff at the Commission's High Frequency Direction Finding (HFDF) Center communicating repeatedly on 14.313 MHz ..."

  • 1
    $\begingroup$ What makes the HFDF system more accurate than the old CDAA-based systems? $\endgroup$ – rclocher3 May 15 '19 at 15:48

The demise of the military CDAA is not an issue of capability, accuracy or even economics, but the primary driving factor is the change in mission, targets and technology. HF; whether it's SSB, CW or even digital, is simply not today's prevailing set of tactical or operational frequencies. In addition, satellites and other passive technologies do not require the target to transmit a signal!

One comment: By the late 1960's, CDAA target location was highly automated, including multiple computer generated bearings, and refined elyptical search areas with each bearing. Fron the time of the automated bearing request, location times were better than 2min for NB, with WB better than 5min. WB times were NOT dependent on the antenna or the recievers, but on the technology required to address the variation of the transmitter center frequency for the short duration signal. New technology in mid 1970's solved this issue, automating WB bearings. Repeated localizations provided target tracking data to commanders, supporting field observations and field data. By the 1970's target tracking was part of a secure database including date, target, track and confidence included...

Accuracy: The CDAA design, technology and field implimentation means that the accuracy of each site's bearings are highly impacted by the level of maintenance and consistency of the antenna and the individual elements. Without daily antenna and signal handling maintenance, each site deviates from the "best case", which is listed in the classified tech documents and performance testing as much better than 2.5 degrees!!

Typo: CDAA. not CBAA NB = Narrow Band. (CW, SSB, etc) WB = Wide Band (Burst) (Soviet Submarine)

  • $\begingroup$ Hello and welcome to ham.stackexchange.com! $\endgroup$ – rclocher3 May 15 '19 at 15:44
  • 1
    $\begingroup$ Also - is CDAA and CBAA (in your post) the same thing? $\endgroup$ – mike65535 May 16 '19 at 12:10

Response times are not primarily driven by the antenna characteristics, but the data sampling rate (electro-mechanical, ecectronic, or otherwise), data transmission time (including both receiver's tuning and bearing response) and lastly the ABI calculation algorithms.
So, beyond the fixed sampling rate, the biggest limitation is the confidence algorithm in the ABI calculation. What signal quality and how many "valid" samples are necessary to satisfy and complete the calculation. Obviously a marginal signal requires more time to satisfy the calculation. ABI = Automatic Bearing Instrumentation


being a circular antenna has very little to do with it. The Pusher is a circular antenna it is however smaller in diameter. Hams still use circular antennas for the lower bands BECAUSE they have excellent properties when it comes to DFing signals. This allows them to rotate around the globe to pick the strongest signal available. The resolution is often better in the newer antenna types however with the best DF cut being approximately +/- 2.5 degrees. What stops better resolution is the non-homogeneous nature of the ionosphere.

One further thing that did help to bring about the demise of the CDAA in the early to mid 90s is the drop in CW and other HF signals to obtain bearings on with the bulk of the targeted traffic moving to satellites by that time.

By the way these CDAA's and their successors work quite well on CW, SSB, AM, FM and even some of the newer modes even though the latter replacements might have been quite a bit faster when obtaining a bearing on a particular signal.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.