Review of the Battle of Dong Hoi

By Larry Rouse

To introduce myself and state my bona fides, I am a retired Senior Chief Electronics Warfare Technician, having served in the U.S. Navy from 1980 until 2000. During my early career, I was a technician and operator on the WLR-1C/G, the same system as installed on USS Sterett during the Dong Hoi engagement. I have heard the story of Sterett in the Dong Hoi battle told and retold many times throughout my career by numerous people who claimed to have been there. These stories, together with my own experience, have led me to develop my own theory as to what Sterett's EW crew saw and what actually happened.

First of all, the question of whether the Soviets had supplied the SS-N-2 to North Vietnam by 1972 cannot be adequately answered by consulting Soviet sources. The culture of secrecy, distrust and deceit was so pervasive within the Soviet Union that they routinely lied even to themselves about nearly everything. The Soviet archives are full of documents that are nothing more than one Bureau telling another the lie it thinks it wants to hear. In my opinion, no Soviet source can be trusted as accurate unless it is backed up by at least one independent source. That said, I personally believe that there were no SS-N-2s in Vietnam during this time because we have no other instance of them being seen, and this "sighting" is rather tenuous at best.

In his discussion of the EW capabilities of the USS Sterett, Mr. Slade cautions us to avoid mistaking the WLR-1 of the 1970's with the WLR-1H of 1980's vintage as they are "...completely different animals..." Unfortunately, he then proceeds to do just that. The console he describes as being used in this action is actually the console for the WLR-1H, not that of the WLR-1. In order to clear this up and perhaps add to the understanding of the conditions Sterett's EWs operated under, I will describe the intercept process below.

The WLR-1C/G was completely manually operated, there was an automatic scan function but it was mechanical, and was never used because it was slow and could knock the tuners out of alignment. There were nine bands available, which were selected by turning a selector knob on the console, and there were three directional and two omni-directional antennas. In order to intercept a signal, the operator had to select which band he wished to search, then select the appropriate antenna. Its rotation speed had to be adjusted slowly because having the speed control at maximum when the selector knob was turned could damage the servomotors. Then the operator would tune the receiver from one end of its band to the other. An intercept was indicated when audio was heard through the headphones or, more commonly, over a loudspeaker attached to the headphone jack on the IP-480. The operator would then read the frequency off the mechanical frequency readout to the EW Evaluator who would then log the intercept. The IP-480, which was the display of the WLR-1C/G, had two displays. The first was the Polar DF display which had a strobe which spun in sync with the rotation of the selected antenna, and the second was the panoramic display which showed four traces: a pulse as the top trace which was calibrated in microseconds to indicate pulse width, and a series of spikes on the middle two traces which gave estimated pulse repetition frequency (PRF) and the bottom trace which was the panoramic video display and was used to help identify the scan. It displayed a rapidly scanning spike that appeared like a spinning ring viewed end on. The scale of this video was adjustable in increments.

After reading off the frequency, estimated PRF and pulse width to the Evaluator, the operator would then stop the antenna's rotation and slew it onto the emitter, rocking it back and forth looking for the highest signal strength, in order to determine the precise bearing to the emitter. He would then tune an audio signal generator connected to the IP-480. This would cause the spike on the panoramic video trace to slow or even change directions. When the spike stopped, it meant that the output of the signal generator now matched the audio from the incoming signal. This represented the PRF of the signal and was read off a frequency counter attached to the signal generator.

During this time, the Evaluator would be comparing the parameters as he received them to various publications and locally prepared electronic orders of battle, looking for a match. Finally the operator would determine the scan type and time of the emitter by ear and stopwatch. Each scan type makes a unique sound and EWs were trained to recognize them. Bi-directional and circular scans were timed using a stopwatch, unidirectional and conical scans were measured using the signal generator and frequency counter. The scale of the panoramic video display would be turned up and the frequency counter adjusted until the display stopped moving. The frequency of the scan was read off the frequency counter. A uni-directional scan displayed as a row of humps, a conical looked like a tiara with one large hump and smaller ones on either side.

That sounds like a cumbersome process and it was, but a well-prepared and trained team could do it in less than twenty seconds. EWs drilled incessantly on threat recognition and most could identify many emitters, especially missile seekers, without the aid of publications. In a general quarters situation as many as six people would be involved in this process, which sped it up greatly.

Now to cases; the shoreline at Dong Hoi is not as described In Mr. Slade's essay. It is not semi-circular but is located in what can be described as a scallop in the coastline. This scallop is too shallow to have affected radar performance in any way. The bay itself has a narrow entrance flanked by a wide flat beach on either side. There are hills beyond Dong Hoi, but they are not semi-circular either and are too far from the scene of the action to significantly affect the radar picture. Mr. Slade goes on to describe the use of the missile illuminator as a surface search radar. I assume, in this case, he is referring to the SPG-55 radar. The SPG-55 was not and could not ever be used as a search radar. The Constant Boat Track (CBT) mode was used for the surface attack mode of the Terrier, and later SM-2 missile. The radar had to be queued onto the target using another radar, usually the SPS-10. It would then lock onto the target and track it continuously until the target was destroyed.

Mr. Slade is correct in stating that the conditions favorable to anomalous propagation exist about 80% of the time, but the actual effect is quite rare, especially in the South China Sea. These effects are only routinely reported in the Persian Gulf and Indian Ocean during certain times of the year; they simply do not happen that often anywhere else, and especially not in close proximity to land masses. The effect he refers to as "second time around" is actually called "second sweep echo" in the U.S. Navy. This is somewhat more common, but on an analog radar display, such contacts do not have the same visual quality as a real contact and are thus easily detected by experienced operators.

Because Mr. Slade does not understand how the WLR-1 worked, his assumptions about what the EW operators saw and how it affected their reactions are simply wrong. The Square Tie radar's scan, frequency and PRF are very different from those of the seeker on the SS-N-2. There is simply no way one could be mistaken for the other. The Square Tie had a circular scan in search mode and a steady scan in the track mode. The Styx had a conical scan. The sound it makes is unique and readily identifiable. A shift from search to track mode in the Square Tie is detectable as a change in PRF, and corresponding change in audio pitch. Even when it happened on friendly radars my heart would jump a little, but it was a cue for the operator to tune into the frequency range for the missile seeker, because the PRF change indicated that it could be coming next.

His statements regarding decoy launches by the NV patrol boats in response to being illuminated by the SPG-55 are plausible and I believe that these are a contributor to subsequent events. Chaff rockets are most useful against airborne radars because they generally burst high in the air, much too high to make them effective against surface based fire control radars. A chaff rocket launch would provide the vertical video separation that the crew on Sterett reported. But he states that the Vietnamese shut down their Square Tie Radars and this accounts for the emitter ceased reports. Again, the EW operators would probably not see the Square Tie cease because they were in a different part of the frequency band looking for the Styx missile seeker. They wouldn't know it was gone until they returned to look for it.

Finally, we come to the question of detonations of the Terrier missiles. Mr. Slade asserts that they detonated because they were unable to locate an airborne target and thus entered ballistic flight. This does not hold water as the missiles could not have been launched unless the SPG-55 was locked onto a target. The Terrier was not a "fire and forget" missile and as such it did not "look" for a target. The target had to be tracked from launch to destruction. The version of the Terrier used in this engagement used semi-active homing. This means that the missile received echoes coming from a target illuminated by the SPG-55 and it needs to receive these echoes from launch to intercept, otherwise it will not be able to guide to the target. In order to destroy a missile in flight, it is true that the operator would simply slew the radar off the target and the missile would enter ballistic flight. After a few seconds, the missile would self-destruct. However, the missile cannot be launched unless there is a valid, locked-on target.

That target could have been the chaff described earlier, but that does not account for the ESM intercept, which I believe was valid. I think that the target was a MiG-17P. This was the all weather version equipped with the Scan Odd radar, which was similar to the seeker for the SS-N-2. This makes all of the pieces fit together. A MiG-17P flying just above the overcast crosses the scene of the action just as the retreating patrol boats launch their decoys. The SPS-49 spots vertical separation and cues the SPG-55 to shift modes. The SPG-55 slews up looking for the new threat and locks onto the MiG-17P. The MiG pilot, hoping to strafe the U.S. ships without being blown out of the sky, turns on his radar to locate them, with the intention of popping out of the clouds on top of them. This is picked up on the WLR-1 and mistakenly identified as an SS-N-2. The SPG-55 locks on and a pair of Terriers are fired. These fly up into the overcast and the first missile strikes the MiG and destroys it. The second missile strikes the debris less than a second later. The "missile seeker" ceases and the U.S. ships retire from the area.

Given what was a confused situation, I also agree with Mr. Slade that the crews of all of the ships involved performed quite well. They managed to get all of the ships safely away and inflicted some damage on the enemy. The EW crew did an excellent job given the equipment they had. They evaluated the threat and carried through all phases of the detect to engage process, along with the likelihood of having successfully downed a MiG in the process.

Analogous Propagation and Radar Set AN/SPG 55B

By William W. Berry

For the record, I am a former Fire Control Technician who worked with the Terrier Missile Radar Set AN/SPG 55B (analog variants, mods 2, 4 and 5) prior to commissioning. I retired from the Navy as a LCDR LDO [Surface Ordnance, 6160]. I'm now a systems engineer working for a defense contractor.

This paper concludes that the so-called "second time around" radar echoes described in the article were unlikely.

USS STERETT (CG 31), at the time of the Battle of Don Hoi, was a "mod 4" ship. It was equipped with a Master Synchronizer AN/SPA-42A. This device ensured that neither AN/SPG-55B's track radar transmitter, nor the surface search radar AN/SPS-10, radiated in a manner to interfere with one another. The synchronizer told one radar to transmit, then the next, and so on. Another piece of equipment, the AN/SLA-10A, issued signals to radars that blanked their receivers during the short period of time that another radar was transmitting. Thus, the hugely powerful transmit klystrons of one AN/SPG-55B would not interfere with another radar nor with the AN/SPS-10, presuming each of these support equipments was operational. Each was a very reliable device requiring minimal maintenance.

The AN/SPG-55B had three methods of transmission: Long Pulse, Short Pulse and Narrow Pulse. Each had their own use for specific target tracks, and the radar operator could alternate between these to attain the best target track. Generally speaking, the long pulse operation was the best, because the receiver bandwidth could be narrowed (bandwidth is an inverse function of pulse width – the shorter the pulse, the more wide open the receiver has to be), thereby limiting the amount of extraneous noise that could sneak into the receivers. However, it was miserable to align the AN/SPG-55B receivers, very time consuming in a very hot space, and I don't have any idea about how good (or otherwise) the radars in STERETT were working that day.

The radar also had a feature that, if turned on, would limit the amount of second-time-around echoes. If this feature was turned on, a second-time-around echo would appear as a blur to an operator instead of a well-defined receive echo. Again, I have no idea as to whether this was turned on, but most AN/SPG-55B radar operators would have done so by virtue of their training and operational experience.

Consequently, I think the concept of second-time-around echoes would be credible only if a radar operator screwed up his normal operational settings, or if a multitude of radar casualties prevented each of these features (master synchronization, pulse width selection, etc.) from properly working. I consider this to be very low in probability.

Radars and Computers aboard USS Sterett and other Warships during the Vietnam War

By William W. Berry

The Terrier system that STERETT had aboard took its "mod" number from the type of computer used.

Radar Set AN/SPQ-5 was the original equipment in DLG 6, 7, 14 and 15, as well as CAG 1 and 2. It was a maintenance nightmare, with much of the sensitive electronics enclosed in the huge antenna. The antenna used three-axis tracking (train, elevation and traverse axes). Its antenna was the predecessor of the Talos AN/SPG 49.

Radar Set AN/SPG 55 was the original variant that only guided beam-riding Terrier (BT) missiles. It was installed in CGN 9, CV 63-64, DLG 8-13, and DLGN 25.

Radar Set AN/SPG 55A was an upgrade that allowed it to illuminate a target and guide Homing Terrier (HT) missiles as well as retaining BT missile capability. In addition to the above ships, it was original equipment in DLG 16, 17, 19, 21 and 22, and a couple of Italian cruisers.

Radar Set AN/SPG-55B was a major upgrade that included a very high powered 7-cavity klystron transmitter mounted below decks, which doubled the radar's range and tracking capability. This radar was extensively modified in coming years, requiring multiple OrdAlt plates. By the mid-1970's, all Terrier radars were some variant of this radar. It was original equipment in remaining U. S. Terrier ships DLG 26-34 and DLGN 35, and in one Italian cruiser. It could guide the BT, HT and the new Standard Missile 1 (SM-1). In later years, the BT and HT missile capabilities were dropped and Standard Missile 2 (SM-2) capability was added, as well as a co-located Doppler tracker that operated on an entirely different frequency than the track radar: a bad guy would have to jam two separate radar frequencies to defeat what had become a very capable radar.

My first encounter was with Radar Set AN/SPG 55B (mod 2 - see below) in USS WORDEN (DLG 18). It had 900 vacuum tubes, 1700 diodes and 456 transistors in each of the two radars installed. I suspect I replaced each one of these at least once. The radar was filled with brother-in-law circuits, for which someone was probably receiving usurious royalties. I can almost see someone in NAVORD looking at things and saying "golly, gee whiz" and writing the check.

My next encounter was USS LEAHY (DLG 16), which had just been upgraded to a Mod 5 system (see below). It was a little better, and I'm especially proud of a midnight OrdAlt I generated for each radar that provided the radar operator with a range rate (DMr) meter. That way, the radar operator could see at a glance that he was tracking a moving target. I repeated this midnight OrdAlt in USS ENGLAND (formerly DLG 22, then CG 22 at the stroke of someone's pen). Later digital variants provided this range rate monitoring capability on-screen.

Computer Mk 119 mod 0 was an analog machine that solved the fire control problem using angular rates (lead angle offsets from the line of sight). The mods 1, 2 and 5 variants were similar. The mod 4 variant was an expanded design that eliminated many of the dual-purpose servomechanisms and was therefore a tad easier to align, but computed the same ballistics.

In the late 1970's, the digital computer took over, with the Mk 152 computer replacing the Mk 119. Much later on, that computer was deemed obsolete and was replaced with a more modern variant.

The overall Terrier system originally took its name from the computer being used.


Mk 76 mod 0

Radar Set AN/SPG-55 (later upgraded to AN/SPG-55A): Mk 119 mod 0 analog computer.

Mk 76 mod 1

Same as Mod 0, but as installed in USS LONG BEACH (CGN 9), and the computer was a Mk 119 mod 1.

Mk 76 mod 2

Radar Set AN/SPG-55B, Mk 119 mod 2, installed in the first four AN/SPG 55B ships (DLG 18, 20, 23, 24). Incidentally, they had run out of money, so these ships initially received only one AN/SPG 55B radar at each end, but had a full complement of four computers (go figure).

Mk 76 mod 3

Same as mod 2, but installed only in CV 66.

Mk 76 mod 4

AN/SPG 55B, Mk 119 mod 4, installed in DLG 26-34 and DLGN 35. This is the variant that STERETT had installed aboard during the Battle of Dong Hoi.

Mk 76 mod 5

This was the upgrade to previous DLG's that changed them from AN/SPG-55A to AN/SPG-55B, and used the Mk 119 mod 5 computer. The four Mod 2 ships were also converted to the common configuration.

Mk 76 mods 6-10

Eventually all Terrier ships (with the exception of the aircraft carriers) went through a series of OrdAlts to improve their capabilities and replace the Mk 119 computers with digital variants.

The radar was extensively modified so that it looked almost the same, but the guts below decks were entirely different. With Mod 8, the Doppler tracker was introduced.

One statistic of which I am very proud: we religiously ran Daily System Operability Tests (DSOTs) in all different configurations. In USS ENGLAND (CG 22), 53% of our missile firings were cross-launcher assign (a forward radar to an aft launcher or vice versa). Other ships of our class were scared to death of this capability and never used it.

We knew it worked, well and often.

Page History

28 September 2005
Updated section "Review of the Battle of Dong Hoi" by Larry Rouse.
4 February 2006
Updated sections "Analogous Propagation and Radar Set AN/SPG 55B" and "Radars and Computers aboard USS Sterett and other Warships during the Vietnam War" by William W. Berry.