Radio Homing

of a SeaTag for Recovery at Sea

Using Radio Direction Finding techniques to recover SeaTag devices from the ocean

Staging an expedition to recover SeaTag devices floating potentially hundreds of miles from shore may not make sense for the value of the tags alone, but can yield a 'gold mine' of data.

Using micro-SD card technology, SeaTag-MOD can hold gigabytes of data and work such as detailed accelerometer based studies of animal behaviour can quickly produce data sets far too large for transmission by ARGOS in its raw (and most valuable) format.

With SeaTag-MOD, physical tag recovery looks to become a realistic option in at least some cases. The tag's solar powered ARGOS transmitter has proven to remain active for weeks and months - enough to stage an expedition or possibly even allow the tag to drift to an area close to shore. As of this writing, some tags deployed one year ago are still transmitting on solar power.

The latest version of this application note (30JUL12) includes a case study of a tag recovered in the Florida Keys after popping off from a Great Hammerhead Shark.

Finding a SeaTag in the Ocean

Updated 30JUL12 by Marco Flagg

Physical recovery of a SeaTag-MOD can be of tremendous scientific value. Due to the devices large memory, it is generally programmed to store a full set of sensor scans at a frequent rate; at least once every 4 minutes and as fast as once per second. While satellite data transmission is generally limited to summaries and snippets of the raw data, physical tag recovery yields the whole, uninterrupted data set.

Tag recovery of course can and does happen ‘accidentally’. One SeaTag-MOD was found recently by a beachcomber in Alaska. Incredibly, two others were retrieved by divers when two of six tagged tiger sharks returned to the same site three months after tagging.

But, due to the tags solar powered design, targeted recovery of after pop-up is now a realistic possibility. Unlike battery powered tags, SeaTag-MOD will generally transmit for months and in fact one of our early test tags floating in the Atlantic is now a couple of weeks short of reaching a year of reporting.

The long post pop-up reporting duration allows time to stage a recovery expedition, or to wait until the tag has drifted within reasonable reach of a port.

Since the first revision of this app-note a few months ago, the described technique has been used for two actual tag recoveries. First was a recovery of a tag by Hopkins Marine Labs in Hawaii; the tag had drifted up on a beach in Oahu and the equipment led the team right to it.

Then, just last week, I was part of a tag recovery team going after a SeaTag-MOD of the specialized ‘Hammertag’ variety belonging to Neil Hammerschlag of RSMAS (U Miami). It had popped off a Great Hammerhead shark about 20 nautical miles offshore in the Florida Keys. Coming up on that tiny tag floating straight ahead in an ocean spanning to the horizon in all directions was a tremendous rush!

The first part of this new revision of the app-note describes the same proposed recovery technique, although with a few minor changes based on what we learned. The second part is a case study of the successful recovery.

Equipment Needed
  • A handheld FM radio scanner, that meets these characteristics:

    • Can operate at 401.65MHz, the ARGOS transmit frequency

    • Has a signal strength (bar graph) display

    • Has an audio output jack

    • Has a removable antenna, typically with a BNC antenna connector

  • A Ramsey Electronics DDF-1 Doppler Direction Finder

    • A 1/8” male stereo to 3/32” male stereo cable to connect the audio output jack of the scanner to AUDIO IN on the DDF-1.

    • A suitable power cable for the DDF-1. It is designed for use with 12V power systems, and you can make a cable that connects to a cigarette lighter plug on a boat for example.

  • A Ramsey Electronics LPY-41 ‘Logi’ Log-Periodic directional antenna

    • A SMA male to BNC male cable to connect the antenna to the scanner antenna connector.

    • A non-metallic mounting pole for the LPY-41, and non-metallic mounting hardware. A 2” PVC pipe is a suitable mounting pole.

  • A SeaTag-MOD that can transmit for test and calibration purposes.

Suitability of Various Scanners
Radio Shack PRO-404No signal strength display.  Not suitable for use with directional antenna.Works well.  Correct audio signal into DDF-1 at about 1/2 volume on scanner and 1/2 AUDIO LEVEL on DDF-1.  PHASE NORM on DDF-1
Radio Shack PRO-107Works well with LPY-41.  5-bar signal strength display.Not suitable.  Audio output not sufficient for DDF-1. AUDIO LEVEL ‘LO’ LED generally stays on during transmissions, only flickering off – even with full scanner volume and AUDIO LEVEL knob on DDF-1 all the way to the right.

Icom IC-RX7

Not tested. High resolution signal strength display (about 15 segments) should make this work well to judge small changes in signal strength.

Not recommended. Switched antenna RF signal into scanner often does not trigger the squelch even at the most sensitive setting. Works with squelch open, but this does not reliably latch the direction indicator at the end of a transmission.


General Search Technique

  1. ARGOS Fix: Determine the approximate location of a SeaTag floating on the ocean by getting its most recent ARGOS position fix. Verify the fix is reliable, by seeing several in the same vicinity and/or making sure the position class is 0,1,2, or 3 but not A, B or Z. Make sure the tag shows a good transmit activity, as repeated signals will be needed for homing. Specifically, look at the ARGOS transmit serial number in two SDPT_MODENG packets transmitted some hours apart during good daylight. From that, compute the tags transmit rate in minutes per transmission. Determine if the transmit rate is sufficient for a realistic recovery effort, and use it to time the search.

  2. Long-distance approach guided by directional antenna: You use the directional antenna to guide your initial approach, because it will pick up a signal over a greater distance than the DDF-1 setup can. The rough change- over point may be at 1km distance, but can vary substantially depending primarily on directional and DDF-1 antenna height above the water. Mount the directional antenna as high as reasonably practical, pointing in the direction of travel. As you are within a number of kilometers of the tag, you should pick up its signal on the scanner. Watch the scanner’s signal strength display closely during each transmission closely and adjust the course to maximize signal strength. Another trick: We found that the typical 1-sec transmissions of the tag are long enough to get a rough idea of direction by quickly rotating the antenna once the 1-sec transmission starts, then watching the signal strength get weaker or stronger.

  3. Change over to DDF-1: Once you are close enough to get four or five bars with the directional antenna (standard five-bar display such as on PRO-107), it’s time to start watching the DDF-1. Even better, just keep both systems running simultaneously and use whichever performs better at any given time.

  4. Close approach: Use the heading indicator on the DDF-1 to adjust the boat course, keeping the display centered on straight forward. Course adjustments should be slow and deliberate; do not overcompensate. It’s important here to judge if a heading indication is good or bad:

    1. If the DDF-1 emits a very pure 500Hz tone along with the received tag signal, the signal is clear of interfering reflections and you can probably trust the heading indication (assuming good prior calibration).

    2. If the 500Hz tone sounds distorted or raspy, there is a problem with RF reflections and the heading indication may be inaccurate or completely wrong.

    3. Both of the AUDIO LEVEL LED must be off during a transmission, indicating the audio signal strength into the DDF-1 is at the correct level. If either LED remains on, the DDF-1 is locked and it will not try to determine a heading. Adjust the AUDIO LEVEL knob on the DDF-1 and the volume on the scanner until both LED turn off during a tag transmission.

    4. Consistent display on repeated transmissions suggests good heading, jumpy performance means uncertainty.

    5. The DDF-1 keeps the LED of the last heading indication on, even after the transmission ended. So, you can look at it as needed. Be a bit careful, however. In particular if squelch should be off and static comes through between transmissions (not a good idea), the heading display may flip to some incorrect reading. Know-your-gear…

  5. Visual ID and recovery: Both in preliminary tests at sea and the actual tag recovery described below, the tag ultimately appeared straight ahead. Use the signal strength display of your second scanner used with the directional antenna to judge the approximate distance. You might even wish to connect the standard vertical antenna back to the scanner. As you approach close, the signal strength will rapidly gain and you will soon see five bars (or equivalent full strength). Now it’s time to place a lookout at the bow to search for the tag. Go to minimum maneuvering speed, and at the most three knots. You may see the tag straight ahead, perhaps at a visual ID distance of 50m or so. Alternatively, your boat may pass by the tag in close proximity. In this case, you will see the heading indicator move in a left or right arc to behind you. A heading indication of 90 degrees to port or starboard indicates that you have reached the closest point of approach. Mark it as a GPS waypoint. Circle your boat in whatever heading the DDF-1 indicates, and look in that direction for the tag. Upon successful recovery of a tag, dance the tag recovery jig and set course for the next one!

Setup for Field Testing

We set up the SeaTag-MOD on a small tower on the roof of our building. That is of course a much better antenna position than a tag floating in the ocean, and so we probably got significantly greater coverage than you will see at sea. To maximize your coverage by setting up the antennas high on your boat. They should also be as clear as possible from rigging and metal boat structures that will cause RF reflections.

The DDF-1 antenna array was on the roof of our pickup truck. The magnets delivered by Ramsey are weak, and so we made generous use of duct tape to back them up. The antennas are spaced about 9.4” apart (4.7” radius) as described in the DDF-1 manual for 400Mhz. The telescoping antennas are extended to a length of 7”. The sequence is antenna #1, #2, #3, #4 clockwise. We placed antenna #1 facing forward, but that doesn’t really matter because you need to calibrate the heading reading to align with the reality of the situation before first use in any case. That is done with the CALIBRATE knob on the DDF-1.

Here is a close-up of the LPY-41 directional antenna secured to a 2” diameter PVC pipe. The pointy end is the direction of the strongest signal; as in this example pointing towards the truck.

Antenna view from the other side. We mounted by drilling four holes through the plastic parts of the antenna only (avoiding the copper traces), then using zip-ties. Zip-ties aren’t the greatest to stabilize the antenna, and so we used some duct tape to prevent things from rotating in the wind.

The DDF-1 inside the car. There are three connections: (1) 12V DC power, (2) INTERFACE connector with direct connections to the antenna array and a coax-cable with a BNC connector going to the antenna connector of the scanner, (3) AUDIO IN from the speaker jack of the scanner. The green LED indicates power ON, and one of the heading LED is lit indicating the transmitter heading measured during the last transmission. The AUDIO LEVEL LO light is ON. This will be the case between transmissions. If this or the HI LED is on during a transmission, you must adjust the AUDIO LEVEL knob and/or the scanner volume: The signal into the DDF-1 is too weak (LO) or too strong (HI), and the device is in lock-out mode where it will not attempt to get a heading.

Calibrating the DDF-1

The DDF-1 direction finder must be calibrated before use.

  • DDF-1 Switch and knob settings:

    • POWER switch ON

    • SCAN switch in RUN position. RUN enables direction finding, and so should always be selected. STOP is available to listen to transmissions without the 500Hz Doppler effect tone generated by the direction finder. That setting is rarely necessary when working with tags.

    • Start with PHASE switch in NORM position. This setting depends on the wiring of your scanner and will always be used in one or the other setting (see below).

    • CALIBRATE, DAMPING and AUDIO LEVEL knobs all in center position.

  • Scanner and tag:

    • Scanner set to 401.65MHZ, the SeaTag-MOD transmit frequency.

    • Volume about half way.

    • Squelch just so engaged (no static)

    • Configure your tag in real-time transmit mode, sending for example sensor packets. Configure for frequent transmission; once every 16 seconds. Ideally, attach a battery pack as well, so that both the sun and battery will re-charge the tag’s capacitor, promoting frequent transmissions for this calibration purpose.

  • Calibration site selection:

    • The ideal site will have no buildings or hills nearby, as both can cause reflections. A flat field or meadow in the boonies is ideal. If you can’t find that, try a large parking lot with buildings small or relatively distant.

    • The calibration distance could be anywhere from 50m to 1000m. Make sure you are getting a strong signal from the tag.

  • AUDIO LEVEL adjustment (DDF-1):

    • When no transmission is taking place, the LO LED should be lit.

    • During a tag transmission, the LO LED must go off (not just flicker briefly), while the HI LED must stay off. This indicates that the signal going to the direction finder electronics is at the correct level needed for heding detection.

      • If the LO LED stays on during tag transmissions, turn the AUDIO LEVEL knob to the right. You may also have to adjust the scanner volume up. If full scanner volume and full AUDIO LEVEL setting is still not enough, then the scanner is not suitable for DDF-1 operation. We have seen that with the Radio Shack PRO-107. Use a different scanner, or possibly a small audio amplifier. (We have seen good results with the Radio Shack PRO-404 scanner).

      • If the HI LED goes on during transmissions, the audio signal is too strong. Turn the AUDIO LEVEL knob to the left, or turn down the scanner volume.

    • Note that turning the AUDIO LEVEL knob will not change the volume of the sound coming out of the DDF-1 speaker. It only adjusts the signal level going to the direction finding electronics.

    • Note that incorrect level adjustment (a LED is ON during transmissions) will completely prevent you from seeing heading indications. The DDF-1 is in lock-out mode when this is the case.

  • PHASE switch selection:

    • With the DDF-1 antenna array mounted on the roof of a vehicle, drive the vehicle slowly in a circle and watch the heading indicator. If you are turning left, the heading indicator should turn right (like a compass would), and vice versa. If the indicator moves in the same direction of your turn, then change the PHASE setting. This setting is scanner specific, and will always stay the same once you determined the correct setting. We found that NORM is correct for the Radio Shack PRO-404 scanner and the Icom IC-RX7.

    • For this test, do not worry which way the heading sensor actually points, just that it moves in the right direction.

  • Heading indicator calibration:

    • Heading calibration must be done after each installation on a boat or a car, as it adjusts the indicated heading to correspond to the reality. This depends for example on the orientation of the antenna array, but also on particulars of the scanner and DDF-1 electronics.

    • Throughout the heading calibration, listen for the 500Hz Doppler tone when the tag transmits. It must be pure. If it sounds distorted or raspy, you are picking up RF reflections and the calibration will not be reliable. Move your vehicle to see if the situation changes. If you are seeing RF reflections no matter where you go, then they may well originate from your vehicle – such as reflections from the rigging of a boat. In that case, your only option is to move the DDF-1 antenna array to a different place and try again.

    • Point the car or boat in the direction of the tag and watch the heading indicator during a tag transmission. If it is to the left, then turn the CALIBRATE knob to the right. If the indicator points right, then move the CALIBRATE knob to the left. Repeat until the heading indication is perfectly centered at 0 degrees when the car or boat points at the tag.

    • Verify your calibration by turning your monster truck or boat in various directions, and watching the heading indicator. During and after each transmission, it should now point in the actual direction of the tag!

  • Adjusting the DAMPING knob:

    • The DAMPING knob determines how fast the DDF-1 responds to a signal. Left is very fast response but a bit flicker. Right is slower response (more averaging), and therefore more steady. We found this knob to be quite uncritical, and the center position works fine.


Test Results
10-50m: Parked and driving around the building. Not useful.  Too close, and maximum signal strength indication in all directions.

Mostly accurate & solid, but with wrong results due to RF reflections in some spots. RF reflection was obvious when it happened: The 500Hz tone was distorted or raspy.

120m: Driving a circle around a parking lot.

Always required an attenuator switched into the antenna cable to reduce signal strength. Even then, much problems with reflections. Not useful.

Similar to driving around the building: Generally accurate, but with some spots impacted by RF reflections.

920m: Driving around a traffic circle. Stopping with truck facing various directions.

Only useful with attenuator. With 20dB attenuation, 3-4 bars when pointing at tag, 1 bar in opposite direction, 1-3 bars in other directions at least 45 degrees off.

Solid performance; no reflection problems.

2.78km: Stopped on raised terrain, about 200’above tag elevation.

Generally useful. 5 bars when pointing at tag. 3-4 bars in other directions at least 45 degrees off.

Solid performance; no reflection problems.

14.3km: Stopped on mountain top, about 1100’ above tag elevation

Generally useful. 3 bars when pointing attag. No detectable signal when 90 degrees or more off.

Sporadic signal reception. When received, it was accurate.

4-0km: Driving towards Desert Star from Fort Ord / CSUMB.

Not tested

First sporadic signals at CSUMB library, 4km distant. Stronger and more frequent as approaching Desert Star. Some false readings initially, but even then more accurate than not.

920m site: The tag is near the control tower in the distance. At this distance, the use of the directional antenna still required an attenuator due to a strong signal, and the DDF-1 worked solidly.

2.78km site: Both the directional antenna, and the DDF-1 worked well. When both systems are available, the DDF-1 will generally be the preferred choice because its readings are instantaneous and more accurate.

Is a directional antenna setup even needed?

Our tests showed very good direction finding with the DDF-1. However, it’s range will always be less than that of the directional antenna. The issue here however is that RF conditions and effective range can be very variable, in particular when operating from smaller vessels with the antenna not much above the sea surface or when the sea state frequently obscures the line of sight to the tag. In addition, ARGOS position fixes frequently have errors of several kilometers. If the fixes are old, the tag may also have drifted. Under any given set of conditions, a directional antenna can be expected to pick up the signal from a tag from a substantially greater distance than the omni-directional antennas on the DDF-1. This edge may enable you to find an initial direction to the tag once you have arrived at the last reliable ARGOS position.

Being prepared to do some directional antenna work is recommended to maximize your chance of success.

A Tag Recovery Case Study

This section discusses the recovery of a floating SeaTag-MOD at a location about 15 nautical miles north of the island of Marathon in the Florida Keys. The discussion focuses on the chain of decisions that resulted in success.

Preparations: Boat, weather and tag transmit rate

Weather: On the planned day of the tag recovery, weather conditions were favorable. It was mostly sunny, meaning solar power tag transmissions would be maximized. Further, the sea was near dead calm. This would mean that line of sight from the tag’s antenna just above the water to the boat mounted receiving equipment should generally be available, i.e. reception would not be lost when the tag transmit while in the through between two waves. Locating the tag visually once in close proximity should also be easier.

Tag transmit interval: ARGOS transmission serial numbers embedded in two engineering packet transmissions (SDPT_MODENG) from the previous day showed that the tag was transmitting about once every two minutes. The tag’s battery would by now be depleted, so the transmit rate depended on sunlight and any fouling of the tag. Once every two minutes indicated good performance. A two minute interval should provide enough fixes to recover the tag with reasonable expediency.

DDF-1 Calibration: Once out on the bay, we tossed a test SeaTag-MOD into the water. It was connected by some monofilament line to a small buoy for ease of viewing and recovery. At a stand-off distance of about 100 yards, we calibrated the DDF-1 so that the indicated direction matched the actual direction of the tag. The test was repeated with various boat orientations relative to the tag. Indications were correct, except in one direction (starboard aft), where possible RF reflection from the boat caused unreliable direction indication. Since the unreliable direction was behind the boat, it shouldn’t be a problem.

Transit to first ARGOS position: We had obtained a few ARGOS positions that morning, and selected a class 0 position as our search start location. By the time we arrived at the position, the fix was about four hours old. During the transit, we pointed the directional antenna straight ahead, but did not hear a transmission.

Initial search with directional antenna (4.5 nautical miles from tag recovery location): A few minutes after arriving at location, we received a weak signal (zero bars) on the directional antenna system, while it was still pointed in a random direction (search had not started yet). So, we knew that the tag was functional in the same general location. Neil Hammerschlag at RSMAS was also texting updated ARGOS positions as they became available. Meanwhile, Cpt. Bill Hardy was monitoring the currents through the drift of the boat. After perhaps 15 minutes, and with only a partial or semi- random directional antenna search completed, a new ‘class B’ position roughly coincided with the direction from which the signal seemed to be strongest, and also the general direction of the currents. We started off in that location, speed 4.5 knots.

Full search with directional antenna (about 3.5 nautical miles from tag recovery location): After signal strength reached one bar on the directional system (no signal on DDF-1), we went to minimum maneuvering speed and steady heading to conduct a full search. The antenna was pointed in 90 degrees intervals, listening in each direction until at least one transmission or received, or the time elapsed was long enough so that a transmission should have been received (about three minutes for two minute transmit interval). Straight ahead and 90 degrees port showed one bar; other directions zero bars or no signal detected. We checked 45 degrees to port, and the second bar flickered. This became our selected course. The signal gradually became stronger, a solid two bars, then three bars.

First signal on DDF-1 (about 2.7 nautical miles from tag recovery position): By the time the directional antenna setup showed three to four bars, the DDF-1 for the first time indicated a position. It was dead ahead, in conformance with the earlier established course. We slowed to three knots to travel less between fixes and avoid overshooting the tag. Monitoring each transmission, we made gradual compensating course adjustments (5 to 10 degrees to starboard or port) if the indication of the DDF-1 was one LED off from dead ahead. A lookout was now placed.

Final approach and visual detection (2.7 nautical miles to 50 meters): Closing proximity to the tag is apparent because required course corrections will become larger. The directional antenna scanner also showed a solid five bars. Course corrections grew, finally becoming 60 degrees. We marked point of closest approach on the GPS after the DDF-1 indicated a required heading change of over 90 degrees. After setting the new course, the tag was quickly spotted dead ahead, at a distance of about 50 meters.

The tag recovery location was about 4.5 nautical miles from the ARGOS based search start point. The time from start of search to tag recovery was 2.5 hours.

Summary of lessons learned from successful recovery

  • Wait for good weather conditions. Sunshine is needed for frequent transmissions. Calm seas will allow uninterrupted radio line-of-sight to the tag and make it easier to spot the tag.

  • Select a boat that allows high antenna mounting, at least 5m recommended. The table below relates antenna height above water to theoretical range in calm seas, based on earth curvature.

  • Measure the tag transmit rate in advance (minutes per transmission) by evaluating the ARGOS transmission serial numbers as reported in two recent engineering packets (SDPT_MODENG). It should be fast enough for recovery. Once on-site, we had a transmit interval of about 1mins 45sec. Anything over 5mins would have been very slow and challenging.

  • Obtain a recent ARGOS position as your search start point, one that is probably reliable on account of continuity with previous reported positions.

  • Use the directional antenna setup for a best chance to obtain an initial direction (greatest range) once on site.

  • Once you have at least one bar with the directional antenna (and no DDF-1 signal yet), take the time to test in all directions, starting at 90 degrees intervals and down-selecting to 45 degrees. Keep boat course steady at minimum maneuvering speed while doing this evaluation. This evaluation will increase the confidence of your search heading.

  • DDF-1 operation:

    • Calibrate prior to search. Watch out for RF reflection problems, indicated by completely wrong bearing indication in some directions, generally accompanied by a distorted sound 500Hz Doppler tone. Keep antenna array away from rigging and metal surfaces as much as possible.

    • Bearing indicator generally good to within 1 LED (22.5 degrees). Make small course adjustment to keep the indicator showing dead ahead, but do not over-compensate.

    • Be aware that bearing indication may change just as the signal cuts out (DDF-1 gets confused by brief noise until squelch kicks in). Use the reading indicated while the tag signal is received.

  • Search roles or task division:

    • Search leader, considers all inputs (directional distance reporting, DDF-1 indication, boat drift and currents) and sets search strategy.

    • Helmsman, with good view of DDF-1. Follows DDF-1 heading or other course, as determined by search leader.

    • Directional scanner operator. At highest point (observation tower), searches for signal.

    • Look-out. For final approach and tag recovery.