Hyperbolic System operating principles

HiFix was a short range precision positioning system primarily for marine
hydrographic survey purposes, capable of providing a fix with an accuracy of
about 1.5m at 30 miles from the shore stations. HiFix was a hyperbolic system
using a chain of three transmitters, one as master and two as slaves. As with
most such systems, for example Decca, the operating principle was that the
transmitters radiated a carrier phase locked to an accurate source, and the
receiver measured the difference in phase between the signals received from the
three transmitters. Unlike Decca however, where the transmitters radiated a
continuous transmission on different but harmonically related frequencies,
HiFix used the time-share principle where the transmitters radiated on the same
frequency, but in sequence. This is explained further in the 'System Description'
section.

Hyperbolic system prnciples

The principle on which all hyperbolic navigation and positioning systems
operate is essentially the same. If two transmitters radiating a radio wave in the
same phase are located at the ends of a baseline, as shown in the diagram below,
then a receiver in the centre of the baseline will receive the wave in the same
phase since the time-of-flight of the wave to the receiver from both transmitters
is the same. Applying this principle in reverse, if the receiver is receiving the
two waves in phase, then it must be located either at the centre of the baseline or
somewhere along the perpendicular line l - l'.

However, in practice the transmitters are further apart than a single wavelength
(which is only about 160 metres in the case of HiFix on 1900 kHz) so
consequently there will be several points along the baseline where the signal is
received in phase. Unlike the simple case outlined above however, when plotted
out from the baseline, these points of equal phase will produce a family of
curves instead of a straight line. These curves are referred to as hyperbolae, and
a set of hyperbolae generated by the master and one slave are referred to as a
pattern. If, at the receiver, we are receiving signals with a particular phase
relationship then we know that we are somewhere along one of these curves.

In order to give a navigational 'fix'- i.e. to find out exactly where on the curve
we are, we need another pair of transmitters and to carry out the phase
measurement process on the pattern produced by these as well. The hyperbolae
from this second pair will overlap those from the first, producing a 'lattice': If
we can establish our location along a hyperbola from each pattern, then we have
found our position. In practice, one of the transmitters can be common to both
pairs- with HiFix, this is the Master or A station; those at the ends of the two
baselines are the B and C slaves. The diagram below shows a section of
coastline with the 3 sites and their overlapping pattern of hyperbolae. It must be
remembered that the fix was always in relation to the position of the chain
transmitters, and these had therefore to be accurately known.

Resolution
Points of zero phase difference are referred to as 'lanes'. It is not necessary to
rely on these however since the receiver can measure phase differences at points
in between lanes. The maximum resolution of the system is expressed as the
smallest distance within a lane that can be measured and is limited by a number
of factors. In the case of HiFix the maximum resolution of the system is given
as 0.01 lane, which at an operating frequency of 1900kHz would be 1.6 metres.
Since the distance between hyperbolae and hence lanes increases with distance
away from the transmitters, the accuracy of the system reduces in proportion.
The operator would have to take this into account when a fix is obtained.

Lane ambiguity
It may have become apparent to you that there are actually several points where
the same phase relationships can be measured from both the A-B and A-C
sectors of the chain. This is true and the result is that a fix is not unique. This is
a characteristic of hyperbolic systems, and is referred to as 'lane ambiguity'.
Different radio location systems tackle the problem in different ways. Decca,
for instance, periodically broke the normal transmission format and transmitted
a signal that enabled the receiver to get a rough fix- to within a lane. On
resumption of the normal transmission, phase measurements were taken and the
exact postion within the lane displayed.

HiFix in its simplest form had no inherent capability to resolve lane ambiguity.
The receiver did, however, have a 'lane counter' display. Therefore, a navigator
using HiFix always had to start from a known position- perhaps by taking
bearings on visual objects or radio beacons on the shore. As the vessel moved,
lane counters incremented every time a complete lane was crossed. Users would
still have to periodically mark a position with a buoy. Ultimately this was a
major drawback, and a system using two radio frequencies was developed
(mode B) which could be used to give lane resolution. This used the fact that at
different RF frequencies, the lane pattern shifts. Subtracting one pattern (the
higher frequency) from the other results in a coarse lane pattern that allows the
user to fix their position to within half a lane.

Range-Range operation
An alternative to hyperbolic mode was Range-Range mode. This involved
carrying the master transmitter actually aboard the vessel, whilst the slaves
remained ashore. The receiver, also aboard the vessel, would therefore be at
zero distance (and therfore phase difference) from the master transmitter. With
this arrangement, instead of hyperbolae, a set of concentric rings around each
slave were obtained, and positional fixes could be obtained where the rings
overlapped. The advantage of this approach was that the spacing of the rings
stays the same as the vessel moves away from the shore, so the resolution may
be higher at greater ranges but the angles at which they intercept become less
favourable. Nevertheless under particular conditions, an improvement in
accuracy over hyperbolic mode could be obtained.

Errors
Several factors influenced the accuracy obtainable under practical conditions.
The first of these was the siting of the master and slave stations. For best results
they would ideally have been located with the baselines at an angle to each
other. They also needed to be located right on the coast, since radio waves
refract as they cross the coastline. Also, the accuracy of the system would have
been affected by radio propagation conditions; in particular after dark when the
signals become reflected by the ionosphere and therefore interfence from the
sky wave would have occurred.

Return to introduction page

System Description
The Hi-Fix system was engineered to fulfil a number of roles, varying form temporary
deployments to fixed chains. To enable this, the equipment comprised a series of units
each built on a 19-inch chassis 4U high (usually) and housed in a light alloy military-style
Creeth transit case. The cases could be fastened together to form a rack, or be free
standing.
 The equipment used both miniature valves and transistors in its circuitry. The valves
were of the type popularly referred to as 'pencil valves', on a miniature wire-ended 8-
pin base. The transistors were a mix of silicon and genrmanium devices. Within each
unit, a transistor invertor supplied the valve HT, and most of the other voltages
required. The DC supply to all units was 24V positive earth, intended to be supplied
from lead-acid batteries. A tap on the battery at 12V also required by the receiver.

The major units of the system were:

Decca type
Item Comments
no.

Master Drive Unit

9216 Used on master station
(MDU)

Low Power Transmitter 9371 10W output; Used on master and slave chain stations.

Medium power
9362 40W output; used on long range and permanent sites
transmitter

Receiver 9217/III Used for both user receiver and slave stations.

Blower Unit 9217/BM Cooling blower. Provides base stand for receiver.

Distribution Unit (DU) 9218 Provides metering and control of power supplies.

Blower and Synchro Provides power for synchros in receiver for remote displays
9369
Supply Unit and plotter; incorporates cooling blower.

Track Plotter 9232 Provides continuous track plot

Click on the links within this table for more information..

Minor units/accessories:
 Receiver antenna and matching unit
 Transmitter Junction Box (TJB)

 Transmitter Antenna Capacitor Unit

 24V Batteries

 Mains Battery Charger

 Generator charging set

HiFix Chain
The HiFix system required 3 transmitter stations. One of these, the Master, was located
centrally between two slaves, and at the intersection of the baseline from Master to
Slave 1, and Master and Slave 2. The sites transmitted in sequence on the same RF
frequency, as shown in the diagram below:

HiFix Transmission Format

Click here to listen to simulated HiFix transmissions!

The main frequency (referred to as the 'pattern frequency') could be in the range 1700 to
2000 kHz, but any one chain and its associated user receiver would be set up to operate
on one of 4 channels within a 25 kHz switching bandwidth. The three sites all radiated a
200mS pulse within a 300mS 'slot', which allows for a 'guard time' to allow for the
operation of the various switching circuits. In addition to the main pulse on the pattern
frequency, the master station also transmitted a trigger pulse on the 'signalling
frequency'- a frequency 60Hz lower than the nominal pattern frequency.

Master Station
The master station consisted of a Master Drive Unit (MDU) and a Transmitter unit. The
MDU contained a highly stable crystal oscillator from which all frequencies used in the
chain are ultimately derived. The MDU provided RF drive to the transmitter unit, which
contained an RF power amplifier.

Slave Station
The slave station consisted of a receiver and a transmitter. The receiver picked up the
transmission from the master station, and kept an oscillator in phase lock with it. It also
used the trigger pulse to synchronise a timer. At the appropriate time, the slave's
transmitter was keyed and radiated the slave pulse.

User Receiver
The same receiver unit could be used either on board ship or vehicle mobile. It picked up
the transmissions from the chain transmitters, and gave a continuos readout of position
relative to the transmitter sites on decimal digital counters. As an option, the receiver
could be connected to a chart recorder to give a historical trace of location.

HyFix Type B
The system described above is referred to as Type A, or Lane Integration working. This
type cannot resolve lane ambiguity and so it is necessary to know where you are are
starting from, and then the HiFix receiver will count the lanes as they are crossed, hence
'integration'. Loss of signal (which can happen due to lightning noise, for example)
means that you must needs back track to your last known position- a marker bouy for
example.

HiFix type B, or Lane Identification working, uses a second set of transmitters working on
a different RF frequency, but located at the same Master and Slave sites. This allows the
lane to be positively identified. The additional RF frequency needs to be about 10% lower
than the main one. This is how it works

Consider an area where, on the main channel of 1900kHz, there are 10 lanes. At the
lower frequency, 1710kHz, there will be only 9 lanes (because of the longer wavelength).
So when the lane counter on the HF receiver reads 1.0, that on the LF one will read 0.9,
and so on. If we subtract the LF reading from the HF reading, we find that the difference
is always one tenth of the HF reading. If we move the decimal point one place to the
right, we have the full lane reading of the HF pattern. Here is an example: HF lanes: 4.50,
LF lanes, 4.05. Difference is 0.45. Moving the DP one place to the right results in 4.5.

In this way we can resolve the ambiguity to 10 lanes- we don't know whether the lane
number in the above example is 4.5, 14.5, 24.5 so we have to be able to determine this
by other means. However, we are still 10 times better off than with Type A!

Return to introduction page

Decca HiFix Low-Power Transmitter Unit

The transmitter unit accepts drive from the master drive unit (MDU) at a master
station, and from the receiver unit at a slave station. It boosts this up to
sufficient level to feed to a transmitting antenna. It is constructed on a 4U high
19 inch rack chassis, and operates from a 24V DC supply with a maximum
consumption of 3 amps. The simplest of the HiFix system units, the RF stages
use valves, while the HT is derived from a transistor invertor which produces:
one HT rail at 215 to 265 volts for the driver stage and the PA screen grid
supply; One HT rail at between 480 and 530 volts for the PA anode supply; One
Bias rail at -50V.

RF Stages
The drive from the associated MDU or reciver is a keyed RF signal of 1.5-2
volts into 100 ohms balanced. The drive input is fed to the grid of an EL91
pentode which is biassed so that it draws a very small anode current under no
drive conditions, i.e in class B. The anode of this stage is coupled via a
broadband RF transformer to the grid of the Power Amplifier (PA) stage. The
PA is a 5B-258M tetrode biassed beyond cutoff (the full -50V bias is applied to
its grid). Therefore in the absence of drive, no anode current whatsoever flows.
The anode is coupled to the output socket via a tuning unit giving anode tuning,
coupling, and aerial tuning controls. In the PA anode circuit there is a 'feedback'
coupling transformer connected to a FB terminal. this is used for Type B
operation.

HiFix Type B operation

Under these circumstances each transmitter station uses two transmitters, each
working one of two RF frequencies. This is done to enable lane identification by
the user receiver.

In order to couple the two transmitters to a single aerial, their RF outputs are
paralleled while the FB (feedback) connection is cross-connected from the
output of one Tx to the FB terminal of the other. The point of this is that the
output of the other transmitter is fed back out-of-phase, this causes it to cancel
out in the PA anode circuit. In this way, the other transmitter becomes
'invisible'.

Aerial Detuning
There was considered to be a psossibility that the resonant antenna of a station
could possibly pick up and re-radiate the trnsmission from another station,
causing distortion in the pattern and therefore positional errors. To eliminate
this, the aerial tuning inductance in the transmitter is shorted out by a relay
contact when the transmitter is not actually keyed. This destroys the resonance
of the aerial making it less likely to re-radiate signals.

Back to system description

Decca HiFix Master Drive Unit

The Master Drive Unit (MDU) is a 4U-high 19 inch rack mount unit. Its function is to
generate all of the frequencies required in a HiFix chain. It generates the pattern
frequency and the signalling frequency, both of which are amplified to a suitable level
(1.5-2 volts RF into 100 ohms) for driving the transmitter unit. It also generates the
master timing sequence, and finally it can be used as a signal generator for setting up
other components of the system. To perform these functions it contains three crystal
oscillators and a ring-of-four timing circuit. It also has a built-in power supply unit that
draws 1.5 amps from a 24-volt, positive earth supply.

Frequency generation
The output frequencies of the MDU are derived by mixing. Two crystal oscillators, the
pattern oscillator produces 132.840 kHz and the signalling oscillator produces 132.900
kHz. Either of these can be mixed with the output of the heterodyne oscillator, which
generates a frequency in the range 1833-2133kHz. The difference frequency forms the
output of the MDU. The heterodyne oscillator uses one of four switched crystals, thus
enabling operation on one of four RF channels. The unit can be aligned on a 25 kHz
segment of the band 1700-2000kHz. The four channels must be within this 25kHz sub-
band. The various crystals within the MDU are contained in a crystal oven. This si
maintained a temperature of 70°C by an electronic feedback control circuit using
transistors.

Timing generation
A ring-of-four oscillator provides outputs that key the signalling and pattern frequencies
in the correct sequence. Time constants in the oscillator give a 100mS on period for the
signalling (trigger) pulse and the following 300mS main pattern pulse. During the
signalling period, the signalling oscillator is selected to mix with the heterodyne
oscillator, producing output 60Hz below the pattern frequency.

Components
The MDU uses 10 valves (Mullard EF730, EF731 and EF734) and 16 transistors.

Back

HiFix Receiver

The HiFix receiver can perform two roles.

As a user (field) receiver

The unit receives the RF signal from the master and slave transmitters and
displays the phase difference between the master and slave 1, and between the
master and slave 2. Using the trigger pulse radiated by the master station on the
siganalling frequency, it synchronises a timer to run in step with that in the
master drive unit at the master station. During the master pattern frequency
pulse, a crystal oscillator is adjusted to maintain phase lock with the master
station. During the slave transmission periods, the phase of the slave
transmissions is compared with the oscillator and the difference is displayed to
the operator on two mechanical digital lane counters, each displaying 5 digits: 3
digits before the decimal point which represent up to 999 whole lanes, and two
digits after the decimal point which represent fractions of a lane down to 0.01
lane.

As part of a slave station

The unit receives the signal from the master station, and provides drive to the
slave transmitter. Using the trigger pulse radiated by the master station on the
siganalling frequency, it synchronises a timer to run in step with that in the
master drive unit at the master station. In this way, the slave pulses are radiated
at the correct time. It also receives the master pattern pulse and during this
period a crystal oscillator is adjusted to maintain phase lock with the master
station. The oscillator is used to generate the RF drive for the slave transmitter.
This drive signal can be adjusted in phase to give the correct transmission phase
for the slave station.
General description
The unit is a superheterodyne receiver employing miniature valves and both silicon and germanium
transistors and diodes. The power input is 24V DC positive earth, and the HT supply and bias
voltages for the valves are derived from a transistor invertor. In general, the RF signal stages and
oscillators use valves while transistors are used in the timing and servo circuitry. The unit is
constructed in a 4U high 19 inch rack and extensive use is made of plug-in modular circuitry. A
crystal oven is used to maintain the oscillator crystals at a constant 70°C.

The antenna input is fed to an RF amplifier, which is followed by a mixer stage. A local oscillator
consists of a crystal oscillator using one of four channel crystals, followed by bandpass amplifier
stages. The IF signal from the mixer is at 132.840 kHz for the pattern frequency, and 132.900 kHz for
the signalling frequency. Crystal IF filters are used to give an IF bandwidth of 100 Hz, with a response
that encompasses both pattern and signalling channels. The 'composite' IF signal is amplified by
three IF amplifier stages.

Signalling frequency channel

The signalling frequency is extracted from the composite IF signal and fed, via further amplification
and a crystal filter to an envelope detector stage. This produces an output pulse whenever the
trigger pulse from the master station is radiated. This pulse synchronises a ring-of-four oscillator
which produces the timing signals required.

This oscillator consists of 4 stages connected in a ring. At each stage is a timing circuit, and each
stage produces an output signal from a relay. The first stage time constant is 100mS and all others
are 300mS- so essentially, the outputs correspond to the trigger and 3 pattern pulses.

Pattern frequency channel

The pattern frequency is likewise extracted from the composite IF signal and amplified further
before application to one input of each of 3 identical discriminators. The first of these forms a phase
locked loop that maintains a 132.840 kHz crystal in phase lock with the master transmission. The
other two disrciminators are used together with goniometers as part of the display, one for pattern
1 (master - slave 1) and the other for pattern 2 (master - slave 2).

To properly understand the following description of how the HiFix receiver measures and displays
phase differences, it is necessary to understand the operation of two essential components: the
discriminator, and the goniometer.

Discriminator
In modern electronics we would undoubtedly call this stage a phase comparator. Essentially, it has
two inputs and one output. When the two input signals are in phase, the output of the discriminator
will be zero volts. When the two inputs are of different phase, the output will be a DC voltage whose
magnitude depends on the amount of phase difference, and whose polarity will depend on the
direction of phase difference.

Goniometer
The goniometer is a device that can produce any desired phase shift with respect to an input
voltage. Since it may not be apparent how this can be acheived, the description below expands on
this.
 In this diagram are a fixed coil A-B and an additional
coil C-D which may be rotated about its centre with
respect to the fixed coil. In the situation shown, if an RF
voltage is applied across A-B then a current will flow in
the coil, and by transformer action, a current will be
induced in the rotatable coil C-D. The voltage produced
across C-D will be in the same phase as the the voltage
across A-B.
In this diagram, the moving coil has been rotated
through 180 degrees. The voltage produced across C-D
will now be 180 degrees out of phase with the voltage
across A-B. By rotating the moving coil to other
positions, corresponding changes in phase will result.

There is however a flaw in this theory, which is that

when the two coils are at 90 degrees to each other, no
magnetic coupling occurs and hence there will be no
output when the moving coil is set as shown here.
To overcome this, an extra fixed winding is provided
which is connected to the same voltage A-B but through
a capacitor C1. This component is chosen so that the
current in the horizontal coil is 90 degrees out of phase
-in phase quadrature- to that in the vertical coil. Now,
when the moving coil is rotated to the 90 degree
position, an output results which is 90 degrees out of
phase with the voltage A-B. In fact, any phase may now
be obtained by positioning the moving coil
appropriately with respect to the fixed coils.
We now have a means of generating any desired phase shift in an RF signal
between 0 and 360 degrees. In the HiFix receiver, the goniometer is used as part
of the display system. Consider the diagram below:
Let us assume for a moment that the IF input is in phase with the master
oscillator and the goniometer is set for zero phase shift. Then the two inputs to
the discriminator will be in phase, and its output at point A will by zero. The
servo amplifier output will also be at zero and the motor will be at rest. Under
these conditions, we can mechanically zero the display if we wish to read
000.00.

Now if the IF signal phase has changed, because we have moved with respect to
the master and slave transmitters- then the phase of the two inputs to the
discrimiator will be different, and correspondingly an ouput voltage will appear
crresponding to the amount and direction of phase difference. The servo
amplifier will now drive the motor in the appropriate direction, thus turning the
goniometer, until the zero phase condition has been restored. Here's an example:
We have moved 0.1 of a lane away from the master and slave stations. This will
correspond to a phase change of +36 degrees. The discriminator output will now
be a small positive voltage, which will cause the servo amplifier to drive the
motor in the direction needed to move the goniometer until its output is also at
+36 degrees, i.e until the difference between the discriminator inputs is restored
to zero. When this has been achieved and the motor is once again at rest, the
mechanical gearing will have moved the goniometer by 0.1 turn, and the
mechanical display counter will read 000.10.

There are of course two displays, one for each of the two patterns: master -
slave 1, and master - slave 2.

Slave Drive channel and Automatic Phase Control (APC)

In addition to the pattern displays described above, the receiver provides an
additional channel for use with a slave transmitter when the receiver is part of a
slave station installation. The pattern frequency IF is mixed with the output of
the four-channel oscillator to bring it back up to the correct RF frequency. A
goniometer and associated servo operate to control the phase of the drive signal.

Alarm lamp
An alarm indicator is built in to the receiver. When a phase difference of greater
than 90 degrees exists in one of the channels, the lamp is lit.

RF Attenuator
When part of a slave station installation, the receiver would be overloaded when
the associated slave transmitter is keyed. To prevent this, an attenuator is
switched in during the slave transmit period.

Photographs of HiFix receiver installations

Back to introduction page

Last update 27/4/05

Decca HiFix Antennas

The standard HiFix antenna system consisted of a 32ft vertical pole, standing on a base
insulator and earth spikes. The capacitance of the antenna was increased by using a top
capacity hat, which utilised the top part of the upper set of guys.
The antenna requires an effective earth counterpoise consisting of a large number of
radials extending out as far as possible from the base.

The photograph of the station at Seahouses, Northumberland is a typical example of this

type of antennna. Ian Jackson, G3OHX points out that this antenna design is designed to
minimise the skywave component, which of course would reduce the effectiveness of Hi-
Fix. Ian built antennas to this pattern for use on the 160m amateur band and confirms
that they only really gave groundwave range contacts. Ian also points out that the top
sections of the middle guys aren't insulated from the tower. Like him, I would have
expected them to be- in fact the original version of my drawing above showed this.

Back to system description

Back to introduction page

Last Update 14/5/05 prev 27/4/05 AC

Hi-Fix Station at Seahouses, Northumberland

Seahouses station has a continuous history of over 40 years providing precision

positioning in the North Sea.
Seahouses in the 1960's
The photograph below was taken in 1965, and shows a typical east-coast Hi-Fix station
with the standard design of antenna described on the antenna page.

All else about this station is also fairly typical incuding the equipment accomodation
which couldn't have afforded much protection- either to the equipment, or to an unlucky
engineer who had to visit the site in bad weather!

Photograph by kind permission of Ian Jackson, G3OHX.

Seahouses in the 1970's
Allan Papworth was working for Decca in the 1970's and took this picture of the
Seahouses station in 1971. The caravan apparently was considered an eyesore by the
Golf Club (I wonder why?!), who consequently wanted Decca off the site. At this time, the
station was the northern slave of the North Tyne chain, with the master station at
Hartlepool and the southern slave at Whitby.

Photograph by kind permission of Allan Papworth, G3WUW.

Seahouses in the 1980's and 90's

Some time around 1982, the station was converted to Hyperfix. At this time, the mast
was replaced by a guyed lattice tower. The wooden 'garden shed' hut was also most
probably provided about this time.

Seahouses in 2006
Following the demise of the East Coast Hyperfix chains in the mid 1990's, Seahouses was
converted to a Deltafix station by the addition of a GPS receiver and a modification to its
transmitter to allow transmission of the phase-shift keyed DGPS signal. It still radiates
this signal which sounds like a 2-second on, 2-second off data burst alternately on around
2830 kHz and 1900 kHz. The signal is easily audible throughout most of the UK.

The photographs below were all taken just before sunset. In this part of Britain there is a
unique quality of light at this time of day that makes the beautiful coastal scenery look
even more spectacular.
 Photograph: Alan Cordwell G0NFY.

The following photographs show bits of detail:

 Photographs: Alan Cordwell G0NFY.

Clockwise from top left:

 Signery on the hut door dates from the Hyperfix days.

 Base of mast showing feed point. You can just see, nestling in the grass, the rain
bell covering the insulator, and just to the left, the static discharge balls.

 The GPS antenna near the corner of the compound.

 The hut has seen better days. The board nailed onto it above the GPS cable entry
appears to say 'WASHING' in very faded paint: I'm intrigued to know where this
came from!

About Seahouses
The station is located at Seahouses in Northumberland, between Alnwick and Bamburgh
on some of the most scenic coastline in Britain. The station is situated on a headland
between the lake known locally as 'Logan's Loch' to the north, and the 10th hole green of
Seahouses Golf Club to the south. It is a local landmark, easily visble from the B1340
Seahouses to Beadnell road.
 Image produced from the Ordnance Survey Get-a-map service.
Image reproduced with kind permission of Ordnance Survey
and Ordnance Survey of Northern Ireland.

Seahouses: The future

Fugro, who operate the Deltafix chain in support of their hydrographic and surveying
operations, will likely keep the system operating for the foreseeable future as it provides
them with an in-house facility for precision positioning. So it looks as though the station
will remain here for a while yet. I wonder if it can make 50 years?

Last Update 25/10/06 Previous 10/9/05, 14/5/05.

About Seahouses
The station is located at Seahouses in Northumberland, between Alnwick and Bamburgh
on some of the most scenic coastline in Britain. The station is situated on a headland
between the lake known locally as 'Logan's Loch' to the north, and the 10th hole green of
Seahouses Golf Club to the south. It is a local landmark, easily visble from the B1340
Seahouses to Beadnell road.

Image produced from the Ordnance Survey Get-a-map service.

Image reproduced with kind permission of Ordnance Survey
and Ordnance Survey of Northern Ireland.

Seahouses: The future

Fugro, who operate the Deltafix chain in support of their hydrographic and surveying
operations, will likely keep the system operating for the foreseeable future as it provides
 them with an in-house facility for precision positioning. So it looks as though the station
will remain here for a while yet. I wonder if it can make 50 years?

Last Update 25/10/06 Previous 10/9/05, 14/5/05.

Front Panel views of Hi-Fix units

This page shows views of the front panels of Hi-Fix units.

Master Drive Unit

Receiver
 Low Power Transmitter

Distribution unit

Return to index page

Last update 2/5/05 AC

Hi-Fix/6

In June 1973, about 10 years after the introduction of Hi-Fix, Decca announced
that the Hi-Fix/6 positioning system was shortly going to enter production. The
system got its name from the fact that up to 6 chain stations may be used, but
Hi-Fix/6 bears only minor similarities to Hi-Fix in its transmission
characteristics, while the equipment used is completely different and much
more modern in terms of technology.

Naturally the operating principles are the same- that of position fixing by phase
comparison, in either the hyperbolic mode where the mobile station carries a
receiver only, or in circular or range-range mode where the mobile station
carries a transmitter as well. In addition, Hi-Fix/6 introduced 'compound mode'
which is a combination of hyperbolic and circular modes. The chief advance
with Hi-Fix/6 over its predecessor is that the compound mode can be used
simultaneously by up to 3 mobiles, or circular mode only with up to 4 mobile
stations. Hyperbolic mode may be used by an unlimited number of mobiles, of
course.

A Hi-Fix/6 chain consists of a 'Prime' station and up to 5 secondary stations.

The function of the prime station is to control the timing of the chain, and to
provide the transmission that the secondary stations phase-lock to. However, the
prime station does not have to be the 'Master' station in the context of patterns-
when using Hyperbolic Mode; the 'master' can be any secondary station
physically located in between two others. The prime station should in fact be
located where all of the chain stations may receive its transmission.

Hi-Fix/6 could provide both simple (single frequency) and lane identification
(two-frequency) working, and where the latter was used, the function was
automatic as far as the user was concerned. Unlike Hi-Fix, which required two
seperate sets of equipment for 2 frequency working, Hi-Fix/6 incorporated the
facility in one set of boxes. The operating frequiencies could be anywhere in the
band 1.6 to 5.0 MHz, in 100Hz steps. The second frequency used for lane
identification (or 'coarse' working) needed to be between 80 and 98% of the
main frequency. The relationship between operating RF frequency, lane width,
and coarse (identification mode) lane width is as follows:

Lane Width = 1/2 X Wavelength at the main frequency: 300/f(MHz).

For example, at 2 MHz the lane width will be 150/2 = 75 metres. The maximum
accuracy of the system is typically 0.01 lanes, which is therefore 0.75m.

The Coarse width is the lane width given by the difference between the two
frequencies. For example with f1 = 2MHz and f2= 1.8 MHz, the course lane
width will be:

1/2 X 300/(2 - 1.8) = 750 metres.

The complete transmission cycle occupies 260 miiliseconds and consists of an

trigger transmission from the prime station, immediately followed by that
stations main transmission. The other chain stations then transmit in sequence.
The trigger transmission may be clearly identified as such by all stations.

Click here to listen to a simulation of a Hi-Fix/6 chain in operation (22kHz, 16-

bit mono, .wav format; 7 seconds 304KB).

Receiver 90015
The Hi-Fix/6 receiver type 90015 is a solid-state unit employing integrated
circuits and frequency synthesis, and is constructed using plug-in circuit
modules, some of which are common to the transmitter unit to ease servicing. It
is designed to have the minimum of operator controls, and setting up
adjustments such as manual calibration and referencing have been eliminated.
The operator enters the operating frequencies on thumbwheel switches, and
three numerical displays give both the course and fine readouts from the
selected station pairs, in lanes and fractions of lanes. Additional controls allow
the operator to select the operating mode and pattern (i.e. station pairs).

The receiver contains a total of twelve highly stable phase locked loops, one for
each of the two frequencies transmitted by each of the six stations. It has
outputs that may be connected to external equipment such as plotters and
remote displays.

Rather than timing and phase locking being seperate functions as they are with
Hi-Fix, the timing of the Hi-Fix/6 system is derived from the phase-locked
oscillator. This means that once synchronisation is acheived, the timing will
remain in sequence while ever phase lock is maintained. This renders the
system reasonably immune to loss of the trigger signal.

Hi-Fix/6 Configurations

This details the different ways in which a Hi-Fix/6 chain may be deployed and
utilised.

In Hyperbolic Mode
In hyperbolic mode, a set of hyperbolic lines is plotted for each station pair. The
stations ar at the focii of the hyperbolae, and they cross an imaginary line (the
baseline) joining the two stations. Each station pair and its associated set of
hyperbolae is referred to as a 'pattern'. A minimum of three stations and 2
patterns is needed for a fix, but the more patterns are available, the more
accurate the fix will be.
 Simplest configuration: 3 stations, 3 baselines.

4 stations, 6 baselines.

5 stations, 10 baselines.

6 stations, 15 baselines.

Can anyone tell me why there isn't 16 baselines- why can't prime and Sec 3 be
on one?
In Circular Mode In circular mode, the position lines form a set of concentric
circles around each station. The circles represent range. The minimum
requirement for obtaining a fix is, therefore, to be able to determine the range
from two stations.

Simplest configuration: 2 shore stations, 1 vessel.

2 stations, 2 vessels.

3 vessels, coarse/coarse/fine positioning available for 1 only.

4 vessels, coarse/fine position not available.

In Combined Mode In compound mode, 3 of the stations are used to generate
hyperbolic patterns but also up to 3 vessels may use two of the stations for
circular mode positioning..

3-station hyperbolic chain, 1 vessel.

3-station hyperbolic chain, 2 vessels. Coarse/fine postioning is available on the

hyperbolic patterns only.

3-station hyperbolic chain, 3 vessels. Coarse/fine positioning not available.

Summary
Hi-Fix/6 is interesting because it represents a half-way house between Hi-Fix
and the much later Hyper-fix in terms of flexibility.

Return to introduction page

Last Update 5/6/05

 Sea-Fix
Sea-fix was a derivative of Hi-fix and is similar in its operating principles and modes except
that it can operate with a master chain station and up to 3 slaves.

I am indebted to Peter Ebsworth G8CKB for the following information, taken from the
Decca Survey Catalogue.
Sea-fix Units

9430 Transmitter Unit

9431 Master Control Unit
9432 Slave Control Unit Block diagram
9435 Receiver Unit
9436A Coil Unit, 6ft whip antenna
9438A Coil Unit 15ft Antenna

DECCA SURVEY Sheet No. 14a

CATALOGUE Issue No. 3
Sea-Fix, a derivative of Hi-Fix, is a small, solid state, accurate electronic position fixing
system for short-range hydrographic and oceanographic operations. The chain stations may
be either shore based or buoy-borne permitting such projects as: Sea-mount Surveys,
Minesweeping, Beach Control, Dredging, Hydrographic Surveys, Control of Unmoored
Vessels, Ships Trials, Construction of harbour works and Buoy Laying.

A chain comprises a Master Station and three Slaves [see images above]. If desired one slave
may be omitted. The transmitting stations may be installed in either the hyperbolic or ranging
mode (T,R,M,S for transmitter, Receiver, Master Unit and Slave Unit respectively). The
former is more flexible but the latter gives more precision at a distance owing to the nature of
the lattice.

The transmitters emit CW signals in sequence, all sharing one radio frequency. The signals
from the Master and each Slave in turn are phase-compared in the Receiver, giving two
pattern co-ordinates which are read off counters and provide the fix. For a continuous run of
fixes, a page copy or perforated tape is made available; when other data such as rectilinear
co-ordinates, depth-soundings, and time are needed, suitable computing and data logging
equipment is added.

The electronic units are housed in aircraft type covers size 1/4 ATR (long) excepting the
Marine Receiver, which is presented in a blue enamelled alloy case. Rugged sealed fibreglass
cases in an attractive finish are provided for the equipment at shore-based chain stations.
Environmentally tested and approved components are used; each unit is contructed of
modules to simplify servicing. The equipment is designed for world-wide use, which
includes desert, tropical and arctic climates.

Operating (pattern) frequency: Normally in the band 1600-2100kHz extension to 3000kHz

if required.
Trigger frequency: 60Hz below the pattern frequency.
Type of transmission: Interrupted Continuous Wave, time multiplex.
Master:F.9. Slaves: A.0
Switching Rate: Five times per second.
Radiated power: 1W (approx.) from 15ft (4.6m) vertical whip aerial
of 40pF capacity.
Operating range over sea water: 20 miles (37km) for world-wide operations.
Receiver IF bandwidth: 200Hz between 3dB points.
Maximum receiver speed: 600 knots (1110km/hr). (Note: One lane/sec with
type 9412 mechanical displays.)
Power supply: From 22 to 28V d.c. normally provided by secondary
batteries. Positive, negative, or isolated earth.
Positioning accuracy of system: Better than 2 metres under optimum conditions.

DECCA SURVEY Sheet No. 14c

CATALOGUE Issue No. 3
 THE SEA-FIX SYSTEM-INSTALLATION OF CHAIN STATIONS

While survey sheet numbers 14a and 14b say what Sea-Fix is and does, and how it does it,
the object here is to indicate the accuracy and operational coverage which can be expected of
the System. Both these requirements influence choice of position for the Chain Stations.

Selection of Station Sites

The Sea-Fix System is not critical as regards the selection of suitable sites for the Chain
Stations, although certain precautions should be observed if the best results are to be
obtained. In general, similar considerations apply to Sea-Fix as to the operation of
communications equipment working on ground-wave propagation at frequencies in the
region od 2 to 3 MHz, but owing to its desgin features Sea-Fix operates successfully in areas
of noise which preclude the use of communications transmissions in this frequency band.

Intended as they are for hydrographic or oceanographic use the stations of teh Sea-Fix chain
may be installed on shore or afloat. When sited ashore a station should be installed ideally on
level ground, preferably on moist soil rather than on rock or sand, and should be as close to
the water's edge as possible. Clear avoidance of tall natural features such as large rock faces
and trees is advisable, while contstructions like pylons, buildings, cranes, overhead traction
wires, high tension cables and telephone lines should be given a wide berth. When sited at
sea, in a buoy for example, the aerial should be fitted a few feet above water to avoid too
much sea spray. Dependoing on the amount of l;ateral drift allowed, the type and length of
mooring should be considered.

Position-fixing accuracy - Repeatability

Sites for stations (in both hyperbolic and ranging chains) are chosens to give a good angle of
cut of the position lines (as near as possible to 90 degrees) in the operations area. It is useful
to note that in hyperbolic chains accuracy can be improved by lengthening the base lines,
since thi snarrows the lane width in front of the chain, a characteristic which applies to three-
slave chains aslo.

For theoretical urposes it is convenient to consider Sea-Fix errors in two categories: first,
those inherent in the nature of the system, known as random errors, and secondly terrain or
'fixed' errors. Random errors arise from fluctuations in the propagation medium, small
variations in the equipment, the human reading of the lane counters and so on. They apply to
each pattern position line, and can conveniently be lumped into a figure known as the
'Standard Deviation' of that position line. In Sea-Fix chains with stations situated on the shore
and operating from short baselines the Standard Deviation is normally taken as 0.01 of a
lane, but when the base lines are longer and operational areas larger, it may be advisable to
use a larger figure. When the angle of cut of position lines, the lane width, the standard
deviation and a probability factor are mathematically expressed together, a family of
'accuracy contours' for the chain is obtained. The figure noted on a contour would expect a
repeat of the ofrmer erading within the limiting figure given on the contour; this is known as
the repeatability at that point.

Position-fixing accuracy - Absolute

Radio transmissions travel in free space with the velocity of light. For chains with signal
paths predominantly over the sea the propagation speed is slightly less, but the salinity of the
surface water and the amount of land intruding into the signal paths influence this speed.
Decca Survey Ltd has great experience in the selection of the correct propagation velocity
and this experience is put to good use in the calibration of Seaf-Fix chains.

In a shore-based system, when teh stations have been surveyed into the geodetic net, the
propagation speed has been established and the chain has been calibrated, the Sea-Fix
readings can be converted into geographical or grid co-ordinates as required. However,
certain residual or 'fixed' errors will remain. Should the project so demand, a table of these
errors can be compiled so that the degree of positioning accuracy of which the chain is
capable can be acheived.

In an oceanographic chain where the stations are waterborne, the station positions are located
by astro-fix, Decca, Loran, satellite or whatever method is available at the time and distances
between them are checked by lane counts. The ships course is plotted in geographicals at the
desired scale. The absolute accuracy therefore depends upon the initial fix of the chain's
position; the repeatibility accuracy is greatly influenced by the lateral drift of each station on
its mooring.

Operational Coverage
The extent of the coverage depends largeley upon the relative positions of the station sites,
teh type of antennae used, and the amount of land intruding into the signal paths. A certain
amount of chain planning is therefore necessary, and Decca Survey Limited offers a full
consultancy service in these matters.

Decca Survey Limited, Kingston Road, Leatherhead, Surrey

Telephone Leatherhead 76971 Cables:DECSURVEY LEATHERHEAD Telex:928437

SEAFIX TECHNICAL MANUAL II.I April 1968 Page 9

Station configurations

Sea-fix master station (hyperbolic)

 Sea-fix slave station

Sea-fix receiver type 9435 station (hyperbolic)

 Sea-fix two-range receiver station

Whilst I believe that the information above is probably definitive, I had earlier found
references to Sea-Fix in the NOAA Hydrographic Manual, which was first published in 1976
and was revised annually until 1981. This manual also gives a comparison between the
specification of Sea-Fix and Hi-Fix, and it states that the frequency range of Sea-Fix extends
down to 1600 kHz and the output power is higher than Hi-Fix at 100W. Surprisingly, this
makes no difference to the stated maximum working range. The manual also includes the
picture below, ahich appears to show 2 standard Hi-Fix receivers and a unit I cannot identify.
 Photograph reproduced by kind permission of The Hydrographic Society of America.

Since it is unlikely that the Hydrographic Manual published by as reputable an organisation

as NOAA would contain incorrect information, I am speculating (until I find out for sure)
that there was an American version of Sea-Fix which was different to the European one. The
unidentified unit could be a high-power transmitter; this would have to carried by a ship for
range-range working.

Perhaps this will remain a mystery- who knows? Well, maybe not. There is another
explanation, as Peter told me:

"Reading through your site and going over the SeaFix picture with two colleagues induced
me to search through old work files, and too I have remembered something and found an
explanation.

Quite a few of the Hi-Fix receivers, 9217, were modified to operate with SeaFix chains, I
have found a copy of the modification instructions. The mods centred around replacing most
of the relays and their driver circuits with multiple reed relays, making relevant component
changes to reduce timing constants and delays, etc.

One could say there were at least two development stages in the 9217,
a) with the valve cathode followers, balencing potentiometers and signal integrating circuits;
and
b) the next stage with the FET source followers on a small plug-in PCB. these were more
stable.

In fact I must have worked at these mods in my first week at the Decca Survey base depot in
'71.

Then the mods for use with SeaFix came along and Hi-Fix equipment life was extended. So
that NOAA photo could well have been accurate, and as I've written I have seen it in another
publication and been puzzled by it. I did work with some american sourced receivers in 1977,
[which had] UNC-UNF screw threads."

In the lack of further data, I'm happy to assume that NOAA sourced their SeaFix equipment
this way.

Last updates 28/5/06, 3/3/06