FARNHAM VHF GROUP
GB3FX - Specification
GB3FX is located on a site 3 km North of Farnham, Surrey, UK at:
National Grid Reference; SU821494
The site is 187m above sea level and is shared with GB3FN and GB3FM, the group's 70cm and 23cm repeaters. It has been operational since 10 October 1997.
A brief specification of the repeater follows.
Some philisophy behind the design: the coverage from well sited 6m repeaters is tremendous, although the band is still quiet. 6m is quite sensitive to electrical interference, for example ignition noise from older petrol engines, which is why the receiver has an AFC loop to try and minimise the effect of this. The logic has been designed to try and be helpful to users without annoying them too much. That's why the repeater has the simple 1 sec hang on initial access - if someone really wants to try to get in without giving his or her callsign, they can get brief confirmation without subjecting other listeners to lots of morse code! The signal strength indicator allows a station to judge their signal if there is no one else around. The overdeviation indicator is fairly unobtrusive - but as it's set quite high, stations really should adjust their deviation so as not to hit it, if they are going to avoid distortion and squelch closure on deviation peaks. We also recognised that the CTCSS-only access on 6m would cause some stations problems as they adjusted or found the right tone, which is why stations without the correct CTCSS tone can get a few seconds of low deviation talkthrough - hopefully enough for others to help them set up correctly.
The repeater runs a single dipole (home made by G4VDF) at the top of the tower. A replacement dipole has been purchased, to be put up in place of the home made antenna one day, due to the age of the current antenna.
We originally used a duplexer made out of Heliax from the WB5WPA design. It worked very well, cost virtually nothing, but because we only had LDF5-50 Heliax, was a bit lossy especially on transmit. Thanks to the extreme generosity of Rupert, G4XRV, who donated both his time and the materials, we now have a custom made duplexer made with real helical filters, which has been running since January 2003. And a good couple of dB less loss. This has helped increase the ERP slightly.
The repeater has had a couple of improvements over the years as well, especially in the receiver section where the IF and front and have been modified further to improve the sensitivity. We are now limited by the level of intermodulation products that come from an NTL site across the road, which doesn't seem to have the greatest filtering in the world on its transmitters. This also prevents us putting out more TX power, although the box itself is capable of it.
From 2009, like GB3FN, the repeater has been modified to transmit CTCSS tone at all times, including when sending its callsign when out of use. This allows mobile operators to drive around using a tone squelch to cut out the high levels of noise and interference on 6m, whilst still hearing periodic callsigns as a confidence and coverage check. This is not to everyone's taste, but the majority of those surveyed prefer this. Note that the CTCSS tone is not sent during callsigns sent after a timeout - this is to avoid any latch up situations involving local intermodulation sources close to the site.
|Transmitter||Storno CQF9334 modified from 70MHz band|
|RF Output power||28W|
|Power into feeder||16W|
|Effective Radiated Power||14W|
|Frequency accuracy||Better than ± 3Hz
Typically better than ± 1Hz
|Receiver||Storno CQF9334 modified from 70MHz band|
|Sensitivity||0.3µV EMF (0.15µV "PD") for 12 dB SINAD|
|Sensitivity into filter||0.4µV EMF (0.25µV "PD") for 12 dB SINAD|
|Antenna||Homebrew dipole with gamma match|
|Height||20M above ground|
|Filter system||Helical duplexer using two notches in each leg built by G4XRV|
|Desensitisation of RX by TX||Less than 1dB|
|Frequency reference||Racal 9420 ovened 5 MHz oscillator
(shared with GB3FN, separate from GB3FM)
|Access||82.5Hz CTCSS tone (RSGB tone D)|
|Reaccess||Carrier with CTCSS|
Every 5 minutes as beacon
On full access - from cold after 5 seconds modulated transmission
On shutdown - after a QSO
On timeout (every 20 secs)
|Reply signal||Tone pip(normally)
"B" if operating on standby batteries
"T" (long) after timeout
|Keying frequencies||1092 Hz (normal callsign)
546 Hz (timeout callsign)
|Tone frequency of pip reply signal
varies with signal strength, from greater than 2kHz if signal is 1µV EMF or more, to
600Hz for threshold. Less than 2kHz indicates to mobiles that faded signal is marginal. A
dual rate sample and hold circuit smooths out variations over last few seconds of
The receiver has limited range (+/-500Hz) AFC operation.
Callsigns suppressed on access and shutdown after repeated access, or no QSOs.
Audio muted by 20dB with no subaudible tone; allows signals with no or incorrect CTCSS to be heard quietly for maximum 10 seconds.
Full audio "coast" timer allows full audio to be heard up to 1 sec after CTCSS loss (allows decoder to be talked out by weak signal or overdeviation).
Overdeviation signalled by short audio mute and low level 546Hz pip.
Reaccess possible in timeout, needs either carrier without CTCSS over the top, or repeated keying on and off for a couple of seconds to shut down the tone decoder.
|Battery lifetime||In excess of 24 hours of mains interruption given normal use patterns.|
|Battery type||Dryfit 12V (2 x 6V) 150Ah|
|The repeater is
reciprocal with a mobile running 12W and a receiver with 0.5µ EMF (0.25µ "pd")
Coverage Maps Scanned in from prediction program. There are two of these, with different parameters entered.
The 'high' coverage map gives an indication on where a high specification mobile installation (greater than 25W, quarter wave antenna, good groundplane for antenna) should be able to access the repeater, but with some limits. The more solid central area gives 90% probability, the hatched area gives 50% coverage. The 50% coverage area should be OK for a well sited mobile, or a fixed station. The prediction maybe doesn't allow fully for the extent of electrical noise experienced at 50MHz, and performance may be a little worse depending on environmental conditions. It also doesn't fully account for clutter loss in urban areas, so results towards London are a little worse than this map predicts, and results to the South and South East may also be slightly worse due to some shadowing by the mast. North and West seem about right.
The 'low' coverage map was drawn up with more pessimistic parameters (and sorry about the disgusting colours, we'll do another one some day). This is where a high specification mobile should definitely have good access, and a lower specification mobile (less than 25W, loaded whip less than a quarter wave, maybe poorer groundplane on antenna) or maybe where a well sited portable should be able to access.
Especally with antenna efficiency and background noise on 50MHz, the maps are less predicable than at higher frequencies. Many mobiles will achieve somewhere between the two - depending on which direction they are pointing in and so which direction their antenna works best in. We'd welcome Feedback on users' experiences. For more info, see coverage.
NB: Note that the prediction program prints the names to the right of towns' locations without showing the towns. So the town itself is just to the left of its printed name.
|'High' coverage map|
Update Apr 2010
Copyright © 1997-2010 The Farnham VHF Group