PROTOTYPE INSTALLED IN THE G4EZT SHACK Jan 2011 under soak test before moving to its permanent location outside in the shed, note it is Bias Tee powered and rig 2 in use.
There are quite a few commercial or other nice (and not so nice) SO2R antenna switch implementations documented on the web, however after a couple of pints of beer (Ringwood Fortyniner) Cliff G4PZK and myself decided to do something slightly different.....
This unit is intended for general dxing as we both have 2 radios and multiple antennas, 5kW power handling is not a requirement! However it is designed for 1 kW continuous carrier into a VSWR of 3:1 or upto 1.5 kW into a good match. It will handle two 1kW Transceivers and select one of 6 antennas for either radio or bypass either radio to an independent bypass connector. The bypass could be used for a dummy load, chaining units to make a 12,18,24 way switches, a VHF antenna, a multiband vertical, remote ATU or any other user function. A further use with one transceiver is to use its external antenna input for rx only connected to the switch port 2, then a selection of separate TX and RX antennas can be used.
We have some other applications in mind for the future to complement the switch e.g. a single radio 12 way remote switch (designed) with the same specifications, 1kW dc tap (finished) and the G4PZK remote balanced 1 kW ATU (in progress) - as usual no parallel port control decoders or multiple dc relay switching wires will be needed. However in deference to existing primitive systems and linears (which we also have) the PC App provides a fully cutomisable parallel port interface for both radios to control via the PC. The 2 serial ports in the remote switch can also be used to control legacy or new equipment via a simple hardware decoder.
Many thanks to CDXC members Chris G3SJJ and Ian GM3SEK for further ideas to improve the product, following on from which we have designed in a dc bias tee tap circuit at the A antenna input of the switch pcb, this will allow remote operation and dc powering of the switch with only the one coaxial cable (the exisiting feeder) to the switch. The B channel also has a Bias Tee which only powers external circuitry allowing DC at +/- 30V at upto 1.5A. The dc tap circuit will be also be available in a screened diecast box to fit at the shack end to inject the DC at 1.5 Amps. It is also planned to multiplex the WiFi signal or other 2.4 GHz signals along the feeder if certain applications benefit from this approach using the same tap design.
Prototype standalone Bias Tee.
Back to the switch.... Control software and hardware prevents both radios selecting the same antenna 'deliberately' or in a fault mode. When 'deliberately' selecting the same antenna the last radio to request that antenna is routed to the bypass port which would sensibly have a dummy load on it or maybe a remote all band antenna.The switch is fully software controlled over WiFi or via a USB cable when used locally in a shack. It can also operate in a 'stand alone' mode (if there is no WiFi) once antennas have been defined, needing only 12 V power and a short burst of low power on the band required to set it up. Internet Remote control operation is built in due to the internal web server. Power meters and frequency counters are also included. The uncommitted relays can control a 4 square antenna band select relays or alternatively remote a atu such as the SGC230 allowing lock and reset functions, DC power can be fed out from the switch making these external installations remote controlled and powered ( Port A DC must be 12V nominal as it powers the switch electronics and relays, leaving 0.5A for external units, Port B bias Tee is uncommited and can be upto +/- 30V DC at 1.5A for any external connection.
Colin G0CUZ of Winlog32 fame has kindly offered to provide us with an interface to WINLOG32 when we finish the software DLL definition, this could pass over Power, SWR, Freq, Temp, antenna in use etc to generate automatically QSO details for the QSL card and of course command the switch to follow the radio(s) frequency.....
Design Spec: Power handling <54 MHz: 1 kW CW, SSB 750W RTTY. Load < 3:1 VSWR
Frequency range Antenna Ports: 1.8 -72 MHz, (150W 70 MHz)
Bypass ports (x2) Freq range: 1.8 - 148 MHz. (150 W 70- 148 MHz)
Antenna ports (x6) Loss: <0.1 dB to 30 MHz, <0.2 dB 50-72 MHz,
Isolation any antenna port >60 dB upto 30MHz, >50 dB. 50-72 MHz.
Isolation rig to rig port as antenna port but may be greater depending on bank selected, 80 dB below 14 MHz - tbd on pcb v2.
Bypass Ports Loss <0.05 dB 1.8 - 72 MHz, <0.15 dB below 148 MHz
Power Meter One per channel, fwd and ref power (1.5kW to 1W)
Freq counter One per channel, 1 kHz resolution (for auto function)
Freq counter error <5 Khz 1.8 to 72 MHz over temperature.
Supply: External 13.8V nominal at 1.5 Amp, optionally via port A coaxial input using internal bias Tee with output from switch to other units from Bias Tee on cable B. Built in A/D to check tap DC A o/p is ok.
Control: Only via WiFi (built in), USB or automatic.
Control Range WiFi 50m, upto 1 km with high gain external antenna
USB via 5m cable for in shack application.
Control Outputs: 2 * relay isolated spco terminals, WiFi, 2*Serial ports.
Indicators LED's for WIFI/USB link active and 12V dc power on pcb.
Temperature pcb sensor read by software used for rf calibration also.
Safety: automatic hot switching prevention by rf detector
TCVR A or B to same antenna lock out and routing to load.
Automatic operation: standalone mode available (antennas bands pre-defined).
Size: 14 x 8 x 2 inches including external Aluminium Box. External waterproof cabinet is required outdoors!.
Connectors: N-type or SO239 or mixed, + SMA male socket for external WiFi ant, latching XLR for Tap DC out, 6 pin locking Mic connector for relay o/p.
PC CONTROL SCREENSHOTS: JAN 2011.
Test: Channel A -20m IC7700 plus ACOM1000 to Steppir at 1.25:1 VSWR, Channel B-IC756 to 50 Ohm load on 40m.
The frequencies at the bottom are those measured by the switch, at the top is the rig freq from the CIV if connected.
Example of antenna to frequency set up,
In my case port A1 is a 3el Steppir, Port A3 is a 60 ft vertical - The Aux relay1 selects 160 or 80 on the vertical and Aux relay 2 does SSB or CW switching. Port 4 is a 40m or 30m antenna.
Parallel Ports control page:
Radio control of switch at G4EZT
This shows part of the User interface, most parameters are configurable, e.g colours, display units (W,dBW,dBm,RL,degC,degF, peak hold etc).
January 2011. No logging interface yet due to successful direct rig control, but have tested port isolation and it found to be far greater than the antenna to antenna isolation. In fact the ic756 operated ok on 14MHz with the IC7700 at full power cw on 40m, no external filters.
December 2010, hardware complete, 2 complete units and bias Tees now built and operational at G4PZK and G4EZT. Software for automatic antenna selection by the spare IC7700 RS232 port complete.
September 12th, External Bias Tee feeding 12V via coax to switch using its internal Bias Tee tested at 1kW, no problems.
September 7th, Internal Bias Tee and power meter tested at 1KW 1.8 to 50 MHz and 3:1 VSWR all ok. Web control and display tested between G4PZK and G4EZT with 2 Telnet one web IE8 and a USB connection concurrent. Web interface tested on IE8, FireFox, Safari. Discrete Bias Tee for shack end designed and proto built. Software for the Power Meter calibration per band and at 4 power levels implemented. Software for Auto switch function implemented. BITE function added to check 12V supply via Bias Tee.
September 1, Bruene couplers and power calibration software finished. Internal DC bias Tap working and tested with 1KW 160 - 6M. Web server basic functionality checked over internet.
August 17 th Back from Holidays
Bruene coupler partially tested over 160m to 6m and 1kW to 1W, unfortunately my dummy load blew up.
REV2 PCB (July 2010) CAD design finished, pcbs arrived from Germany on July 14th.
Testing and contruction will be delayed by the summer holidays hi !
The new pcb allows for internal screening divisions in an attempt to further improve the isolation inside the box between radios and also the new Bruene coupler power meters. The RF connectors are now fully tracked on the pcb which will simplify construction and help isolation and losses.
Prototype box below:
The A/B chanels and bypass ports can be seen and the 6 antenna positions. The 12V DC connector will either power the switch or power an external unit if the switch is powered by its internal Bias Tee. The relay connector allows control of basic external units requiring dc switching. The WiFi antenna rounds off the unit. The indicator LEDs are not fitted.
REV1 PCB (April 2010)
The double sided thru hole plated pcbs have returned and the first prototypes are built and operational 1.8 to 148 MHz. The extensive software is being developed by G4PZK and I'm optimising the RF hardware which also includes a power detector and frequency counter for each Radio channel as well as the RF switching circuitry.
Testing so far confirms operation at 1 KW and low path losses up to 72 MHz, see Antenna path plot below:
The power detector for hot switching protection is functional as is the frequency counter up to 72 MHz with 1 kHz resolution. The power meters are now under development, a Bruene compensated bridge type is used with low cost Log Amp demodulators.
Photo showing the first pcb software development model. Wifi controller (antenna not fitted) and rf detectors are in the screened box area, safety control relays and the antenna path relays can be seen with their GCPW (Grounded Co-Planar Waveguide) microstrips. The RF switch bypass ports are not on this pcb version.
BREADBOARD PROTOTYPE: (Jan 2010)
The proto-type gives 0.1 dB loss at 50 MHz and ran cool at 1000W carrier for 60 seconds into a dummy load, I didn't want to upset my amp doing it any longer!
Obviously this is just a single relay cell, the PCB, Software and control electronics is much more complex and currently under development.
Proto-type single antenna 2 radio version photos below:
The insulated wires are nominally 50 ohm impedance over the ground plane.
<0.1 dB Loss at 50 MHz
>55 dB isolation at 50 MHz