76GHz Equipment
The system shown above comprises a head unit and a PLL local oscillator housed in a diecast box. The head unit is mounted on the rear of a 35cm ex-link dish fed with a penny feed arrangement. An FT290 acts as the 145Mhz IF. The photo above shows the equipment in use at Tog Hill near Bath (IO81TK48) working G3FYX at his home QTH.
Transverter
This photo shows the head unit of the 76Ghz DB6NT transverter with, from left to right, the 9.5 to 19Ghz doubler, the 19 to 38Ghz doubler and finally the mixer / doubler pcb using a dual flip chip diode type MAE1318.
A newer unit shown below dispenses with the 9.5 to 19Ghz doubler and uses a Millicom multiplier to go straight from 9.5Ghz to 38Ghz.
PLL LO
The WA6CGR PLL is shown above, which controls the 9.5Ghz oscillator and provides the LO. This is locked to a surplus ovened 10MHz reference oscillator which can be seen as the shiny cube in the bottom left of the picture. On the underside of this box is another containing the G4DDK 004 oscillator/multiplier and a WDG 003 multiplier pcb.
Like the 24Ghz system described elsewhere high stability and good phase noise are a must. Thus the crystal used in the LO must be of high quality and age predictably. The crystals are usually specified for operation with an oven temperature of 60degC and are supplied by Eisch-Kafka.
This system has now been superceded by a newer system using the G8ACE Reverse DDS system locked to a 5Mhz source. This is easier to use as it employs a PIC to set an AD9851 DDS chip to the correct synthesis frequency to allow the OCXO to be synthesised down to 5Mhz for comparison with the reference frequency. For further RDDS info also check the site of Brian G4NNS
Listen here to a typical FM QSO on the band
A "high power" multiplier source
This 76Ghz X2 diode multiplier uses a single flip chip GaAs varactor diode type MA46H146 in a DB6NT transverter type pcb, pumped with a 38Ghz Millicom module working as a x3 multiplier. The Milliwave module is driven at ~20-30mW by a 12.66GHz crystal controlled source which can be keyed or FM modulated. Power output at 76GHz has been measured as being ~2.5mW.
Components for the systems shown are available from DL2AM and pcbs from DB6NT
The crystal controlled source block diagram is shown below. This was quickly "cobbled together" when it was realised that better power could be obtained using an FM/ CW source than the transverter with its limited SSB uW output, although the receive portion is still used.
A DDK004 module is used, but its internal oscillator is disabled and an external ovened crystal oscillator substituted instead. (A G8ACE OCXO and a X24 multiplier might be a better solution but were not available at the time). A X5 multiplier, based loosely on an old WDG 11GHz design using 3 NE325 GaAs fets, follows which then drives the Millicom module.
An alternative transceiver system
Experimentation has been done recently using two dielectric 36dBi horn antennas (above) instead of a single dish. This gets around the need for a waveguide switch for transmit and receive and its thought the horns built for 60Ghz may have better gain than say a homebrew feed system in an ill-illuminated dish of unknown characteristics. This configuration has the high power FM source on one horn and the transverter on the other. Thus one can receive and transmit thro' the tranverter on the right horn or transmit only at slightly higher power on the other while still ensuring alignment on the distant station.