10 GHz 50 Watt Home Station

Coming Soon!!!! Using a TWT Amplifier all the time for just general QSO's on 10 Ghz is a bit cumbersome so i set about to provide an alternative local system using a HPA/LNA remotely mounted at the dish. This has been achieved and is being used almost nightly for local work. More detail on this and the RG214 coax used to feed the system SOON!!

At the beginning of this project, a list of criteria was set down. After building 2 previous 10 GHz transverters (primarily for portable use) the "specification" for a Home QTH based system was a relief. After all Power consumption and operation down to 10.5 Volts with a cold battery was not an issue, along with Mil Spec shock resistance for the occasional 6-foot drop! The following points were considered:

Because of my parallel interest in ATV, bandwidth was required to cover the nominal 10,350 MHz ATV allocation. The final result is such that a 414 MHz FM TV signal can be inputted to the TX mixer to produce 10 GHz ATV atleast. For the time being, however, I have no immediate plans to use of 50 Watts of ATV on 10 GHz though!

The transverter system is to feed a 650mm offset fed dish mounted on the rotating section of the tower. Currently the dish has a Masthead 1W HPA / LNA combination with the eventual aim of waveguide feed from the base of the Tower where the TWTA will reside.

Having built DB6NT transverters before, I have used a Qualcomm "OMNITRACK" modified for 10 GHz. The Qualcomm transverter provides an excellent base for the project. The Receiver Noise figure is in the region of 1.5db (thanks to Fujitsu FHX13LG HEMT's). The Transverter provides about 15 mW's output at 10 GHz. In this application the TWTA will only need < 1 mW to drive so a combination of coax loss (The TWTA is mounted remotely) and attenuation has been used to provide the exciter with a good 50 ohm load and reduce the power level to ~0dbm.

The Qualcomm PLL LO has not been used in this instance. As I have been experimenting with PLL's for 100 MHz overtone oscillators, I thought this project would make a good test bed for this alternative. In practice, the PLL system used in this project provides superior phase noise and spurii characteristics, when measured on the Spectrum Analyser. This was thought to be especially important with weak signal operation.

The PA is a NEC 40Watt TWTA, originally tuned to 14 GHz for use with satellite uplinks. The modification of the TWTA is to be covered shortly. Modification is largely the removal of the original FET driver amplifier and the WR62 Plumbing of the TWTA output. A simple manifold/matching section using WR90 has been made up to replace this. At this point only the RF monitor coupler has been included; the SWR protection circuit has not been re-included.

The TWTA needs about -2 dbm to drive it to about 55Watts output on 10368 MHz. The coax and attenuators between the Transverter and the TWTA have been selected so that it is impossible for the drive level to exceed 0dbm under conditions of overdrive/saturation.

The TWT is intended to me mounted at the base of the tower ( For Tropo work) or the base of the dish for EME work.

It is also possible to use a standard 1 Watt Qualcomm PA, inplace of the TWTA. As of 1/1/99 I now have the remotely mounted 650mm dish with masthead HPA / LNA ( All Qualcomm) mounted up the tower a 40 feet and fed by RG8 style coax. The 20 db loss is easily over come by the 33 db gain LNA and the -10dbm drive requirement of the HPA.

A standard motorized WR90 changeover switch has been employed. A PNP transistor triggered by sequenced outputs from the EME66 sequencer (see further) drives each Motor coil with 28VDC. The transfer relay provides a virtually lossless changeover, easily handling the 50 or so Watts of the TWTA. The The output of the TWTA is switched to a 50 Ohm load during receive. ( as is the LNA during tx).

The Omnitrack Transverter requires LO injection at one forth the final LO frequency, at about +7dbm level. As previously mentioned the Qualcomm PLL has not been used rather a PLL locked Overtone oscillator.

The actual local oscillator strip, is 1992 technology, borrowed from my first transverter (which now has a Qualcomm PLL!) and is best described as a test bed oscillator. The line-up was originally used for my experiments with active temp. Compensated local oscillators. Stability in these systems, translates to about + or - 1 KHz at 10 GHz. The VK5VF beacons on 1.3, 2.4, 3.4 & 10 GHz, as well as my main portable Transverter, use this technology. I am now using LM35xx Temp sense IC's instead of varistors, subject to a further article.

With the PLL system, the requirement for a well-isolated Oscillator has become redundant. In fact you can play with the Oscillator tank circuit with your finger and hear no change to a beat note on 10 GHz. Fancy that! The 106.500 MHz oscillator is a design borrowed from the KK7B 540 MHz LO. A Butler oscillator could have also been used although differing opinions seem to take both sides as to which has a lower phase noise contribution.

A 2-stage multiplier follows the Oscillator to 639 MHz (a cut down G4DDK001 PCB). This is followed by a KK7B X4 multiplier modified from 2.3 GHz to 2.556 GHz (Filters cut down to 32mm). The output is a MAR4 MMIC giving +8dbm.

If I was to build a new strip, a G4DDK004 or G4DDK009 LO would replace the three separate modules. For this project, atleast, I am pleased that I have used something out of the junk box!

The PLL system simply locks the Overtone oscillator, on 106.500 MHz, to a more stable 10 MHz Oven Controlled Oscillator. The PLL system can be adapted to lock Crystals from 70 - 116 MHz. An external reference could be used from say an ovenised frequency counter or perhaps a referenced locked to the Sync pulses from ABC TV (locked to a Cesium standard).

Why worry about a PLL system? I have 6 Microwave transverters for portable use, all using Crystal oscillator chains. To upgrade portable stability spec's I am slowly adding the same PLL system, locked to a single Reference oscillator placed in the Portable box. Then all I have to do is put digital readout in the IC202! Click HERE to view the complete article on the PLL.

Mark Kilmier, VK5EME, was doing a parallel project for 5.7 GHz / 10 GHz Transverters and had already developed a Transverter Sequencer. The sequencer makes the older YU3UMV and EME16 IF PCB's redundant! The project also is to be descibed soon on its own page.

An important factor when running TWT Power levels is proper sequencing of the Receive - Transmit Changeover. The Transfer relay takes about 250 mS. to changeover. In this time the TWTA must remain disabled and the Exciter turned off. The Exciter must also be turned on after the TWTA has settled for 100-200mS. The reverse must occur when going from Transmit- - Receive.

I use the standard convention of 9 Volts on the IF feed during transmit for changeover. This is preferable to RF sensing as the number of relay operations is reduced significantly. The IC202/TR9000 & TR751 IF rigs all have 9 Volts supplied via a 4K7 resistor to the antenna terminal. If you use the transceiver into a standard yagi with a DC grounded feed the 4K7 resistor easily handles this case scenario.

Mark's sequencer can be set up for either standard (DC on RX or TX) plus also has a RF sensed failsafe circuit to change over the system. Three sequential Transmit outputs are provide, capable of switching 5 Amp loads, plus a single receive rail. Mark's Website for kit details Tx Sequencer

Nothing special here. Standard 3 terminal regulators and a 7662 Inverter IC have provided five supply voltages.

The transformer is toriodal to reduce EMI problems. A front panel fuse disables all regulators from the Source voltage should a problem occur.

A rack width box has been used (formerly a 1.5GHz sig gen upconverter box). For a change I have allowed ample room to add more circuits. The Transverter sits on the TWTA so the Waveguide output of the TWTA can be connected to the WR90 Transfer switch. The TWTA still dwarfs the Transverter, see photo!