The Transmit Line-up
9 M Hz pre-driver and ALC

The Mark 1 Pic-a-Star IF produces a 9 M Hz signal. This is raised to about 10 milliwatts using the CDG2000 post mixer board circuitry. This also incorporates the post roofing filter amplifier (4 by J310) for the RX side. The layout used is essentially the CDG2000 post-mix board cut in half without the filters and switching relays. The PCB layout already incorporates the alc circuitry and this is reliable and works well. The idea was to raise the 9 M Hz signal to the maximum tolerable (about 10 milliwatts) by the mixer. This reduces the amount of signal frequency amplification required and helps to maintain stability. Following the mixer my pre-amp (2 by mav11 in push pull) acts as a post mixer/pre-filter amp and raises the signal to 10 to 20 milliwatts giving 5 to 10 milliwatts at the output of the bandpass filters.

PA Circuits

The PA circuitry comprises a 2n3553 pre-driver. This transistor is still available (Sycom) and running 100 ma can produce over 200milliwatts of drive in class A.    In the picture below it is in the little diecast box.

The main amplifier uses a pair of mrf 476 because I had them though the 2sc1969 is cheaper and might be better. The final uses two 2sc1969 citizen band transistors. These are available at Ł3:50 from several sources (including sycom) and are tough. I have inadvertantly run mine at full drive into an open circuit without damage! The circuit uses the excellent bias arrangment of "Harry" SM0VPO and more or less his layout. The circuit is a combination of Harry's/the Motorola circuit and my own!! The pcb I used has a few more pads than needed.

HEAT SINK
Perhaps the most innnovative part is the heat-sink. This is an early Athlone CPU sink but similar ones are available from computer suppliers (Dabs, Simply etc) for about Ł10. I tapped the bottom and bolted the transistors on through a hole cut in the PCB. The thermistor is araldited to the sink between the transistors.   (I had never tapped a hole before, but a metric 3mm tap set and correct drill is available from Axminster tools for about Ł6 and is a doddle to use). The transistors are insulated with high conductivity military spec pads (Rapid). The fan is downrated from 12 to 6 volts. A soak test at 2 AMPS (25w DISSIPATION) standing current barely warmed the sink and the fan is almost silent at this speed. The whole replaces the lid of a die-cast box.

For the Mark 2 STAR I used a low noise heatsink Athlone from Rapid. This runs a class A PA at a standing current of 2amps each in the FET finals. At 18 volts this is over 70 watts dissipation but the fan-cooled sink runs cool and quiet!

OUTPUT TRANSFORMER

My first version used two brass tubes to make a one turn primary on FT77 cores. This worked fine but the FT77 was too lossy at 30 M Hz. The final version uses FT50-43 cores.    This did not provide enough primary inductance with one turn so the primary is made of two turns of 2mm silver plated copper. This is made by forming two "U" shaped loops each with a longer and shorter leg. The loops are covered with silicone insulating tubing and slipped through the cores (6) from one end. The two short legs are flattened with pliers and soldered together. The longer legs are shortened and the whole connected to the board with 2.5mm solder tape. The secondary is 4 turns of 20 gauge copper coming out the opposite side of the core. Theoretically 1 : 3 is a better ratio but the Motorola technical leaflets suggest 1:2 is better in practice. Suffice to say I can get 60 volts PP out of mine and up to 20 watts on the lower bands. There is no low pass filter as I am driving an FL50B and this neat little unit incorporates filtering. The drive needed is 5 watts so the amp has a bit in reserve.

The Mark 2 is tidier and uses FETS in the PA ...in class "A"   BLF247's well underrun at 18volts.
It uses toroidal cores as specified by Phillips and available from Barend Hendriksen

Well there you are. There is still some work to do but its on the air.

Mike G3XYG