by John Brown - G3DVV
PHOTO 1: Top view of completed unit
This unit is an attempt to make the simplest possible signal source for 13cm without the usual grief of not finishing up on the right frequency. It has, of course, been constructed with Phase 3D in mind. It is built on a 4.25 x 1.4 x 1/16 inch epoxy glass fibre board.
Inspection of the circuit diagram shows that there are just three transistors: an FET oscillator, a BJT frequency multiplier (times 5), and an output transistor frequency quintupler. The oscillator starts with a readily available 96MHz crystal and multiplying this by 25 goes up to 2400MHz. The use of filters takes the guesswork out of finding the correct frequencies.
Download the circuit diagram
The output is -10dBm (100 microwatts) and the nearest sprog is on 1920MHz. This is -38dBm down. The output is too large for front ends (use an attenuator) but not sufficient for checking antenna systems (use a MMIC).
The FET oscillator uses a J310 in a standard circuit. No attempt has been made to obtain a signal clear of sprogs as the two filters will give all the cleanliness required. The oscillator gives 1 milliwatt drive to the 2N5179 quintupler. The output from the 2N5179 is fed into a Toko filter (Toko filters are the unsung workhorses of the VHF/UHF bands - what we do without them!). Some attempt has been made to match the 480MHz output to the filter and C12 is used to peak up the signal. The output from the Toko filter is approximately 1mW and is very clean.
Some care is now required over the final matching. The output from the Toko filter at 50 ohm is L matched to a transistor type AT41486 ( not too expensive) whose input impedance has been taken as 12-j8. The final in and out stages use microstrip lines. I use Chartpack rub down lines on the PCB before etching - but that's another story.
Figure 2: 2400MHz Signal Source pcb layout
Etched side of board showing tracks and placement of components. The board is 4.5" long, 1.5" wide and 1/16" thick and is double sided epoxy fibre glass. It is etched on one side only, the other side forming the ground plane. Careful use of a file will be needed to fit in the box.
Download the pcb
The fifth harmonic from the AT41486 is matched into a Neosid filter. An earlier version used a variable capacitor but that did not work very well, so a microstrip line with the appropriately placed matching stub is used. The output impedance has been taken as 80-j48. For further details on this see UHF/Microwave Experimenters Manual chapter 6.
The output from the Neosid filter is again 50 ohm microstrip and the BNC socket matches up against the edge of the PCB, the metal extension on the back of the socket having been removed in the lathe. You may have to tune the Neosid filter to peak up the signal but adjustment of the Toko filter is not necessary.
PHOTO 2: Underside view showing component layout:
Use the photo above to determine the position of various components with respect to the pcb layout diagram. "Earth" connections are made through to the groundplane side of the pcb, most of the components being fixed on the track side (see photo).
Components marked with a * are mounted on the plane side. The coffin lid capacitors go through the board and Q3 is flush with the board. The name of the UK supplier is shown for the more unusual items.
The unit consumes approx 38mA