|
|
 |
|
|
Multipliers for Microwave LO’s
|
|
|
|
Copper pipe cap filters and MMIC gain stages work well in delivering a clean LO signal for microwave transverters. Jim Davey WA8NLC investigated MMICs as multipliers some years ago in his article “Frequency Multipiers using Silicon MMICs”, avialable on the ARRL UHF/Microwave Projects CD. For my transverters I didn’t need to stick to the normal scheme requiring an intermediate stage around 1Ghz, so the thought of taking a 500Mhz signal from a crystal oscillator and multiplying that directly up into the microwave region was appealing.
|
|
|
|
|
|
On the left is a very basic test fixture using two MMIC stages either side of a 1” pipe cap filter. It’s an effective x6 multiplier with an input of 568Mhz @ 10dBm the output is shown on the specan at 3408Mhz at 4dBm with spurious responses better than 35dB down. The MMIC stages were mounted on Minikits experimenter boards and lashed to the same (nominal) 9v supply. It’s quite instructive to vary the supply voltage while monitoring the multiplier output, in fact it would be preferable to run the first MMIC stage (the multiplier) from a separate supply so the current can be set for maximum output on the desired frequency. A number of factors affect the MMIC response as a multiplier, the DC conditions are just one important factor.
|
|
|
|
|
Multiplier schemes for 9, 6 and 3cm:
|
|
|
|
|
|
I tried each of the basic multiplier schemes at left with success using the test fixture shown above. I was limited to testing the first multiplier stage only, however this resulted in sufficient information for me to come up with basic circuits for each of the multipliers. My first attempt was a module for 9cm’s delivering an LO frequency of 3256Mhz for an IF of 144Mhz, keeping in mind the 9cm band allocation is 3400Mhz here in VK.
|
|
|
|
|
Below are two photos of the completed multiplier for 3256Mhz. The tuning range of one inch pipe caps easily covers this frequency, and the circuit was etched on Rogers RO-4003 laminate. The line up is an ERA3 as the x6 multiplier then an ERA3 between two filters followed by an ERA5 on the output This configuration helps to preserve a 50ohm termination at the last filter. FR4 laminate could likely have been used in this case, I just didn’t have any double sided FR4 at the time. The overall dimensions are 45mm x 145mm, fitting a 220 degree stripline section between the multiplier and first filter added significantly to the overall length at this frequency, however there would be options I suspect for fitting that into a smaller board area. Overall the board design is very simple using 50ohm microstrip and de-coupling around DC feeds. The output from this module is in excess of +10dBm with an input of 0dBM at 542.667Mhz.
|
|
|
|
