Two years ago I was arm twisted to provide the 432MHz ststion on VHFNFD for the new Pembrokeshire Contest Group.
I could write a book on the trials and tribulations of resurrecting a system that hadn't been powered up for probably ten years plus. One challenge was to turn it into a twin system transmitting into two antenna arrays. This required an unequal power split as the two PAs needed different power levels. After some frantic last mninute experimentation, a coupled line power splitter was made by hacksawing some copper board to make the two lines and supergluing them to an earth plane - it worked remarkably well. The split was about -11dB. The design was created using the coupled lines model in QUCS.
Roll the clock on two years - can we do better in terms of a power splitter? A lower split of around -8dB would be nice, but would that be physically possible with very closely spaced lines? The design was done in QUCS, implemented in DesignSpark and then the board produced using my homebrew UV exposure unit. The design called for lines 12.5 thousands of an inch apart.
The interesting part was just how long it took to etch the narrow gap between the lines. I normally reckon about 90 seconds to etch a board. It took a further eight minutes to etch that fine gap. Looks good but does it work?
The impedances look good on the network analyser but sadly the coupling is about -11.5dB. I nor actually believe that coupled lines reach a limit at around this degree of coupling.
I will try the old hacksaw produced version and this rather better looking version on the VHFNFD site to compare performance. But an interesting exercise to use two software packages and the home PCB facility to produce a nice looking product.
No comments:
Post a Comment