i’ve been aware of this project for quite a while, however Glen VK3YY reminded me of it with his two talks at the VK3 SOTA conference last month. It was building his multiband efhw that I finally had the need for it as I initially had some problems working out where it was resonant using my older tenna-dipper.
I built mine on veroboard without any particular care regarding grounding or rf techniques.
I initially couldn’t find any germanium diodes, despite being sure I’d added some to a recent order thinking they would be useful, so I used some 1n5711 Schottky diodes instead. This turned out to be a mistake despite the small difference in forward voltage drops. It didn’t measure much of the forward power, but it did return small readings for reflected.
A further search revealed the OA95 diodes and these were substituted resulting in much improved voltages. Dropping R6 on the DDS board from 3.9k to 2.2k increased the output to 1v p-p with a massive improvement. A LM358 was used as well, as far as I can tell there are no stringent requirements on the opamp in this circuit.
I installed processing and Glens script on the netbook (running Ubuntu) and the following graphs were produced for the end fed:
These are in rough agreement with Glen’s results, certainly the shape, though the swr figures are a little low in places. The performance also rolls off before 20mhz, probably due to construction or the fact I didn’t fully raise the 7m squid pole. I can live with this for the moment as I’m mostly interested in where the dips are.
Not bad for a few evenings work and quite a useful tool. I have a second very similar project to build sometime which is a scalar network analyser.