Actually put this together some years ago, and buttoned it away before I could take a picture...
Three lines, from a 240VCT isolation transformer; the three inductors of varying size at the left end, ballast the transformer's leakage (biggest for neutral, a direct connection obviously, the others for H1/H2 -- evidently at different depths in the windup, hence the differing leakage), giving close to 250uH I think it was, and this is terminated by, not 4.7 ohms for some reason, but 7.8 actually (hey, I have a ton of those resistors..?), in series with 5.6uF. Then the traditional 50uH chokes to the output, and coupling caps and highpass filters to the RF ports.
Chokes are designed for soft saturation around 30A or so I think, so, good enough for the 15A RMS rating of the transformer it's tuned for, and won't be quite right at the peaks of a full load FWB, but at lower power, or higher PF, it's good. Anyway, that only affects LF response, so even if switching ripple might read in error in that case, the higher stuff is fine.
Given the stub lengths and terminal strip construction,
Oh yeah a construction detail about that terminal strip in the middle, that's copper clad (I don't have any old-school strips suitable for 14AWG solid wire), it's 2oz double sided, with a strip along the base to make a nice wide solder fillet to the bottom panel, then separate squares along the top, with holes drilled through to clinch the wires on.
No, no plots or anything quantitative; no fancy limiters. It's pretty basic, it'll work well over 30MHz which is good enough for me. I suppose stub lengths will start pulling in by a few hundred MHz, as well line-to-line crosstalk; but anything I'm going to wire up to it, is already going to be dubious by such frequencies, so yeah, no worry. Also based on my experience with a different network, which I did measure up to fairly high frequencies (from which, I'd expect maybe ~20dB crosstalk between adjacent taps).
Tim
Home
Help
Search
Login
Register
