Circuit above. Please excuse the skewed image - smartphone camera.
Looks like I picked the original FET specified and didn't notice. First thing I did was to scribble the circuit out on paper and toss the book back on the shelf.
I checked leakage through the zener on the breadboard (with a marginally superior HP3478A
) and the FSD voltage at the FET is 1v which is way above the worrisome leakage point so it doesn't affect measurement as far as I can see. I think it resulted in a 0.005% error at FSD which is "meh" territory for a class 2.5 moving coil meter movement. I blew the front end FET up twice on the breadboard (at 100v in 1v range) and this stopped it from blowing up so I'm not 100% convinced it isn't necessary. This might be a hooky batch of 2n3819's though; they were from China off ebay
There's a version two being planned at the moment which has a FET op-amp in it (nice TI TLC27x series), ohms, current and RF power into 50 ohms ranges. It may end up with a power supply and frequency counter as well. I start with big ideas and remove the unreasonable bits