more JFET switching leakage discoveries - I'm getting there

[tempus]

Hey all;

Last week I posted this schematic:



I found that it was pretty quiet when switching, but that there was some leakage, i.e., there was still signal getting through even though it was switched off. RG suggested that this was probably due to the fact that the V_gs(off) was as high as 8v for and MPF102, and that it was likely that it wasn't switching completely off. I was trying some different things with it today, and discovered that when I pulled the JFET out of the circuit, the leakage was still there (picture the above schematic with no JFET in it). The leakage must be the result of some interplay of the 1M resistors or the voltage divider circuit (not pictured) that's giving me the 4.5v drain/source bias. Any ideas on how to get rid of the leakage?

Thanks

[Eb7+9]

you need jFETs with Vgs(off) less than 4.5volts ... like the 2SK30, J201, 2n5484 ...

[earthtonesaudio]

The leakage must be the result of some interplay of the 1M resistors or the voltage divider circuit (not pictured) that's giving me the 4.5v drain/source bias.

Make the voltage reference lower impedance, either by buffering it or adding a big capacitor.  I like J113's for switching.