jfet drive radio noise

[seedlings]

This circuit is on the breadboard.  The last build with only three jfet stages haunts me with faint radio noise, so I'd like any suggestions to eliminate the problem before a 5 stage circuit goes under foot.  Right now on the breadboard, the radio noise can be heard occasionally.  There is also a 2k2 resistor and a 220uF cap RC filter on the power before going to all the fets.

Thanks,
CHAD

[Jdansti]

Try shielding it.  First wrap it with an insulator such as newspaper and then go over that with a sheet of aluminum foil connected to ground with an alligator clamp. Just make a big burrito out of it. You probably don't need to have perfect 100% shielding. It would be a shame to chunk something that might work in a shielded enclosure.

[psychedelicfish]

I'd try an RC filter on the input, 33k, 220p a LA fetzer valve. JFETs have a high input impedance, meaning that they make good RF amplifiers.

[tca]

Too many gain stages! I would remove the 2nd and the 3rd, and change the first bypass capacitor to 10u (or keep t 22u) and the second bypass cap to 1u.

Or add a 1k to power and 100u to ground at each stage.

Do you really need all that? That should sound fizzy.

Cheers.

[seedlings]

Too many gain stages?  Heresy!   You're probably right, though, except it sounds great.  Throaty.  Working on each gain stage and playing with the oscilloscope.  Probably add a tone stack.  Most of my breadboard experiments don't pan out... but you never know.

I will try the RC input filter later.

CHAD

[R.G.]

You see where there is a 10k in series with the last few JFET gates? Bypass each gate to ground with a ceramic cap of 100pF to 470pF. This should not affect anything in the audio band, but will kill the RF gain.

Even this will not help if the power supplies and grounds are not clean. You need at least a 10uF from V+ to ground bypassed by a 0.1uF and possibly another 0.01uF ceramic capacitor(s). With as much gain as you have stacked up there, you really need to go to RF layout practices, because the tiniest bit of stray capacitance from a drain to an earlier gate can make it oscillate and become a heterodyne radio tuner. At RF, the inductance of the ground wires matters, too, and a bussed ground can be a feedback network unintentionally.

If you keep increasing the gain, everything will eventually oscillate.

[seedlings]

The RC filter on the input seems to do the trick.  These j201s are from tayda, and I know they're leftovers and generally underperforming chips.  Each fet did need varying resistance values to bias for symmetrical clipping.  They seem extra finicky, and he oscilloscope leads introduce noise.

CHAD

*edit- I'm not necessarily going for monster gain, which is what the 100k trim pots are for.  I'm trying to see how the clipping looks, changes and interacts from stage to stage.  Plus I'm terrible at layouts.  It will probably look a lot like the schematic if a board is ever etched.