JFET Switching and Isolation, question

[wizard333]

In the Fender Hot Rod circuit, I'm changing the function of the J111 JFET labelled Q1 to switch the cap at C3 rather than C8 (removed C3, C8, ran a cap from the pos of C3 to neg of C8).

While the 100k resistor R23 between source and drain is more than sufficient to isolate a Ck, it is wholly insufficient to isolate a cap at the C3 position. The problem I'm encountering is, if I change R23 to 1M to isolate the cap when "more gain" isn't on, a loud white noise is introduced into the circuit when "more gain" isn't on (i.e., when the JFET N channel is not conducting).

Anyone understand why that would be happening or how to get around it?

[loudthud]

Link to Hot Rod Deville schematic: http://www.el34world.com/charts/Sche...od_deville.pdf

The signal at C3 is probably too big for a JFET to handle. That's why Fender used a relay.

[wizard333]

In what way? It seems to handle it fine with 100k between source and drain, it's just that 100k is not sufficient isolation for the "less gain" channel.

Also, the AC signal there isn't very big, and there is no DC, especially the part of the present AC that gets through a 150pf cap, which is what I'm using.

Conversely, the signal at C8 would be over 1V dc, never mind the AC, and the JFET handles that fine.

[Enzo]

Nonetheless, Fender chose to ground THIS cap with a relay as opposed to the others grounded with a JFET. My advice is "never think up reasons not to check something." I suspect the JFET myself, I could be wrong. SO remove the JFET. Just run the resistor to ground. Is your white noise still present? The JFET may be OFF, but that isn't the same as not being there. You are in a high gain situation at that grid, and the diode in the JFET gate circuit could well be involved. The JFET off is equivalent to it not being there in terms of the cap. SO by removing the JFET, we find that either the 1 meg alone will cause the noise, or we find that the 1 meg itself is not the issue, but the presence of the JFET in that grid circuit is. Clearly a relay plus 1 meg does not produce this noise, but leaving the resistor but replacing relay with JFET results in the noise.


And just to be nit-picky, we are not isolating the cap, the resistor is there to keep it charged.

[wizard333]

By isolation I mean putting a large enough resistor there that the cap does not audibly affect the low gain channel. I know why the resistor is there. 1M allows the cap to charge but shouldn't audibly affect the circuit when it isn't bypassed.

Are you thinking the JFET might be bad? I can't think of a reason why 1M instead of 100k there would cause the white noise to occur.

How would the diode in the gate circuit be involved?

Any suggestion for a workaround?

[Enzo]

I wouldn;t want to design a workaround until we knew just where the problem lies. I'd hate to spend an hour cooking something up only to find your JFET was bad.

Did the JFET work in its original job? If so, probably OK. Or swap it for the other one. Chances are both will not have the same problem if one is defective.


I can't think of a reason the 1 meg could be involved either, but the experiment will prove either way, then we will know if adding the JFET causes the problem. We will also find out that the layout is not involved. What if the placement of your new here to there cap is causing some sort of noise and the JFET is not involved? If that is the case, the noise would still be present without the JFET. It may not be the resistor or the JFET, it might be the new cap or even just its placement.

Diodes or transistors wired as diodes are often used as noise sources. I can't swear that is your situation, but with the 1 meg there and the JFET off, it is not unlike connecting a diode from the cap ( and thus a direct couple to the grid for high freqs) over to the switching circuit. Cathode circuits would not be sensitive to that.

[wizard333]

Looks like the J111 was bad. It was switching correctly at first although being noisy, but eventually started to show a resistance of only about 1.7k even when "off". Replaced it and things are working correctly for now.

I guess those J111s are touchy; I'm holding my breath when I measure resistance now as I'm concerned the VOM's test voltage will zap it. Solder heat from changing R23 probably got it going before, even though I'm using minimal heat and not leaving it there long.

Hope it keeps working as intended, I really didn't want to make a relay board just to switch that one cap.

[Steve Conner]

Was it bad, or did you blow it with excessive signal voltage?

[wizard333]

Explain to me how there would be more signal voltage at C3 through a 150pf cap than C8 through a 1uf cap?

The cathode should be a near unity gain with the signal at the grid in terms of AC, and the 1uf cap will let more signal through. Even if the cap were shorted, there is no DC at C3, but well over a volt of DC at C8. This signal is also post 2 gain stages with the attenuation of a volume or gain control and a tone stack. Not what I'd call a lot of AC.

Neither position should provide anywhere near enough voltage to blow that JFET, but if either were, it's original application and position would be the one to do it.

Or am I missing something?

[loudthud]

The signal at C8 can maybe go to +2.5V under extreme overdrive. It will never go below ground. The signal at C3 can go about the same +2.5V, the voltage is clamped by the tube grid. However the signal at C3 can easily go 100V negative, there is nothing to stop it. There are two stages of gain between the guitar and C3 so depending on how the level controls are set the gain is over 1000 at some frequencies

[wizard333]

Under what condition could it go that negative?