Sometimes when I see a schematic that switches an audio signal it uses some sort of transistor with the Source to Drain in the signal path like this:


I figure that it would act as a failsafe to ground the signal instead in case there is a failure the audio would continue. Something like this:


Having not seen this, I am wondering if I'm barking up the wrong tree. Is this inappropriate use of the JFET? Regardless, I can't seem to make a breadboarded circuit work with this idea.

This message has been edited. Last edited by: mark123,

A couple of things.

First, it looks from the datasheet like the source-drain resistance of the J175 JFET when "on" is in the hundreds of ohms. That would be part of why there is a buffer before and after the JFET. But with the buffers there, using the JFET to ground as in the second circuit won't kill the audio. The low output impedance of the first buffer is capable of driving its output voltage equal to its input voltage despite the few hundred ohm connection to ground.

Second, even if you use a different transistor with a very low "on" resistance, that will put a lot of electrical load on the first buffer - the output voltage gets dropped by pushing a lot of current to ground. You risk burning out the buffer. You would really want to put an output resistor after the buffer but before the transistor. This would decrease the load on the first buffer and let the transistor drop the output voltage further.

Last, in the first circuit, when the JFET is "on," audio goes through the circuit. In the second, when the JFET is "on," audio does not go through the circuit (I mean, ideally, in terms of what you want to have happen). You would need to reverse the on/off gate voltages for the second circuit vs. the first one.

Ah, I see now why it's used this way. I didn't consider the load on the input buffer. Thanks.

I have the on/off signal coming from the output of a comparator so I can switch its output by swapping the inputs depending on what's required.

Glad I could help!

I'm not an EE but I took a few of the classes and I've built enough audio and guitar equipment to fake it on small circuits.

The P-FET turns on (current flows from source to drain) when the gate voltage is (edit) LESS than VA in the first circuit.

In the 2nd circuit shorting VA to ground doesn’t appear like a good thing to do. Since the input impedance into the amplifier to the right is very high, the FETs appear to be performing some type of clipping function. Without seeing the rest of the circuit it’s hard to tell what the context or intent of the circuit it. I would guess it’s doing some kind of limiting / clipping function.

What is connected to the signal off the bottom of each circuit?

seems like there are better jfets than the J175. Try maybe 2N5485?? I've been out of the game for a few decades, but they keep improving them.

For the first circuit, you can feed the DC through a high value resistor like a few hundred k ohms. Might try both sides for the DC feed, each with a resistor.

For the 2nd circuit, yup, the first stage may not want to drive into the ground.

quote:

Originally posted by Jimineer:
The P-FET turns on (current flows from source to drain) when the gate voltage is (edit) LESS than VA in the first circuit.

In the 2nd circuit shorting VA to ground doesn’t appear like a good thing to do. Since the input impedance into the amplifier to the right is very high, the FETs appear to be performing some type of clipping function. Without seeing the rest of the circuit it’s hard to tell what the context or intent of the circuit it. I would guess it’s doing some kind of limiting / clipping function.

What is connected to the signal off the bottom of each circuit?

It's trying to be a noise gate. The other part of the circuit is a buffer into an opamp set for a bit of gain, into a peak detector, into a comparator and that feeds the gate of the JFET.

quote:

Originally posted by mark123:

quote:

Originally posted by Jimineer:
The P-FET turns on (current flows from source to drain) when the gate voltage is (edit) LESS than VA in the first circuit.

In the 2nd circuit shorting VA to ground doesn’t appear like a good thing to do. Since the input impedance into the amplifier to the right is very high, the FETs appear to be performing some type of clipping function. Without seeing the rest of the circuit it’s hard to tell what the context or intent of the circuit it. I would guess it’s doing some kind of limiting / clipping function.

What is connected to the signal off the bottom of each circuit?

It's trying to be a noise gate. The other part of the circuit is a buffer into an opamp set for a bit of gain, into a peak detector, into a comparator and that feeds the gate of the JFET.

A noise gate? So like for a low or no audio input sigal present, do you want to cut off the circuit at the input to the 2nd amp? E.g., preventing a dc hum on some audio output?

Actually, maybe your 2nd circuit was on the right track then. But maybe you need some smallish value of resistance to limit the current from the output of the first amp for when the JFET turns on. but I don’t know for sure without looking at the details. What is VA in the circuit? You need to make sure you don’t put 15v or whatever on the source of your FET. I guess a resistor would also help limit that.

Maybe you were on the right path after all!

quote:

Originally posted by Jimineer:
A noise gate? So like for a low or no audio input sigal present, do you want to cut off the circuit at the input to the 2nd amp? E.g., preventing a dc hum on some audio output?

Yep, exactly.

quote:

Originally posted by Jimineer:
Actually, maybe your 2nd circuit was on the right track then. But maybe you need some smallish value of resistance to limit the current from the output of the first amp for when the JFET turns on. but I don’t know for sure without looking at the details. What is VA in the circuit? You need to make sure you don’t put 15v or whatever on the source of your FET. I guess a resistor would also help limit that.

Maybe you were on the right path after all!

VA is 9v. I have a variable 3v-6v available at VB and can add any other reference voltage up to 9v.
Your thought process is 100% match to what I was thinking. There are other circuits out there that all have the audio going through the JFET like the first example. I thought the second example would be a improvement for audio quality.

quote:

Originally posted by radioman:
seems like there are better jfets than the J175. Try maybe 2N5485?? I've been out of the game for a few decades, but they keep improving them.

For the first circuit, you can feed the DC through a high value resistor like a few hundred k ohms. Might try both sides for the DC feed, each with a resistor.

For the 2nd circuit, yup, the first stage may not want to drive into the ground.

The J175 was chosen for a very specific reason. That reason is I have a bunch here.

quote:

Originally posted by mark123:

quote:

Originally posted by radioman:
seems like there are better jfets than the J175. Try maybe 2N5485?? I've been out of the game for a few decades, but they keep improving them.

For the first circuit, you can feed the DC through a high value resistor like a few hundred k ohms. Might try both sides for the DC feed, each with a resistor.

For the 2nd circuit, yup, the first stage may not want to drive into the ground.

The J175 was chosen for a very specific reason. That reason is I have a bunch here.

Your schematic symbol for the FET is a MOSFET and not a JFET if that matters. I wonder if a MOSFET would do better instead though. It would certainly have a lower Rds on resistance.

I’m curious why you need VA where you have it though. That might be your problem but I don’t know what type op amps you have in there. Wouldn’t a bias resistor be needed? When the FET turns on VA is the voltage seen at the input of the 2nd buffer I think.

^^^ C6 and C7 block the DC while allowing the audio to pass.

quote:

Originally posted by Jimineer:

quote:

Originally posted by mark123:

quote:

Originally posted by radioman:
seems like there are better jfets than the J175. Try maybe 2N5485?? I've been out of the game for a few decades, but they keep improving them.

For the first circuit, you can feed the DC through a high value resistor like a few hundred k ohms. Might try both sides for the DC feed, each with a resistor.

For the 2nd circuit, yup, the first stage may not want to drive into the ground.

The J175 was chosen for a very specific reason. That reason is I have a bunch here.

Your schematic symbol for the FET is a MOSFET and not a JFET if that matters. I wonder if a MOSFET would do better instead though. It would certainly have a lower Rds on resistance.

I’m curious why you need VA where you have it though. That might be your problem but I don’t know what type op amps you have in there. Wouldn’t a bias resistor be needed? When the FET turns on VA is the voltage seen at the input of the 2nd buffer I think.

Yeah, it's a MOSFET symbol. For some reason my library has a mosfet symbol for the J175. I'll have to fix that. I was wondering if a MOSFET would make a better choice.

quote:

Originally posted by mark123:

quote:

Originally posted by Jimineer:

quote:

Originally posted by mark123:

quote:

Originally posted by radioman:
seems like there are better jfets than the J175. Try maybe 2N5485?? I've been out of the game for a few decades, but they keep improving them.

For the first circuit, you can feed the DC through a high value resistor like a few hundred k ohms. Might try both sides for the DC feed, each with a resistor.

For the 2nd circuit, yup, the first stage may not want to drive into the ground.

The J175 was chosen for a very specific reason. That reason is I have a bunch here.

Your schematic symbol for the FET is a MOSFET and not a JFET if that matters. I wonder if a MOSFET would do better instead though. It would certainly have a lower Rds on resistance.

I’m curious why you need VA where you have it though. That might be your problem but I don’t know what type op amps you have in there. Wouldn’t a bias resistor be needed? When the FET turns on VA is the voltage seen at the input of the 2nd buffer I think.

Yeah, it's a MOSFET symbol. For some reason my library has a mosfet symbol for the J175. I'll have to fix that. I was wondering if a MOSFET would make a better choice.

You might want to compare the drain to source capacitance of each type FET to make sure you don’t degrade the low end frequency response with a high parasitic capacitancee. Without PSpice I can’t do this stuff in my head anymore (esp. the small signal AC analysis).
Did I mention just replacing the FET in the first circuit with two parallel diodes in opposite directions?

quote:

Originally posted by Jimineer:
... Did I mention just replacing the FET in the first circuit with two parallel diodes in opposite directions?

I don't want to clip the signal, I want to silence it when the signal is below a threshold. When it's not silenced I'd prefer it be high quality audio.

quote:

Originally posted by radioman:
^^^ C6 and C7 block the DC while allowing the audio to pass.

Good point radioman. I neglected C7.
C7 might pass some power supply noise from VA - but I assume that VA is well filtered.

Also, you might hear a pop at the output when switching the FET on or off but I'm not familiar with the signal levels.

I'm also not familiar with audio applications so this is a bit of nostalgic circuit design stuff for me - which isn't helping mark123 at all.

It looks like he's got a good handle on.

quote:

Originally posted by Jimineer:

quote:

Originally posted by radioman:
^^^ C6 and C7 block the DC while allowing the audio to pass.

Good point radioman. I neglected C7.
C7 might pass some power supply noise from VA - but I assume that VA is well filtered.

Also, you might hear a pop at the output when switching the FET on or off but I'm not familiar with the signal levels.

I'm also not familiar with audio applications so this is a bit of nostalgic circuit design stuff for me - which isn't helping mark123 at all.

It looks like he's got a good handle on.

It does help and I thank you. I was in electronics 31 years ago and now it's just a hobby. I enjoy building my own stuff rather than buying.

I'm hoping the buffer takes care of the switch pop.

Hmmm. Does this make any more sense?

quote:

Originally posted by mark123:
Hmmm. Does this make any more sense?

Mmmmmaybe. It looks like it should switch the signal to ground but I’m not sure what these components will do to your audio performance. Wouldn’t C6 be better to move closer to IC3 on the other side of where C7 ties in?

quote:

Originally posted by Jimineer:

quote:

Originally posted by mark123:
Hmmm. Does this make any more sense?

Mmmmmaybe. It looks like it should switch the signal to ground but I’m not sure what these components will do to your audio performance. Wouldn’t C6 be better to move closer to IC3 on the other side of where C7 ties in?

Probably. When the switch is on leaving it there the capacitance will drop to 500nF but I'm not sure it matters since it's effectively shorting to ground. It does make sense to move C6. Ultimately, that's a question for the breadboard. I need a bigger breadboard.