can anyone help me with jfets?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

salomonander

Well-known member
Joined
Aug 28, 2011
Messages
913
hey guys

im going mad trying to use a jfet for audio muting. it took me a while to understand how these fets work and in my led test everything seems fine.
i insert ground to the source and an LED between drain and 5v (with resistor) and the led lights up. as soon as i put -5v on the gate the LED goes off.... so i guess this is all fine.

but it wont work with audio at all. no matter how i move the pins around i cant seem to make the audio shunt to ground. i tried connecting audio in and out to the source and both grounds to the drain (tried it the other way as well). shouldn't it mute in that case (connect audio go to ground)?
it doesnt do anything. when connecting audio in to source and audio out to the drain i have a constant signal. i can connect -5v to the gate but still have no mute. i dont get it  :(
i even tried to copy the studer 169 mute circuit but still no luck. im using bf245a. is that the problem?
 
My only experience is with J111's, and there, you have to pull the gate much much lower than the signal.

There is a great chapter in D.Selfs book on it.

/R
 
basically i tried everything down to a simple thing like the attached schematic.
shouldn't the signal be interrupted when applying -5v to the gate? it doesnt. it does it when using an led for testing though....

 

Attachments

  • jfet.JPG
    jfet.JPG
    5.3 KB · Views: 40
i also tried this - assumin that the signal would be muted if no -5v were applied to the gate (ground to signal). same result - i have a signal all the time. no way to interrupt the audio.
 

Attachments

  • jfet2.JPG
    jfet2.JPG
    9.3 KB · Views: 34
Rochey said:
My only experience is with J111's, and there, you have to pull the gate much much lower than the signal.

There is a great chapter in D.Selfs book on it.

/R

i even tried a schematic from the book. didn't work either. i guess either my cabling is bad or my fets or im brain damaged :)
 
That schematic doesn't help me much.

I have used many JFET mutes over the years and they are relatively simple. I have even inherited one deign that used a transistor output mute (old school and cheaper).

To begin understand how JFETs work. There is a conductive channel, that can be turned off to not conduct by the charge from biasing the gate voltage above or below the source voltage (baed on polarity of JFET).

I am more familiar with using the JFET as a shunt to ground when turned on, so a fixed input resistor forms a voltage divider with the variable resistance JFET to ground. When turned off the JFET can be very high resistance, for almost no attenuation and when turned on (0v gate to source), the JFET can be as low as hundreds of ohms for lots of attenuation.

Alternately the JFET can be used like a switch in series, while it is a little more difficult to manage the gate voltage relative to a moving source voltage.  I suspect a JFET in series with an inverting opamp -input could operate with a fixed source voltage, but I would worry about other issues (like clicks). .

So a simple JFET shunt mute would involve a fixed input resistor (say 10k), a JFET with drain connected to the input R and the source connected to ground.  The junction of the input R and JFET probably should be buffered by an opamp. Finally the gate is driven between 0V for on=mute, or past cut off voltage for off=not mute. Wether the gate voltage is positive or negative depends on the Nch or Pch sex of the JFET.

JR

 
JohnRoberts said:
I have used many JFET mutes over the years and they are relatively simple. I have even inherited one deign that used a transistor output mute (old school and cheaper).

I've seen this recently -- using a BJT penny transistor (MMBT2222A) on the output of a line driver. (post series resistor).

Nasty nasty nasty.
It was for a $29 set top box though :)
 
Rochey said:
JohnRoberts said:
I have used many JFET mutes over the years and they are relatively simple. I have even inherited one deign that used a transistor output mute (old school and cheaper).

I've seen this recently -- using a BJT penny transistor (MMBT2222A) on the output of a line driver. (post series resistor).

Nasty nasty nasty.
It was for a $29 set top box though :)

IIRC the bipolar mute used a special part number transistor that was optimized for the application (perhaps higher base zener voltage?, Vsat, etc).  Another old school trick, is you can make a bipolar transistor work upside down or with emitter and collector swapped, so the base-collector junction works like a base-emitter but with a lower saturation voltage for less DC in a DC coupled audio mute situation, while these are still cheap and inferior to JFET mutes where there is no base or gate current to corrupt the audio DC.

Sorry for the veer...  Seeing a bipolar mute, either means an old design engineer, or a young engineer copied an old design.  8)

JR

 
JohnRoberts said:
That schematic doesn't help me much.

I have used many JFET mutes over the years and they are relatively simple. I have even inherited one deign that used a transistor output mute (old school and cheaper).

To begin understand how JFETs work. There is a conductive channel, that can be turned off to not conduct by the charge from biasing the gate voltage above or below the source voltage (baed on polarity of JFET).

I am more familiar with using the JFET as a shunt to ground when turned on, so a fixed input resistor forms a voltage divider with the variable resistance JFET to ground. When turned off the JFET can be very high resistance, for almost no attenuation and when turned on (0v gate to source), the JFET can be as low as hundreds of ohms for lots of attenuation.

Alternately the JFET can be used like a switch in series, while it is a little more difficult to manage the gate voltage relative to a moving source voltage.  I suspect a JFET in series with an inverting opamp -input could operate with a fixed source voltage, but I would worry about other issues (like clicks). .

So a simple JFET shunt mute would involve a fixed input resistor (say 10k), a JFET with drain connected to the input R and the source connected to ground.  The junction of the input R and JFET probably should be buffered by an opamp. Finally the gate is driven between 0V for on=mute, or past cut off voltage for off=not mute. Wether the gate voltage is positive or negative depends on the Nch or Pch sex of the JFET.

JR

thanks all of you for taking the time!
this is pretty complicated for me but im trying to understand.
so far i though that if 0v  is applied to the gate the jfet is conducting like a piece of wire.
but i guess this isn't the case. am i right in that:

- i have to keep the gate at 0v in regard to the source. not at 0v in general?
- so if i want to run a signal through it i have to find a way to keep the gate at 0v in relation to the audio signal?
- when using a jfet to shunt the signal to ground i dont have to worry but simply apply 0v?

in case this is right i trying to understand  the studer 169 mute schematic. is C50 used to regulate Q8 in regard to the source? or how do they go about biasing the gate properly?

sorry if im a pain in the ass... i just trying to learn. thanks so much!
 

Attachments

  • 169mute1.png
    169mute1.png
    81.5 KB · Views: 77
salomonander said:
this is pretty complicated for me but im trying to understand.
so far i though that if 0v  is applied to the gate the jfet is conducting like a piece of wire.
Yes pretty much, but not exactly wire... RDS-on can be low tens of ohms.
but i guess this isn't the case. am i right in that:
Nothing is completely simple !dss is current max but not generally an issue.
- i have to keep the gate at 0v in regard to the source. not at 0v in general?
yes exactly
- so if i want to run a signal through it i have to find a way to keep the gate at 0v in relation to the audio signal?
Yes, typical way is to connect a resistor from gate to source, then drive the gate with a current source.  Zero current means Vgs=0V, applied current means Vgs= I x R
- when using a jfet to shunt the signal to ground i dont have to worry but simply apply 0v?
yes, pretty much... there are tricks to linearize it for lower distortion during trasition, but for now lets KISS.
in case this is right i trying to understand  the studer 169 mute schematic. is C50 used to regulate Q8 in regard to the source? or how do they go about biasing the gate properly?

sorry if im a pain in the ass... i just trying to learn. thanks so much!

Yes, I get a headache from looking at other engineers inferior designs.  8) 8), From a quick glance it looks like a series-shunt, so first JFET is being opened up, while second is being made to conduct, and vice versa... .

JR
 
JohnRoberts said:
salomonander said:
this is pretty complicated for me but im trying to understand.
so far i though that if 0v  is applied to the gate the jfet is conducting like a piece of wire.
Yes pretty much, but not exactly wire... RDS-on can be low tens of ohms.
but i guess this isn't the case. am i right in that:
Nothing is completely simple !dss is current max but not generally an issue.
- i have to keep the gate at 0v in regard to the source. not at 0v in general?
yes exactly
- so if i want to run a signal through it i have to find a way to keep the gate at 0v in relation to the audio signal?
Yes, typical way is to connect a resistor from gate to source, then drive the gate with a current source.  Zero current means Vgs=0V, applied current means Vgs= I x R
- when using a jfet to shunt the signal to ground i dont have to worry but simply apply 0v?
yes, pretty much... there are tricks to linearize it for lower distortion during trasition, but for now lets KISS.
in case this is right i trying to understand  the studer 169 mute schematic. is C50 used to regulate Q8 in regard to the source? or how do they go about biasing the gate properly?

sorry if im a pain in the ass... i just trying to learn. thanks so much!

Yes, I get a headache from looking at other engineers inferior designs.  8) 8), From a quick glance it looks like a series-shunt, so first JFET is being opened up, while second is being made to conduct, and vice versa... .

JR

thanks robert
ill try to get some j112s tomorrow and get going.
ps: if you know of a better schematic let me know!
 
Most JFET muting circuits in the SSLs use 2 "switch" circuits, one to "disconnect" the source from the destination, and one to ground the input to the destination.
 
does anyone maybe have a schematic for the ssl system? anything helps me to understand this

i just tried to emulate the system explained in "small signal audio" but im having no luck. it works at first but then the signal slowly creaps back after a while http://www.falstad.com/circuit/#%24+1+5.0E-6+10.20027730826997+50+5.0+50%0AR+224+112+160+112+0+1+40.0+5.0+0.0+0.0+0.5%0Ar+224+144+304+144+0+22.0%0Aw+224+112+224+144+0%0Aw+224+112+304+112+0%0Aj+320+144+320+112+0+-4.0%0Aw+320+144+304+144+0%0Ar+400+112+400+176+0+10000.0%0Ag+400+176+400+240+0%0Aw+336+112+400+112+0%0AO+400+112+512+112+0%0Aw+320+144+320+240+0%0Ad+320+240+224+240+1+0.805904783%0Ar+224+240+224+192+0+220000.0%0AR+224+192+160+192+0+0+40.0+17.0+0.0+0.0+0.5%0As+224+240+224+304+0+0+false%0AR+224+336+224+352+0+0+40.0+-23.0+0.0+0.0+0.5%0Aw+224+336+224+304+0%0Ao+9+64+0+34+0.0048828125+9.765625E-5+0+-1%0A
 
If you want to shunt your audio to ground using the FET, don't forget about the Drain to Source resistance. You will not get a 100% off state with only one FET using this shunt method. Look at the datasheets for the FET's you are talking about. For example, the D->S resistance of a J112 is 50R. It will be the same as a 50 ohm R shunt to ground. You have to also be aware of the preceding device and the drive capability. Can the preceding device drive a direct short to ground? Maybe some series resistance before the FET is required.

Many times, the audio would run thru the FET used for muting. Look up some of the the D&R schemo's in the tech docs area here. I think they used J111's or J112"s with the audio running thru them.
 
simply pull the fader down or rip out the 1/4 inch or xlr going into the strip,

that's what us po folks do,  ???

or use a push button switch, jus sayin...
 
hi solomonander,

I have built this fet mute for my mixer project, using J112 fets. I use a fet in parallel to the input (shunt to ground Q2) and one in series (Q1). one fet is not enough to get good rejection, you need to to get  around -100dB. . they need to be operated inverted. when Q1 is passing audio the shunt fet is 'open'. when the mute is active the series fet is 'open', not passing the signal and the shunt fet is bringing the voltage divider (R5, Q2) to ground (almost at least....)

have a look at my schematic here:
http://www.groupdiy.com/index.php?topic=48120.msg609789#msg609789

two things are important: the inverting opamp (U3) brings the series fet Q1 to 'virtual ground', so that it conducts when the gate is at 0v and Q2 must be operated inverted compared to Q1. I have the fet gates on a little 2 pin header so that I can easily swap the drive voltages.

I hope this helps a little, otherwise feel free to ask for more details.

- michael
 
salomonander said:
basically i tried everything down to a simple thing like the attached schematic.
shouldn't the signal be interrupted when applying -5v to the gate? it doesnt. it does it when using an led for testing though....
I think you've just forgotten a very basic thing. In order to attenuate sufficiently the signal, you wanr to have a series resistance that is much higher than the Rdson; I see you have no significant series resistance here. You also want the load impedance to be much higher than the series resistance (unless you operate in current-mode, but I think it is confusing to you).
 
Back
Top