A gentleman and frequent contributor to this forum has kindly granted me permission to discuss a circuit he developed that uses JFETs to select/switch between audio sources.
The circuit (seen below) only shows one audio source, but the concept can be extended to multiple audio sources by replicating the portion of the circuit within the box marked "Repeat as required".
Using this concept with multiple input sources creates a virtual earth mixer, yet I don't think it was the really intended to be a mixer, but rather just a source selector.
My question is, what would happen if more than one source was selected? Would their be smoke or just the risk of some odd sounding music as the result of the mixing?
The circuit (seen below) only shows one audio source, but the concept can be extended to multiple audio sources by replicating the portion of the circuit within the box marked "Repeat as required".
Using this concept with multiple input sources creates a virtual earth mixer, yet I don't think it was the really intended to be a mixer, but rather just a source selector.
My question is, what would happen if more than one source was selected? Would their be smoke or just the risk of some odd sounding music as the result of the mixing?
No smoke, I think.
This is a complicated circuit with 2 control voltages and an inevitable nonlinearity in the 'ON' jfet, however small. Plus there'll be some added noise. It's not really a good circuit. I'd much rather see a relay used. Primitive and electromechanical as they may be, they do have their advantages.
w
This is a complicated circuit with 2 control voltages and an inevitable nonlinearity in the 'ON' jfet, however small. Plus there'll be some added noise. It's not really a good circuit. I'd much rather see a relay used. Primitive and electromechanical as they may be, they do have their advantages.
w
A gentleman and frequent contributor to this forum has kindly granted me permission to discuss a circuit he developed that uses JFETs to select/switch between audio sources.
The circuit (seen below) only shows one audio source, but the concept can be extended to multiple audio sources by replicating the portion of the circuit within the box marked "Repeat as required".
Using this concept with multiple input sources creates a virtual earth mixer, yet I don't think it was the really intended to be a mixer, but rather just a source selector.
My question is, what would happen if more than one source was selected? Would their be smoke or just the risk of some odd sounding music as the result of the mixing?
Easy one first... there would be no smoke if more than one input were selected. The logic I use doesn't allow that though, but if it did it would just act as a mixer.
You might like to read more on it here starting at post #184 and particularly post 201 which explains it a little more and covers the main reasons why I used this method.
http://www.diyaudio.com/forums/soli...-amplifier-designed-music-10.html#post1560313
If I were doing it again I would this for the PIC controller I think,
FPRC5RX - DIY learning IR decoder
2nd time today I posted that link... it's a brilliant bit of programming and does exactly what the designer says. My PIC skills are pretty much zero I'm afraid but I programmed a PIC using the code in the above link and it's brilliant.
FPRC5RX - DIY learning IR decoder
2nd time today I posted that link... it's a brilliant bit of programming and does exactly what the designer says. My PIC skills are pretty much zero I'm afraid but I programmed a PIC using the code in the above link and it's brilliant.
Using switches is easy. Just thinking out loud so if I make a daft mistake or comment lol...
I take it your using a split -/+ supply so set up two rails from the main opamp supply of say 8 volts and if possible another of around -0.3 volts. The 8 volt can come from a zener... the current needed is essentially zero and the -0.3 from a germanium diode (as I used) although I can't see why just resistive dividers couldn't be used if the supplies to the opamp are stable. Tie the gates to the -0.3v via say 2.2meg resistors... that put's both FET's off. Then use the switch to overide that by feeding the 8 volts via say 220k
I take it your using a split -/+ supply so set up two rails from the main opamp supply of say 8 volts and if possible another of around -0.3 volts. The 8 volt can come from a zener... the current needed is essentially zero and the -0.3 from a germanium diode (as I used) although I can't see why just resistive dividers couldn't be used if the supplies to the opamp are stable. Tie the gates to the -0.3v via say 2.2meg resistors... that put's both FET's off. Then use the switch to overide that by feeding the 8 volts via say 220k
Last edited:
You might find the JFET a bit tricky to source, its a Sony part. Doug Self's book 'Small signal audio design' has a whole chapter dedicated to signal switching circuits and goes into some depth with JFET switching arrangements. He prefers to use the J112 which is very widely available.
Thanks abraxalito,
It's nice to have choices!
I'll also say that the more I try to learn about JFET's the more I get confused. For example, here are specs for several popular n-channel JFET's that might be candidates in this application:
Datasheet for 2N3819: Vgs (off) = minimum not stated, maximum +8.0 Volts
Datasheet for BF245C: Vgs (off) = minimum -0.5 Volts, maximum -8.0 Volts
Datasheet for J112: Vgs (off) = minimum -1.0 Volts, maximum -5.0 Volts
So why do some of these have a positive Vgss (off) value and others are negative?
It's nice to have choices!
I'll also say that the more I try to learn about JFET's the more I get confused. For example, here are specs for several popular n-channel JFET's that might be candidates in this application:
Datasheet for 2N3819: Vgs (off) = minimum not stated, maximum +8.0 Volts
Datasheet for BF245C: Vgs (off) = minimum -0.5 Volts, maximum -8.0 Volts
Datasheet for J112: Vgs (off) = minimum -1.0 Volts, maximum -5.0 Volts
So why do some of these have a positive Vgss (off) value and others are negative?
There's a very simple answer to this question 
Datasheets sometimes contain errors - you've found one in the Fairchild datasheet for the 2N3819. Onsemi's is just as bad. The correct (and most comprehensive) data is found in Vishay 2N3819
Datasheets sometimes contain errors - you've found one in the Fairchild datasheet for the 2N3819. Onsemi's is just as bad. The correct (and most comprehensive) data is found in Vishay 2N3819
Hi,
I wouldn´t recommend JFETs as swiching devices, especially not those named here. The 2N3829 for example features an Rds(on) of 150Ohms, the J112 50Ohms. What You want is a as low and linear-over-signal-voltage as possible device. I´d rather opt for a low Rds(on) MOSFET instead or a dedicated switch or multiplexer like the ADG- and DG-series of AnalogDevices or the 460x-series (and others) of Maxim.
jauu
Calvin
I wouldn´t recommend JFETs as swiching devices, especially not those named here. The 2N3829 for example features an Rds(on) of 150Ohms, the J112 50Ohms. What You want is a as low and linear-over-signal-voltage as possible device. I´d rather opt for a low Rds(on) MOSFET instead or a dedicated switch or multiplexer like the ADG- and DG-series of AnalogDevices or the 460x-series (and others) of Maxim.
jauu
Calvin
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.