Hello!
I built a 2-channel tube amp using 2N5638 JFETs as switches, first time using them. The switches turn on and off seamlessly but I have distortion and attenuation problems, I think because the signal swings ~1-2 Volts, plus or minus...
Here is the way I use them:
How can I correct this?
Thanks!
I built a 2-channel tube amp using 2N5638 JFETs as switches, first time using them. The switches turn on and off seamlessly but I have distortion and attenuation problems, I think because the signal swings ~1-2 Volts, plus or minus...
Here is the way I use them:
An externally hosted image should be here but it was not working when we last tested it.
How can I correct this?
Thanks!
The 'correction' will be not to use SS devices as 'mute switches'. Their non-linear capacitances make for horrid sound. No BS, use a relay if you must mute the output.
(One of the best and most audible mods I've ever done to commercial gear was to remove the muting transistors on the output of a CDP. Better to have a little 'thump' than have these muting devices.)
(One of the best and most audible mods I've ever done to commercial gear was to remove the muting transistors on the output of a CDP. Better to have a little 'thump' than have these muting devices.)
You forgot a couple of resistors. You need DC path through resistor from gate to source to discharge it when the switch is "on". Also, you need DC path to ground from it's source. But now you have a problem with some DC on output. To solve this problem you need more than one JFET for the switch, like "T" of 3 JFETS, couple of horizontal and one vertical controlled by inverted levels.
I guess I forgot to say that there are 2 groups of FETs. When Group1 has -13VDC, Group2 has 0VDC. As far as I can tell, the Control voltages are VDC, there is no swing although I don't have an oscillator to verify this.
I am probably wrong in this, but wouldn't a small capacitor couple the changes of the signal voltage, so that Vgs is always 0V when there is no negative voltage to turn it off?
Just a noob thought...
Here is the circuit that provides DC to the gates.
This switching works perfectly and there is no voltage swing. The wanted here is to offset the negative side of the signal to the gate voltage for every JFET.
A friend suggested the placement of another diode like this:
I will choose to try this first due to available room issues on the eyelet board. Otherwise I'll start experimenting with the suggestions you guys provided here.
I'll post feedback when I'm done!
An externally hosted image should be here but it was not working when we last tested it.
This switching works perfectly and there is no voltage swing. The wanted here is to offset the negative side of the signal to the gate voltage for every JFET.
A friend suggested the placement of another diode like this:
An externally hosted image should be here but it was not working when we last tested it.
I will choose to try this first due to available room issues on the eyelet board. Otherwise I'll start experimenting with the suggestions you guys provided here.
I'll post feedback when I'm done!
FET's are my preferred option for switching but you have to be careful how you drive them.
Take a look at post #2 here,
http://www.diyaudio.com/forums/soli...fet-amplifier-designed-music.html#post1452488
This should work well if you resistively load the output of Q2
Post #201 may help make clearer what happens as the FETs come into conduction,
http://www.diyaudio.com/forums/soli...-amplifier-designed-music-11.html#post1569428
Take a look at post #2 here,
http://www.diyaudio.com/forums/soli...fet-amplifier-designed-music.html#post1452488
This should work well if you resistively load the output of Q2
Post #201 may help make clearer what happens as the FETs come into conduction,
http://www.diyaudio.com/forums/soli...-amplifier-designed-music-11.html#post1569428
Hello Mooly!
Thank you for the links!
Yes, a series-shunt topology will be next step to do. I see you have 22kOhm resistors in the High-pass filter. Should I use the same amount of resistance after the JFET?
Also, by adding a diode like the one in the picture I posted before will it function properly?
Thank you!
The 22K has no special value attached to it. It was just a test circuit to demonstrate the FET properties as a switch. In the actual amp circuit of mine (first link) the FET's work into a virtual earth point at the opamp input. Values of 18 to 22K are typical there.
So if you go to high with the 22K you will at some point encounter problems with stray capacitance and hf roll off. Go too low (lets say 100 ohm to highlight the point) then shunt FET (if used) will form a divider with that 100 ohm. If the on resistance of the FET is 100 ohm too then the signal is only attenuated by half by that part of the circuit. It would also "short" the source component output via the resistor.
The resistor after the FET (the 1Meg in the test circuit) just defines a DC reference and ties the output to ground. You would probably use say 100K ?
So if you go to high with the 22K you will at some point encounter problems with stray capacitance and hf roll off. Go too low (lets say 100 ohm to highlight the point) then shunt FET (if used) will form a divider with that 100 ohm. If the on resistance of the FET is 100 ohm too then the signal is only attenuated by half by that part of the circuit. It would also "short" the source component output via the resistor.
The resistor after the FET (the 1Meg in the test circuit) just defines a DC reference and ties the output to ground. You would probably use say 100K ?
In my preamp circuit (by the way, this is a tube guitar amp I am building, so frequencies above 5-6kHz are not much of importance) there is a 470k resistor.
I am attaching the image of the JFET node in the preamp.
Attachments
So if your points A,B,C and D are the points where you want a switch then I think it should work OK. All those points look ground referenced (no DC present)
The isolation of a single FET is very high (at DC its virtually infinite) and at AC it is due to capacitive and junction capacitance so you may only need a single FET rather than a series shunt arrangement anyway. In my preamp I went for maximum isolation between different signal sources but thats not an issue here.
One question is what the peak to peak signal levels will be as that has to be within the acceptable and allowable range of gate-source voltage to the FET. For example if you turn the FET off with say -8 volts then that means that signals that exceed that level will break through.
The isolation of a single FET is very high (at DC its virtually infinite) and at AC it is due to capacitive and junction capacitance so you may only need a single FET rather than a series shunt arrangement anyway. In my preamp I went for maximum isolation between different signal sources but thats not an issue here.
One question is what the peak to peak signal levels will be as that has to be within the acceptable and allowable range of gate-source voltage to the FET. For example if you turn the FET off with say -8 volts then that means that signals that exceed that level will break through.
The control voltage is a bit higher that that, its at -13.66 VDC. The signal is at around 2 Vpp, so when the JFETs are off no signal comes through. When the FETs are ON, however, they fluctuate, since there is no signal offset to the Gate voltage, which is at 0V when ON.
So, do you think the addition of a diode from Source to Gate would give that offset I need?
So, do you think the addition of a diode from Source to Gate would give that offset I need?
J111 & J112 are ideally suited to FET switching. If the source is treated as the input then it should be strapped to the gate using ~22k to ensure that Vgs remains @ 0V in the presence of an input signal. These FETs are apparently symmetrical.
Rather than try to reiterate the entire section from the book, I would suggest reading pages 408-421 of Small Signal Audio Design by Douglas Self which contain a comprehensive treatment of this subject.
Rather than try to reiterate the entire section from the book, I would suggest reading pages 408-421 of Small Signal Audio Design by Douglas Self which contain a comprehensive treatment of this subject.
You must drive the FET's via a resistor. The diode leaves the gate floating which is no good. Make sure that -14 volts is within the G-S rating of your chosen FET. (I could use much lower "on" drive voltages than you because of the virtual earth opamp).
I used an arrangement whereby the discrete "logic" driving my FET's ran on an -8 volt rail and for "FET on" was clamped to a 0.2 volt rail proveded by a forward biased germanium diode.
I think so. Its the FETs I am new into. But if u have more info please share!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.