If you're not afraid to try something, here's a little circuit with three 2sk170s. It only needs 5 resistors, one capacitor, three JFETs and one LED. Gain is about 2, with plenty of driving current. The LED should be a low brightness type with about a 2v drop. It functions as both a power indicator and as bias for the first JFET. The other two JFETs are a source follower and constant current source. Total current draw is about 13ma. You could probably just knock together this circuit "bugsplat" style in a few minutes to see how well it works.
Attachments
Correction on CS vs. CD gain; FET circuits
Oops. The above quote requires a small edit. I meant to refer to common-DRAIN circuits (CD). Common-source circuits (CS) can indeed have gain.
I apologize for the confusion.
FET gain depends to some extent on the ratio of the drain and source resistors. (Rd / Rs: The internal resistance and load also alter the gain.) For more gain, decrease the source resistor (Rs) or increase the drain resistor (Rd). There are limits, though: The average current must exceed the minimum Idss spec for the FET. Otherwise, your FET will pinch off, and probably quit amplifying.
Oops. The above quote requires a small edit. I meant to refer to common-DRAIN circuits (CD). Common-source circuits (CS) can indeed have gain.
I apologize for the confusion.
FET gain depends to some extent on the ratio of the drain and source resistors. (Rd / Rs: The internal resistance and load also alter the gain.) For more gain, decrease the source resistor (Rs) or increase the drain resistor (Rd). There are limits, though: The average current must exceed the minimum Idss spec for the FET. Otherwise, your FET will pinch off, and probably quit amplifying.
Apologies for restarting a very old thread, however...
I am just learning JFET biasing theory and building up my understanding.
...I have come to understand that every individual JFET has a specific measurable value of IDSS which is a fixed and inherent property of that individual JFET. The datasheet shows a range of values in which an individual sample may fall (not a range of ID which can be used for any JFET of that type).
I can not match my understanding as described above with the comment in post #24 which says
'The average current must exceed the minimum Idss spec for the FET. Otherwise, your FET will pinch off, and probably quit amplifying'
Can anyone clarify?
Many Thanks...
I am just learning JFET biasing theory and building up my understanding.
...I have come to understand that every individual JFET has a specific measurable value of IDSS which is a fixed and inherent property of that individual JFET. The datasheet shows a range of values in which an individual sample may fall (not a range of ID which can be used for any JFET of that type).
I can not match my understanding as described above with the comment in post #24 which says
'The average current must exceed the minimum Idss spec for the FET. Otherwise, your FET will pinch off, and probably quit amplifying'
Can anyone clarify?
Many Thanks...
Instead of "average" it should probably say "peak".
Say the sheet gives "Idss = 2mA to 6mA".
Say you design for 4mA average, but grab a specific part that only does 2mA.
You will only get 2mA, NOT 4mA.
Say you design for 2mA average and grab the same specific part that only does 2mA. It will idle at 2mA, and swing "down" to zero mA, but will not swing "up" above 2mA (much).
Assuming symmetric swings (close enough for jazz), if the spec says "2mA to 6mA" then you should design to idle at 1mA max.
Or else sort-out a baggie of parts to find the higher Idss parts. (Which may be tolerable in DIY but not for mass production.)
Say the sheet gives "Idss = 2mA to 6mA".
Say you design for 4mA average, but grab a specific part that only does 2mA.
You will only get 2mA, NOT 4mA.
Say you design for 2mA average and grab the same specific part that only does 2mA. It will idle at 2mA, and swing "down" to zero mA, but will not swing "up" above 2mA (much).
Assuming symmetric swings (close enough for jazz), if the spec says "2mA to 6mA" then you should design to idle at 1mA max.
Or else sort-out a baggie of parts to find the higher Idss parts. (Which may be tolerable in DIY but not for mass production.)
Many Thanks PRR for a very clear answer,
......When you are learning a concept and think you have nailed it and then read something which contradicts completely you tend to doubt yourself.
The specific question I was really trying to ask but didn't express very well is whether it was correct to attempt to run a JFET at a current greater than its IDSS.
You have confirmed that you can't, which is what I thought, whereas the post originally mentioned implied that you must do.
Cheers again,
Mike
(I tried to edit my previous post to clarify it, but the edit does not seem to have appeared)
......When you are learning a concept and think you have nailed it and then read something which contradicts completely you tend to doubt yourself.
The specific question I was really trying to ask but didn't express very well is whether it was correct to attempt to run a JFET at a current greater than its IDSS.
You have confirmed that you can't, which is what I thought, whereas the post originally mentioned implied that you must do.
Cheers again,
Mike
(I tried to edit my previous post to clarify it, but the edit does not seem to have appeared)
Ah, I see that now. I believe the post is missing a "not".
> The average current must not exceed the minimum Idss spec for the FET.
Maybe line-noise, maybe forum corruption, maybe the mind got ahead of the fingers.
A typical guitar preamp should be able to take 0.25Vrms input and has a gain of 3.
So the JFET in the circuit shown should have a Vto of at least -2V, preferrably more.
Using a J201 as shown above, the output will have a distortion of ~0.5% with 0.25Vrms at the input.
By raising the voltage to 18V and using a J111 instead, you can reduce this to ~0.02%.
Adding one more J111 and a resistor will also allow you to lower Zout dramatically.
Total current consumption is about 3mA.
And you can build the entire thing in SMD as well, so it will be tiny.
The J111 should have a Vto of about -4V.
Patrick
So the JFET in the circuit shown should have a Vto of at least -2V, preferrably more.
Using a J201 as shown above, the output will have a distortion of ~0.5% with 0.25Vrms at the input.
By raising the voltage to 18V and using a J111 instead, you can reduce this to ~0.02%.
Adding one more J111 and a resistor will also allow you to lower Zout dramatically.
Total current consumption is about 3mA.
And you can build the entire thing in SMD as well, so it will be tiny.
The J111 should have a Vto of about -4V.
Patrick
Speaking as both a guitarist and a person who knows a fair bit about audio electronics, lowering JFET distortion is the last thing you want for a JFET electric guitar preamp.
The 12AX7 triode usually used for this job in good guitar amplifiers produces several percent of low-order harmonic distortion, and this softens the harsh sound of a naked solid-body electric guitar appreciably. It is a good thing, and the reason guitarists still put up with fragile, expensive, bulky, heat-generating valves.
When you use a JFET in place of the 12AX7, it would be wonderful if the JFET could be coaxed into producing as much distortion as the half-12AX7 did. We would rather have 5% THD than 0.5% if possible, and a distortion of 0.02% would be a shame, as it's too little to make an electric guitar sound good!
Electric guitar amplification is nothing like Hi-Fi. The goals are very different - not to accurately reproduce an audio signal with minimum distortion and noise, but rather to take an inherently rather cold, harsh, unattractive-sounding instrument, and intentionally mess with its signal until it actually sounds quite good. We aren't reproducing a good musical signal, we're trying to create one!
-Gnobuddy
Very true! And I think a good part of that has been driven by the poor sustain and timbre of those new-fangled electric guitars compared to all the traditional orchestral instruments that existed at the time. If you compare the thin "plink!" of a solid-body electric guitar with the rich sound of a viola, the guitar sounds like a miserable wretch of an instrument. I think a lot of people saw it that way at the time, too.
Before there were solid-body electric guitars, if you look back at electric guitar history, early electric guitars were mostly big hollow archtops, which produced reasonably pleasant sounds of their own (not as good as a viola or cello, but more musical than later solid-body guitars.) It seems that guitar amplifiers of that era weren't designed differently from those for music reproduction, and were usually just lifted from the back of the RCA tube catalog.
But when Les Paul and Leo Fender and several others started making guitars that were basically just solid blocks of wood with pickups, the guitars didn't sound great on their own (at least in my opinion.)
Les Paul himself seemed to like the cold and sterile-clean sounds of his low-impedance pickups, but few others did. Meantime, Leo's favourite surf music used huge dollops of reverb to hide the harshness of his Telecasters. Several percent of harmonic distortion from the "mid-fi" amp designs Leo lifted from the tube catalogs helped.
Nearly seventy years later, our ears have been trained to like the sound of distorted solid-body electric guitars. Not many people today would enjoy Les Paul's guitar tone.
-Gnobuddy
Well it does make it louder also! So it is an amplifier, no? But with special characteristics.
Like a Hifi, it has a preamp, the power stage and the speakers, which all effect the sound. Just in this case all three are tortured.
I know! Lets call it a "Guitar Amp"
Like a Hifi, it has a preamp, the power stage and the speakers, which all effect the sound. Just in this case all three are tortured.
I know! Lets call it a "Guitar Amp"
Last edited:
I can't think of a solution to the problem, but you have to admit, the term "guitar amplifier" has confused a lot of audio engineers, electronics techs, and DIY electronics hobbyists over the years, because they think a guitar amp serves the same purpose as a Hi-Fi amp.
I made exactly the same mistake for some twenty years, before we had the Internet. Specialized knowledge like this was not easy to find back then - only engineers and techs working in the electric guitar amp industry knew, and they kept their mouths shut about all their trade secrets.
Probably for the same reason (trade secrets), you couldn't find any electronics textbooks telling you how guitar amps were weird and different from Hi-Fi, either.
Now we can hunt down stuff easily on the 'Net, but with that came a different problem: no editing or proof-reading, so maybe 99% of the "facts" you find on the 'Net turn out to be utter nonsense! (The earth is flat! UFOs exist! Contrails are secret mind-control chemicals! Elvis is alive! The moon landings were faked! Digital audio sounds bad! )
-Gnobuddy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.