Curvy enough.hips, but a little short in the chest area.

 

Look up Erno Borberly if you are interested in FET amplifiers
https://audioxpress.com/article/The-All-FET-Line-Amp

Ian, I've been reading about JFETs and their characteristics curve (included photo here) for the "constant current" region of operation. That's the proper operation region for JFET amplification, and it's Nearly linear, and I'm Nearly certain of my understanding of that amplification function. By keeping the operational point (Q point) within the constant current region for all input Vgs: the drain current and V drain-resistor are nearly linear for all voltages of D to S, applied. I'm also certain you disagree, so I've included a photo of what I read

Look up Erno Borberly if you are interested in FET amplifiers
https://audioxpress.com/article/The-All-FET-Line-Amp

 

So, You believe that the particular type of Distortion Created by a Class-A Amplifier
when it is over-driven, hard into Distortion, is going to be
the "Magic-Bullet" Tone that "Every-Musician" is striving towards ..........

It's not that simple.

I would suggest that You spend some serious time playing around with Guitar-Effects-Pedals,
especially by using multiple "EQ-Pedals" in between multiple Distortion-Pedals, in a chain,
you'll be amazed at the ""Tones"" that You can create.

You can get Pedals in any Pawn-Shop cheap.

And then there's the hideous level of Distortion from "Musical-Instrument-Speakers" and Cabinets.

All of these "Effects" are at your finger-tips with any one
of the many Guitar-Software-Effects packages that are readily available,
some older ones are even free.
You simply use this Software before your Hi-Fi Stereo System to make it sound exactly
like a really bad Musical-Instrument Amplifier/Speaker.
Total Magic.
.
.
.

 

Ian, I've been reading about JFETs and their characteristics curve (included photo here) for the "constant current" region of operation. That's the proper operation region for JFET amplification, and it's Nearly linear, and I'm Nearly certain of my understanding of that amplification function. By keeping the operational point (Q point) within the constant current region for all input Vgs: the drain current and V drain-resistor are nearly linear for all voltages of D to S, applied. I'm also certain you disagree, so I've included a photo of what I read

I've seen that photo before - but I've never seen a graph like that on any real JFET datasheet. A real JFET has Gfs proportional to Id, so that Id varies as Vgs^2
The same graph from two datasheets taken at random J108 and 2SK208:
Not that in both cases the curves get closer together towards the bottom of the graph. Same applies to the 2SK3557 which I already have posted.
The only one that came close to being linear was a Silicon Carbide JFET but that was at 100A drain current.
A pentode valve might have a decent size linear region, which is why they remain popular.
I can't read the text on the first photo - could you have another go at photographing it, or tell me where it comes from.

 

Here's a similar curve from an EF80 datasheet. You can see that it's reasonably linear around Vgk=-1V, but the same effect occurs at grid voltages that become more negative.
Your only way of achieving greater linearity is more gain and more feedback.

 

Ian, let me review my

Here's a similar curve from an EF80 datasheet. You can see that it's reasonably linear around Vgk=-1V, but the same effect occurs at grid voltages that become more negative.View attachment 258996
Your only way of achieving greater linearity is more gain and more feedback.
[/QUOTE. Ian let me review my books discussion of transconductance and the calculation of Siemens, because it seemed to me that there's an algebraic manipulation that relates the "&m" basic equation to ohm's law. If I can verify to myself the validity of that algebra, I'll walk you through it by later tonight. I might need to photo my steps leading to that conclusion so be aware of any attached file, in my discussion.

 

All ofthe effects from the amplifier, not the pedals or boxes, cause problems when the singer's amplifier dies between numbers and so they plug the mic into the backup lead guitar amp. 25% harmonic distortion is OK for some guitar music but rather bad for a singer's voice. Learned that back in 1967.

 

But it's just the thing that the "Metal-Heads" are looking for !!!
.
.
.

 

Ian, let me review my

Ian, I've finished my algebra without good evidence of what I'm trying to support, but let me try another written attempt to relate the "&m" (Siemens) relation to "resistance" of the JFET channel that is somewhat proportional to the gate voltage of the input signal. I'll also include a better photo of the text so you can better understand my conclusions.
For an 2N5485 "ON CHARACTERISTICS" there a Zero-Gate-Voltage Drain Current listed as min. 4mAdc to max. 10mAdc range with 15Vds. My conclusion is that the resistance of the channel can be calculated to be between R=15V/4mA ohms and 15V/10mA ohms. And the transconductance of the 2N5475 varies from one to another FET, but within those min. max. calculated "channel resistances". And aren't those resistances essentially an ohm's law calculation, especially with note to ohm's law being a linear equation. The fact that gate voltage only partly controls drain current does indeed introduce a non-linear response of a FET, but above the max. pinch-off voltage (shown in that curve photo as 5 volts) 'isn't there' a nearly linear relationship of signal voltage to channel resistance and drain current??? It occured to me (like so many things do) that a voltage divider for 15Vsd biasing the gate of a 2N5475, the nonlinear control of the "channel resistance" or transconductance could be made linear. prrrtt!

 

Feedback is one proven method of improving linearity, unfortunately it also cuts gain. Of course the closer to linear that the device is the less correction is required and so it is useful to start with rather linear devices. The other consideration is noise, a much bigger challenge, really. and a topic for a different thread.

 

Many instrumental pop musicians like amps and speakers with distortion because they like the sound of their instruments better.

Most people who listen to the music prefer amplifiers and speakers with minimum distortion, since they want to hear the music to sound as close to live as possible (as the musicians hear it).

So using a FET in the instrument amplifiers might make sense to get the distortion the musicians desire.
It makes little sense to have that distortion in an amp to listen to that music, since that would just add more distortion.

 

Many instrumental pop musicians like amps and speakers with distortion because they like the sound of their instruments better.

Most people who listen to the music prefer amplifiers and speakers with minimum distortion, since they want to hear the music to sound as close to live as possible.

So using a FET in the instrument amplifiers might make sense to get the distortion the musicians desire.
It makes little sense to have that distortion in an amp to listen to that music, since that would just add more distortion.

So very true! Creating the sound (music) is very much different than reproducing it. That opening chord sound does not need to be modified any more.

 

Ian, I've finished my algebra without good evidence of what I'm trying to support, but let me try another written attempt to relate the "&m" (Siemens) relation to "resistance" of the JFET channel that is somewhat proportional to the gate voltage of the input signal. I'll also include a better photo of the text so you can better understand my conclusions.
For an 2N5485 "ON CHARACTERISTICS" there a Zero-Gate-Voltage Drain Current listed as min. 4mAdc to max. 10mAdc range with 15Vds. My conclusion is that the resistance of the channel can be calculated to be between R=15V/4mA ohms and 15V/10mA ohms. And the transconductance of the 2N5475 varies from one to another FET, but within those min. max. calculated "channel resistances". And aren't those resistances essentially an ohm's law calculation, especially with note to ohm's law being a linear equation. The fact that gate voltage only partly controls drain current does indeed introduce a non-linear response of a FET, but above the max. pinch-off voltage (shown in that curve photo as 5 volts) 'isn't there' a nearly linear relationship of signal voltage to channel resistance and drain current??? It occured to me (like so many things do) that a voltage divider for 15Vsd biasing the gate of a 2N5475, the nonlinear control of the "channel resistance" or transconductance could be made linear. prrrtt!

Waiting for IanO