TILLMAN: HOW TO INCREASE OUTPUT LEVEL?

[JDoyle]

(Now that I think about it, using this "valve" terms contribute nothing but mojo.)

Great point. I resolve to stop doing that right now. It's Ohmic Region and Constant Current Region from now on.

I think I see your point. I haven't been able to obtain a closed formula for the transition between constant current and ohmic regions for a general case yet.

Honestly? I don't think there is one. Keep in mind that the two 'standard' formulas we see, for the ohmic region and the constant current region are approximations only. Along with the approximation of the 'square' in 'square law' - it differs for each JFET and can be as low as 1.64, and as high as 2.89(!), in measurement tables I have seen - it just happens to average out to be 2.

There is the formula for FINDING the line between the two regions, but I don't know of any formulas that cover the ENTIRE range of output operation. And not to toot my own Library, as it were, but I've got several books on just JFET theory and operation. The two best are by Todd: "Junction Field Effect Transistors" and Levin "Field Effect Transistors", the former be the most accessably and complete (and also most recent, if you call the early 70's recent) but the latter (issued by Texas Instruments Press) is a source for nearly every early scientific paper written on JFETs, from Middlebrook and Richer's (discoverers of the 'square law approximation') later works, to Borbeley.

Point being - NO book I own has an all encompassing formula. And you know if engineers like that, who love math as much as sex, didn't find it, we most assuredly won't.

Also, I think a large part of the confusion with JFETs is the use of BOTH Vgs(off) and Vp to mean, well, Vgs(off). Thing is, (I'm sure you know this STM, this is for anyone who happens to still be awake and reading this, a very special, and wonderful type of person indeed), |Vgs(off)| = Vp . It takes a while for that to sink in and you really need to see the depletion zone sketches to understand it, but - The drain to source voltage that causes the channel to pinch off, is close enough for rock and roll (and the rest of semiconductor engineering) to be the SAME value as the absolute value of Vgs(off) (in laymans terms: Vgs(off), but positive). But the two DO mean different things, Vp refers to the Vds value that pinches off the channel, where Vgs(off) refers to the Vgs value that CLOSES off the channel (or only allows 1 nA of current through).

The JFET is such a rudimentary device that it's operation is VERY VERY hard to adequately, and accurately, describe - you essentially have a diode with ONE anode and TWO cathodes, with the whole thing reversed biased, but one cathode more reverse biased than the other. It is about as 'pure' a semiconductor device (thus, it being the 'first' transistor, theoretically at least) as you can get as there is only ONE pn-junction, but it has two 'ends' so to speak...

Fascinating actually.

This suggests, at least for a particular case, that the ohmic region is independent of VCC, thus your statement might be correct.

I hope you don't mind me dropping the bit about the 'magic' biasing point. I think that is surely interesting, but a different discussion entirely.

I think that this shows itself in the transfer curves of Id vs. Vds - the lines separating the regions are solid and stable yet Vds increases until breakdown.

Well, referring again to the experience on the Fetzer Valve optimal biasing is given by:

Vd = 0.6*Vcc + 0.7*|VP|
Rd = 0.9 * (Vcc - 2*|VP|) / IDSS

showing dependence on both Vcc and Vp.

I see your point here, the only quibble I have is that you are designing for a specific and very distinct operating point, instead of an overall view of JFET operation. Not to diminish what you have done in the least, I think it is really cool!

But I don't think it can really be relied upon for anything BUT the Fetzer Valve.

My pleasure.  It is good to have the chance to explore, discuss and exchange ideas about this!

I hope it is allowed to continue, and does.

Kindest Regards,

Jay Doyle

[JDoyle]

my apologies jay, I should have said since our comments are getting personal, I would like to talk to you rather type.
Thanks,
Brian

I'm over it Brian, and on to other things.

Thank you for the gentlemanly overture, but I do not see the need.

Regards,

Jay Doyle

[JDoyle]

I do have to make one final comment though:

Do those who post those pictures in response to disagreements (the ladies handbags for instance) think that they are helping in any way?

The people who are having the disagreement (at least myself for instance) happen to believe passionately in their position, and do you think that making fun of them, no matter how ridiculous it may seem to you, serves ANY purpose OTHER than to fan the flames and make it worse?

Or is that the whole point to begin with?

Honest question. Seriously.

Regards,

Jay Doyle

[gez]

I do have to make one final comment though:

Do those who post those pictures in response to disagreements (the ladies handbags for instance) think that they are helping in any way?

The people who are having the disagreement (at least myself for instance) happen to believe passionately in their position, and do you think that making fun of them, no matter how ridiculous it may seem to you, serves ANY purpose OTHER than to fan the flames and make it worse?

Or is that the whole point to begin with?

Honest question. Seriously.

Regards,

Jay Doyle

Jay, it's one thing to have a disagreement about something in a gentlemanly manner, but another when it descends into an on-line spat.  When that happens it just gets ridiculous and is disruptive to the forum.  If it takes a little humour at someone else's expense to highlight this, then so be it.

How many lives did this thread save?  In the grand scheme of things we're arguing about little bits of plastic and silicon.  It's not that important.  By all means defend your corner; be pedantic (I am myself); just don't loose your head over it.  It's Chinatown...

I don't mean you any ill will, Jay.  Likewise with Brian.  Feel free to think of me as a pr*ick from now on (everyone else does!), but this is my last post on the matter. 

[Ben N]

Wow, excellent thread--or at least the essential core of it, about 1-2 pp in length. Someone ought to edit this down to a wiki article encompassing the problem--adjusting gain in single jfet stage devices--and the three broad approaches to it, including applications and critiques. I'd do it, but I'm sure I would absolutely mangle it.

As for me, I could do without the bitterness, but appreciate the passion. Thanks to Jay, STM and Brian for explicating their approaches so eloquently, Gus for his pithy additions, and to Andy, Pete, Aron, Gerry, et al, for bringing it all back into perspective. Especially Andy: Dude, making cool pedals is all well and good, but making peace is better.

[Gus]

Something to add for people to look up.

what is lower distortion cap bypassed source resistor on non bypassed?  I paper I read has something different than what one might think.

[Renegadrian]

Wow, excellent thread--or at least the essential core of it, about 1-2 pp in length. Someone ought to edit this down to a wiki article encompassing the problem--adjusting gain in single jfet stage devices--and the three broad approaches to it, including applications and critiques. I'd do it, but I'm sure I would absolutely mangle it.

As for me, I could do without the bitterness, but appreciate the passion. Thanks to Jay, STM and Brian for explicating their approaches so eloquently, Gus for his pithy additions, and to Andy, Pete, Aron, Gerry, et al, for bringing it all back into perspective. Especially Andy: Dude, making cool pedals is all well and good, but making peace is better.

Ben, you or someone else could just make a PDF file and put it in the gallery...

[Ben N]

Too big!

[frank_p]

Do those who post those pictures in response to disagreements (the ladies handbags for instance) think that they are helping in any way?

The people who are having the disagreement (at least myself for instance) happen to believe passionately in their position, and do you think that making fun of them, no matter how ridiculous it may seem to you, serves ANY purpose OTHER than to fan the flames and make it worse?

Jay Doyle

Sorry Jay (Should I say Renegadrian) but it was not to make fun of anybody.  There was no debugging so a multimeter was perhaps not necessary.  But I think that a bit of humour and a multimeter can be useful when dealing with FETs.

Again, sorry.
F.H.P.

[JDoyle]

Response to Gez posted in Member's Section to keep from further cluttering up this thread.

Regards,

Jay Doyle

[jimbob]

What a silly thread. 6 pages. Wow.

[rotylee]

there needs to be a forum management tool called "splitting threads".

>

when people come in with grand equations

please site what your variables and abbreviations mean.

Current = I
Inductance = L
is not an obvious thing you know
Electronics is filled with this crap
i think it is designed to obfuscate

jargon is a factor of exclusion.
designed to make those in the knows life simpler, while mystifying the processes to neophytes and onlookers.

[slacker]

when people come in with grand equations

please site what your variables and abbreviations mean.

From what I can see people have done that in this thread, certainly Jay and Sebastian's equations are pretty well notated in my opinion.

I wouldn't say the so called jargon is designed to exclude or mystify anyone. Using terms like I for current is a universal thing and is taught in science classes at school, or at least it was in my day.
Yes it makes it easier for people in the know to write things down but seen as how this is at least in part an electronics forum it's not unreasonable for people to use common scientific terms.

Just my 2ps worth of course

[frequencycentral]

I wouldn't say the so called jargon is designed to exclude or mystify anyone. Using terms like I for current is a universal thing and is taught in science classes at school, or at least it was in my day.
Yes it makes it easier for people in the know to write things down but seen as how this is at least in part an electronics forum it's not unreasonable for people to use common scientific terms.

I think this forum is made up of people who are electronics engineers, people who are musicians, and people who are both to whatever degree. I would guess the majority are musicians with the ability to solder.

I myself have no electronics/higher math training. Hmm, no formal musical training either come to that. Neither stop me writing music or designing pedals/synths.

What I do have, as I think the majority of members have, is enthusiasm for the subject - sometimes it may get a little too technical for those without the 'right' background (as can music itself sometimes). I'm not saying dumbing down is the answer - but sometimes working out some posts is like codebreaking - a lot of research needed to get the jist of it. Then again - reseach is good.

[DougH]

jargon is a factor of exclusion.
designed to make those in the knows life simpler, while mystifying the processes to neophytes and onlookers.

Yeah, that's right... That's why we have a system of measurement that includes units like tablespoons and teaspoons, cups & quarts. It's to mystify the process of cooking so only a select few are capable of doing it.

It's also why we have temperature scales like Farenheit and Celsius- to confuse those who dare to attempt to read the temperature.



It's amazing to me in a thread like this what lengths people go to to rationalize ignorance.

[stm]

I think I see your point. I haven't been able to obtain a closed formula for the transition between constant current and ohmic regions for a general case yet.

Honestly? I don't think there is one. Keep in mind that the two 'standard' formulas we see, for the ohmic region and the constant current region are approximations only. Along with the approximation of the 'square' in 'square law' - it differs for each JFET and can be as low as 1.64, and as high as 2.89(!), in measurement tables I have seen - it just happens to average out to be 2.

There is the formula for FINDING the line between the two regions, but I don't know of any formulas that cover the ENTIRE range of output operation. And not to toot my own Library, as it were, but I've got several books on just JFET theory and operation. The two best are by Todd: "Junction Field Effect Transistors" and Levin "Field Effect Transistors", the former be the most accessably and complete (and also most recent, if you call the early 70's recent) but the latter (issued by Texas Instruments Press) is a source for nearly every early scientific paper written on JFETs, from Middlebrook and Richer's (discoverers of the 'square law approximation') later works, to Borbeley.

Point being - NO book I own has an all encompassing formula. And you know if engineers like that, who love math as much as sex, didn't find it, we most assuredly won't.

The point you make about the square exponent being an approximation is a good one.  I wasn't aware of it.  Nevertheless, using a value of 2 still provides a reasonable framework to work with, just as the (1 + lambda*vds) term for the id current expression is assumed to be very close to unity.  I think we have to accept that math here won't produce an exact solution, but a reasonable approximation.

Regarding formulas, I have to say I SUCCEEDED.  Yes, I found closed formulas for idq, gm, gain, RS and RD for the general case.  I have to admit that I finally swallowed my pride and resorted to a symbolic math program to do the work (Maxima), as it was required to solve symbolically large quadratic and cubic equations.  I will post the results in another thread as I am able to organize it properly.  The good thing is that, if properly entered in a spreadsheet, they will be accessible for everybody.

Also, I think a large part of the confusion with JFETs is the use of BOTH Vgs(off) and Vp to mean, well, Vgs(off). Thing is, (I'm sure you know this STM, this is for anyone who happens to still be awake and reading this, a very special, and wonderful type of person indeed), |Vgs(off)| = Vp . It takes a while for that to sink in and you really need to see the depletion zone sketches to understand it, but - The drain to source voltage that causes the channel to pinch off, is close enough for rock and roll (and the rest of semiconductor engineering) to be the SAME value as the absolute value of Vgs(off) (in laymans terms: Vgs(off), but positive). But the two DO mean different things, Vp refers to the Vds value that pinches off the channel, where Vgs(off) refers to the Vgs value that CLOSES off the channel (or only allows 1 nA of current through).

Good point. The fact is that many textbooks and application notes just use Vp to represent Vgsoff, many times with an implied negative sign in case of N-channel JFETs.

The JFET is such a rudimentary device that it's operation is VERY VERY hard to adequately, and accurately, describe - you essentially have a diode with ONE anode and TWO cathodes, with the whole thing reversed biased, but one cathode more reverse biased than the other. It is about as 'pure' a semiconductor device (thus, it being the 'first' transistor, theoretically at least) as you can get as there is only ONE pn-junction, but it has two 'ends' so to speak...

Fascinating actually.

Beautiful description. I really dig it.

I see your point here, the only quibble I have is that you are designing for a specific and very distinct operating point, instead of an overall view of JFET operation. Not to diminish what you have done in the least, I think it is really cool!

But I don't think it can really be relied upon for anything BUT the Fetzer Valve.

With the general formulas I've found now it is possible to resort to the general case. Just give me some time to organize and post the info.

Best regards,
STM

[StephenGiles]

Hay que tener corazon - eh Sebastian!!

[wampcat1]

jargon is a factor of exclusion.
designed to make those in the knows life simpler, while mystifying the processes to neophytes and onlookers.

Yeah, that's right... That's why we have a system of measurement that includes units like tablespoons and teaspoons, cups & quarts. It's to mystify the process of cooking so only a select few are capable of doing it.

It's also why we have temperature scales like Farenheit and Celsius- to confuse those who dare to attempt to read the temperature.



It's amazing to me in a thread like this what lengths people go to to rationalize ignorance.

I see your point, but there is something to be said for the chef who uses a 'pinch' of this and a 'pinch' of that in order to make a great meal.

[selectortone]

Many moons ago I built an FET preamp, from a circuit I found on the web, for an acoustic guitar I wanted to use in a live situation. It didn't work. It was such a simple circuit and it drove me completely nuts trying to figure out where I'd gone wrong. I didn't know about the variability in FETs then. In the end I had to admit defeat and I threw it away in frustration eventually.

I wish I'd had access to this thread then.

btw, you really need to buy a multimeter!

[Renegadrian]

btw, you really need to buy a multimeter!

Well, I believe you're right, but I feel I don't need it for now...I built several circuits and they all worked, and when they didn't it was for some solder mistakes, so it was easy to solve them...I know that someone don't like this approach, but hey, we're all different and we're all right at the end, given that we arrive at the point we settle...