FET emulation of tubes without trimmer?

[puretube]

ooops - info lost... 

[R.G.]

FETs don`t like me...

FETs - at least JFETs - don't like anyone. They're like precocious but intractibly ill-tempered children. They not only have great promise, but they also have quirks and the quirks vary so you can't easily compensate for them.

[BJF]

FETs don`t like me...

FETs - at least JFETs - don't like anyone. They're like precocious but intractibly ill-tempered children. They not only have great promise, but they also have quirks and the quirks vary so you can't easily compensate for them.

Hi,

Very true, though it is more perhaps polite to look at them as individuals.
There are ways of controlling behaviour, but it involves finding the ones that like to play or specialorder big batches and sometimes circuit decisions, while the circuit may call for a certain behaviour and then compensation is pointless and may not be possible in view of the supply voltage.
There lies no magic in a component name when it comes to JFETs.

Maybe it is so that you have to love JFETs, like a parent for them to do what you tell them to

Patience

Regards
BJ

[will]

Hi Puretube,

How about Jack's mini-booster style configuration. Just about any transistor will work as the top transistor to create the constant current. No trimpot required.

Regards,
Will

anybody got experience in sound
with other methods of biasing a FET
in tube-clone-circuits
than an annoying trimpot RD,
while maintaining the possibility of replacing
FETs with large production spreads?
:?:

[puretube]

too pentodish for me...
too Ormanic...

well, and being the guy having provided the TubeCadJournal with the original document in times when the www wasn`t that advanced
http://v3.espacenet.com/origdoc?DB=EPODOC&IDX=US2310342&F=0&QPN=US2310342,
I do know the nice performance features of an SRPP,
and do use it in some tube-projects -
but I was looking for a single-"triode"-jFET solution.

re-activated this old thread only, to place the N&V links...

found my peace with CMOS hex-inverters for a while, now...

[aron]

I wish I knew as well. In practice I have been finding that the J201s that I sell have worked with a 15K drain, 1M gate and 1K resistor. I've been able to swap them in and out without problems.

[puretube]

FETs don`t like me...

FETs - at least JFETs - don't like anyone. They're like precocious but intractibly ill-tempered children. They not only have great promise, but they also have quirks and the quirks vary so you can't easily compensate for them.

Lately, they changed their feelings towards me...

[earthtonesaudio]

There are clues to be found in some appnotes and the transfer/transconductance curves while choosing a q-point...  Though I still don't quite understand it.  I think you can use a straight edge, or lots of algebra. 

[puretube]

The 1st signs of "Joy To The FETs" showed up here
but now went commercial... 

[Eb7+9]

There are clues to be found in some appnotes and the transfer/transconductance curves while choosing a q-point...  Though I still don't quite understand it. 

the algebra is secondary ... to get class-A (curvature) like behavior from a jFET one needs to operate it with minimal resistive feedback in the Source circuit ... in a typical (common-source like) passive biasing schemes having little to no resistance in the Source circuit will also land you on the edge of the curvature range (ie., at Vgs=0) ... so a lateral shift of the response curve relative to input voltage is required ... there are at least two practical ways of doing this - one is by applying a low-Z voltage source (op-amp) directly to a Gate-grounded jFET's Source and setting desired Vgs ... this biasing scheme doesn't produce any Source circuit NFB so transfer curvature is maintained ...

the reason why jFET's are still a mystery is pretty simple - unlike Bipolar devices jFETs have two degrees of parametric variability ... the matching test commonly referred to here is bogus (maybe it's intentionally put out to mislead people) - when jFETs are tested for both Vgs(off) and Idss the mystery goes away ... then, matching jFET's involves choosing devices that lie closest together in a  2-D sense (inside a smaller circle) ...

[gaussmarkov]

the reason why jFET's are still a mystery is pretty simple - unlike Bipolar devices jFETs have two degrees of parametric variability ... the matching test commonly referred to here is bogus (maybe it's intentionally put out to mislead people)

[stm]

Conspiracy theory? Nope.

Check this more than 3-year-old thread:

http://www.diystompboxes.com/smfforum/index.php?topic=29513.0

It includes the math and a suitable circuit plus several comments from the respected and stable contributors of this forum.

[R.G.]

... the matching test commonly referred to here is bogus (maybe it's intentionally put out to mislead people)

JC do you just try to invent controversies?

First, the test is not bogus for its intended purpose, and second it is downright defamatory for you to imply that it's there to mislead.

The matching test commonly referred to here is (a) not perfect, never was held out as perfect and (b) very, very fast and simple for the limited aims it was intended for - matching JFETs for phasers. Go try to tell the hundreds of people who have produced matched sets of JFETs for phasers that work just fine, thank  you, that the test doesn't work.

As to being put out to intentionally mislead, would you please either issue a written retraction of that or support the statement with fact?

[DougH]

To answer the OP, even though it's almost 3 years old, is yes, sort of... Well, yes as far as Rd is concerned. Michael Ibrahim posted a nice idea for JFET stage biasing/tuning that helps to make gain/freq-resp a little more independent of bias than the typical variable Rd scheme. I tried it, it works and sounds very good (I have my own suspicions why).

Anyway, it can be searched in the older threads. Discover away, children.

[earthtonesaudio]

A quote from an old Temic/Siliconix appnote:

"Gain variations are minimized when the load line is designed to intersect the pair of limiting transfer curves (top) at points of equal g_fs (bottom)."

They have the two graphs one above the other (transfer characteristics on top, transconductance vs. gate-source voltage on bottom), and they just have the load line intersect the two extremes on the transfer graph, then drop down to the transconductance graph, and make the two points have the same forward transconductance.  Yeah, I'm still working it out on paper to try and get the same thing for J201s/9V supply...

[Caferacernoc]

It seems to me everytime a jfet is used in a triode emulation that getting the bias just right is the key to emulating the triode tube. So I would think any configuration that uses a looser bias or self bias will work, yes, but not sound as tube like. Which would explain why there are no mass market Thor's or Thunderchief's out there.

[JDoyle]

... to get class-A (curvature) like behavior from a jFET ...

Could someone, ANYONE, explain to me what in the hell this sentence fragment ('class -A (curvature) like behavior') means? 

"... to get class-A (curvature) like behavior from a jFET ..."

...bias it Class A. The definition of 'Class A' is completely independent of the type of amplifying device - if the device is operating for 100% of the output signal swing, it is Class A; no matter if it is a BJT, a tube, a MOSFET, or a JFET.

And it won't be 'class A (curvature) like' either, it will be honest to God, actual, Class A...

[soulsonic]

Anyone here ever mess with overdriving LinCMOS chips like the TLC2262. The SansAmp XXL uses one for distortion.
Got myself a couple CD4007s to play with the other day.

[earthtonesaudio]

Anyone here ever mess with overdriving LinCMOS chips like the TLC2262. The SansAmp XXL uses one for distortion.
Got myself a couple CD4007s to play with the other day.

The 4007's are nice because they're so closely matched, which helps for certain things. 

By the way, found the appnote I was talking about earlier:

http://www.colorado.edu/physics/phys3330/phys3330_fa06/pdfdocs/AN102FETbiasing.pdf

Figure 8.  I tried to re-do this procedure for a J201/9V supply, but I'm not sure if I have it correct.  Others are welcome to try, please share if it works.

[alanlan]

... to get class-A (curvature) like behavior from a jFET ...

Could someone, ANYONE, explain to me what in the hell this sentence fragment ('class -A (curvature) like behavior') means? 

What he means is operating in Class A and with minimal negative feedback thereby producing lots of 2nd harmonic content as a result because vgs is allowed to span a significant portion of the transfer characteristic.