Question on JFETs

Who says they can't amplify in the active region? As amplifiers, there's probably less distortion in the active region than ohmic.

I don't understand. In the active region they are fully saturated so they're like either on or off; there's no in-between?

Active region is where you have enough voltage supplying the drain , normally above the gate voltage, by biasing the gate voltage you can select a constant current. In this mode if the voltage on the drain increases for example, the darin current will remain the same per gate bias. Gate/Source

Problem is that this active region really isn't very linear so it does have specific needs or use.

WHen the voltage btween the drain and source come low enough to where current can no long be maintain via the gate bias, it becomes ohmic. when you hit the saturation level due to lack of drain/source voltage the gate bias now makes the fet an ajustable R. Variable voltage at the drain will simply make the ohmic fet behave like an R.

That is best I can explain it off the top of my head.

Jamie

They operate like triode tubes except that they can handle AC as log as the drain voltage doesn't get too close to the gate voltage.

If you have the source ad the gate at the same voltage then it is in conduc tion, which could be called saturation. It will be at its Rdson within reas on. Note that you cannot put alot of voltage to it or it will blow.

In an N channel device when you take the gate voltage negative with respect to the source it increases resistance. It is not exactly like a resistor b ecause changes in drain to source voltage affect the current non-linearly, but for many cases it is linear enough that they can be use for example, fo r audio compression without much distortion and without a delta gain stage. I got a design for an amp that uses JFETs for clip proofing. I have not bu ilt it but in theory it should work fine. And the polarity is simply revers ed for P channel JFETs, which have no equivalent i tubes.

Then when you get to the pinch off voltage, drain to source will act as an open circuit. This is the equivalent of a tube's cutoff voltage.

Most transistors cannot do that. Most of them will conduct forward bias col lector to base and will destroy the transistor. Even the new ones like IGBT s and whatnot cannot handle AC linearly, or close thereto. And then some of them cannot be operated in the linear region at all because of their struc ture. There are also depletion mode MOSFETS available but they also will no t operate linearly on AC from drain to source.

I find actual JFETs to be getting sparse in the market. If you go to like D igikey and see how many FETs they have and compare that with the actually J FETs they sell you will find them scarce.

As such, they might not be the best choice for new designs. They might get phased out completely in time. You will have to move a hell of alot of prod uct to be able to get ONSEMI or whoever to keep making them, they want to s ell what the market buys.

I actually have, at this time a design that makes a workaround really diffi cult. I have looked for JFETs and found that I am going to have to use like six pairs of them in parallel to do the job, and this is small signal. It' s just that any other solution adds so much complexity it is almost not wor th doing.

It is almost like tubes. Russians use tubes in the front end of RADAR array s so they are less vulnerable to an EMP. But they do manufacture tubes and sell them worldwide now that most others have stopped. They serve the audio phile and professional music markets. Perhaps we can persuade them to make JFETs ?

But the bottom line is that JFETs can do things no other semiconductor can do. And it is not simple amplification.

Do you have a copy of AoE (the 2nd ed. is enough) Jfets and fets are close cousins and basically behave the same. (Jfets are like depletion mode fets)

George H.

Ancient terminology problem. "Saturated" for a BJT equals "ohmic" for a FET. FET "saturated" equals "normal bias" for a BJT.

FET amplifiers are almost always in normal bias ("saturated"), so that the transistor functions as a not-too-great current source instead of a not-too-great resistor.

Cheers

Phil Hobbs

FET "saturation" and BJT "saturation" are two different things. A BJT is saturated when the collector voltage is lower than the base voltage. A FET is "saturated" when the drain current is relatively insensitive to drain voltage. I cannot, for the life of me, remember what is "saturated" in each of these two cases. Some digging around on the Internet or (gasp!) in some good old 3rd-year circuits books should get you answers.

So a FET in saturation acts like a BJT that is _out_ of saturation: variations in drain voltage don't have a great effect on the drain current, and the drain current is modulated by the gate-source voltage, so you can make an amplifier by working the drain into some resistance.

I have found that particular book to be one of the most confusing in this regard. MOSFETs and JFETs are treated in the same section with the narrative switching from one to the other most times without adequate differentiation. They should have been treated *completely separately* (not without comparison to each other though) but completely separately nonetheless. It turns out that itty-bit of gate insulation MOSFETs have makes for extreme confusion with JFETs if these devices are both treated under the same heading; it's false economy.

That sentence makes no grammatical sense to me at all.

Jfets are like depletion mode _MOS_fets

Sounds like you should take up some other trade. ...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | STV, Queen Creek, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

I'm looking for work... see my website.

See figure 3.8 in AoE. I haven't used a lot of jfet's so I'm not much of an authority.

I thought H&H did a good job of making clear what could be a confusing subject.

(I haven't used depletion mode fet's either... except playing around with mkaing 'em into current sources.)

George H.

I love it when microchip foundry processes include a depletion mode MOSFET... they make fabulous kick-starters for bandgap references, then disconnect from rail-induced current variations. ...Jim Thompson

Hmm.. not sure I fully follow. Hey! You could write a book! Phil's working on his second, you don't want a snooty PhD leaving more of a legacy. :^)

George H.

I'm more into trade secrets that pass on to the offspring >:-}

E.G. My most recent OpAmp model is a cryptic (but not encrypted) equation whose coefficients are derived from unpublished calculations. ...Jim Thompson

Like this (-:

.SUBCKT OpAmpCore IN+ IN- OUT VP VN PARAMS: A=57.739m B=4.01

• Q=4.319 Z=1.5915 H=1 L=1 RG=200K CC=1.5915nF SH=22 SL=22 G_GM VN OUT VALUE {A*SINH(B*TANH(Q*V(IN+,IN-)))} R_RGL OUT VN {RG} R_RGH N_1 OUT {RG} C_CC OUT VN {CC} G_GH OUT VN VALUE {Z*exp(10.3465*TANH(SH*(H-V(VP,OUT)))-9.6535)} G_GL N_1 OUT VALUE {Z*exp(10.3465*TANH(SL*(L-V(OUT,VN)))-9.6535)} E_ISOL N_1 VN VP VN 1 .ENDS OpAmpCore

...Jim Thompson

In that case, would they make good drivers for laser diodes, do you think?

They're "on" (current flow = IDSS) at VGS=0, if that rings your chime, requiring a negative VGS to turn them off.

Where I've used them for bandgap "kick-starters", drain is tied to VDD, gate to VSS, source to load (BG). When BG comes up it turns off the kick-start current. ...Jim Thompson

AoE is brilliantly clear.

They behave very much alike: wiggling the gate voltage modulates the conductivity of the source-drain channel.

You seem to not understand fundamentals, like voltage and current. You might consider backing up and getting the basics right; that will make things like semiconductors easier to understand.

Some will never understand. There might be something in the DNA of people who do, or those who make nuclear bombs, or whatever.

When I run across people who don't get it I use the headlight theory.

Your car has two headlights which pull(ed) 36 watts each, being three ohms or whatever. You put two of them in parallel it pulls 72 watts, double.

Then there are these little transistors that control the big transistors. That is how it works folks.