Unbiased Drain's J-Fets... the same?

[Plexi]

Always that I've built some schematic where the Jfet is unbiased from the Drain (not 100% correct term I guess  ), have used whatever I've had on hand.
At my ears, is the same an J201 than a 2N5457/J112/BF245/etc.

Examples:



Which could be the 'logic' around this?
Due the nonexistent voltage "limiting", they run similar/the same way?

[R.G.]

The circuit uses the JFET as a constant current source. All depletion mode JFETs (that's 99.9%) will let through Idss if the gate it tied to the drain. If you put a resistor in the source and tie the gate to the bottom, it reaches some current value lower than Idss because the resistor and source current make a reverse bias for the gate and lower the current from Idss.

[Plexi]

Thanks a LOT, R.G.
I'm sorry for my totally ignorance, and trouble to traduce: what you mean is that the Jfet isn't reaching the Idss (due the "reverse bias")?

Do I need to tweak the res from S to ground and/or G for each J-fet (J201 ≠ 2N5457 for example) to get the best performance?

In resume: they're the same in this case and voltage input (±9v)?

[R.G.]

I'm sorry for my totally ignorance, and trouble to traduce: what you mean is that the Jfet isn't reaching the Idss (due the "reverse bias")?

No trouble. We all start with the same amount of knowledge: zero. Working to gain more of it is an honorable thing to do.
JFETs are strange, in general. They consist of a bar-shaped area of silicon doped to look like a resistance. This is the "channel", and the drain is at one end, the source at the other. For many JFETs, the drain and source are interchangeable. Without the gate, the drain and source just look like a resistor. In fact, the "Rdson" number listed on the datasheet is the value of this resistance.

This is a Field Effect Transistor, meaning that the electric field of the gate changes the conduction. It does this by producing an electrical field that literally pushes the charge carriers flowing in the resistor channel to one side, pinching it like a garden hose. For complicated electrical field and semiconductor physics reasons, the JFET "pinches off" the channel and the current does not go fully off for most gate voltages, but levels off at increasing voltage. Above a volt or two across the channel, and with a biased gate, the JFET stops acting like a resistor, and acts instead like a constant current source. If the gate is shorted to the source, the current is Idss, literally "drain current with gate shorted to source". This is the maximum current the JFET will conduct under most conditions. So take a JFET, solder the gate to the source, and you have a constant current device.

If you put a resistor between the source and gate, current through the resistor causes the gate connection to be pulled below the source. This "back bias" caused by the resistor lets you set a constant current less than Idss. So yes, the resistor lets you set a current through the drain and source (and through the added resistor) of less than Idss.

This is an interactive setting. At power-on, there is zero voltage across the drain->source, and across the resistor, so the gate, drain and source are all the same voltage - zero. As voltage rises, the drain gets more positive (for an N-channel...) than the source, and current flows. Some voltage is generated across the source resistor, and this back biases the gate a little, but not enough to interfere with the current yet. As current rises, the gate is pulled further negative, until the gate just reaches the point that cuts back on current to just balance at that current and no more. Just like every real JFET has a little different Idss, and VGSoss (the voltage that irrefutably turns the channel off) each will "balance" at a slightly different point.

Do I need to tweak the res from S to ground and/or G for each J-fet (J201 ≠ 2N5457 for example) to get the best performance?

Sadly, yes, if what you mean by "best performance" is "what I intended it to do". Every JFET is different: not only are J201s different from 2N5457s, but every J201 and every 2N5457 is different from every one of the same type number. Any two J201s, for instance, are NOT identical unless they have been carefully selected to be nearly identical. The Idss, Vgsoff, and transconductance are different in each device, and not by a small amount. These parameters may be different by five, six, even ten to one within the same type number. So unless you're skilled at designing circuits to make the differences self-correcting, yes, you will have to tweak each circuit on each JFET. And get different bias points and different currents and gains for your trouble.

In resume: they're the same in this case and voltage input (±9v)?

Sorry - they will need tweaking if you are picky about where they bias. And in 9V circuits, you usually need to be picky about this. This wide spread of characteristics even in the same type number is one reason the electronics industry largely quit using JFETs except in certain very special cases.

[Plexi]

Thanks a lot R.G.
I take my time to read (and traduce) carfully all your helping post.

Do I need to tweak the res from S to ground and/or G for each J-fet (J201 ≠ 2N5457 for example) to get the best performance?

Sadly, yes, if what you mean by "best performance" is "what I intended it to do". Every JFET is different: not only are J201s different from 2N5457s, but every J201 and every 2N5457 is different from every one of the same type number. Any two J201s, for instance, are NOT identical unless they have been carefully selected to be nearly identical. The Idss, Vgsoff, and transconductance are different in each device, and not by a small amount. These parameters may be different by five, six, even ten to one within the same type number. So unless you're skilled at designing circuits to make the differences self-correcting, yes, you will have to tweak each circuit on each JFET. And get different bias points and different currents and gains for your trouble.

In the examples I've posted, we could say that the S resistors can be tweaked to get any desirable Idss (for example 4.5v)?


I'm trying to get "biased" this schematic:
http://www.diystompboxes.com/smfforum/index.php?topic=118936.msg1108692#msg1108692

Now, the doubt comes about the "range" and balance about this Jfets.
As you say, every JFET has a little different Idss, and VGSoss (the voltage that irrefutably turns the channel off) each will "balance" at a slightly different point.

As I remember, I always thought that the S resistor was fixed to each type of Jfet (some kind of limit for each IDss) and then you only tweak the D resistor.