At 09:17 AM 06/12/2012, you wrote:
>Spehro Pefhany writes:
> > Ah, the classic n-channel JFET.
> >
> > The one in the middle is the gate. The other two are interchangeable.
> >
> > (for example, ON Semi and Fairchild show the reverse pinout--
> > S-G-D and D-G-S respectively)
>
>         Wow! I plan to use this as a current regulator for the
>LED in an optoisolator so I guess I can just tie the gate to the
>source (negative) and the load will be between the remaining
>free pin (drain) and the positive terminal. The odd thing is
>that in this particular circuit, I have the 2N3819 stuck in to 3
>pin holes of an IC socket so I could flip it either direction
>and it will probably behave the same way. Correct?

Yes.

>         My understanding of this circuit is that the gate is one
>diode drop above the source when shorted in this manner and that
>is what gives you the current regulation.

Not quite. The gate is at the source potential, since that's
where it is connected, so you get Idss, provided the voltage
is high enough. A JFET so connected looks like a resistor for
small Vds voltages, and for high enough voltages they look more
like a current source. The voltage-current curve rises from
zero V, zero A with a constant slope (like a resistor), then
flattens off as the voltage increases (current doesn't increase
much for increasing voltage).

BTW, Idss is not a well-controlled parameter- at 15V across
the JFET, it's only guaranteed to be between 2mA and 20mA.
"High enough" is when it's approaching the cutoff voltage,
so maybe in the 5V range. You can put a resistor in the
source and reduce the current. Or use an LM317L regulator
with a resistor.

>The other question I
>have is if you connected the positive side of the power supply
>directly to the drain and put the LED between the source-gate
>and the negative terminal of the power supply, would anything
>really be any different?

Nope. It's a two-terminal current limiter, so as long
as it's in series (and connected with the right polarity),
all is well.

>Would you still have 10 or 15 mills
>regulated current between the source and negative terminal? If
>that was a simple resistor instead of the JFET, you would have
>the same current at any point in the series circuit except there
>would no longer be any regulation. This is based on Norton's law
>concerning current in a series circuit.

You're right.


>Anyway thanks for the information.
>
>Martin McCormick

Just out of curiosity.. how can you read a schematic?

--sp

--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.