[sdiy] jfet driver

Michael E Caloroso mec.forumreader at gmail.com
Tue Feb 26 02:55:58 CET 2019

JFETs come in depletion or enhancement types.

The operation mode is very different between those types.


On 2/25/19, David G Dixon <dixon at mail.ubc.ca> wrote:
> Hi Tim,
> Here's how I think about it -- I'm sure the "know-it-alls" will chime in
> and
> tell me how wrong I am, but I know what works for me:
> N-JFETs are ON (like a short circuit from Source to Drain) when the Gate is
> at ground.
> N-JFETs are OFF (like an open circuit from Source to Drain) when the Gate
> is
> held LOW.
> The question is, HOW LOW?  That depends on the JFET.  Some, like the
> 2N4391,
> are fairly unforgiving and will only reliably shut off when the Gate is
> held
> at about -10V or less.  It is best to take these right to the negative
> rail.
> Others, like the 2N5485 (which I see is now "obsolete" on Digikey -- too
> bad, it's a useful one) are more forgiving, and will "usually" shut off at
> a
> Gate voltage of about -4V.
> This assumes that one side of the JFET is held either at ground or at
> virtual ground -- in my experience, it doesn't matter which direction a
> is installed (S and D are interchangeable -- some datasheets will make this
> explicit, others won't -- YMMV).  The shutoff is actually driven by the
> Gate-Source voltage difference, but in most of the circuits where I use
> JFETs, the Source is connected to ground or virtual ground anyway.
> I use 4391 to short-circuit VCO integrators for hard sync.  I drive them
> with a comparator which is wired to generate GND when up and -V when down.
> This works.
> I use 5485 all over the place, but mostly for switching inverter/followers.
> If an opamp is wired up as a differential amp, but with a JFET off of the +
> input to ground, then when the JFET is on, the opamp is an inverter, and
> when off, the opamp is a follower.  This is an extremely useful circuit
> element.  Often, the thing driving this is a square wave which is nominally
> 10Vpp.  For this, I put a diode in front of the gate to remove the positive
> half of the square wave.  The -5V part of the square wave is enough to turn
> off the 5485, but typically not enough to turn off the 4391.
> Hope this helps!
> Dave
>> -----Original Message-----
>> From: Synth-diy [mailto:synth-diy-bounces at synth-diy.org] On
>> Behalf Of Tim Ressel
>> Sent: Monday, February 25, 2019 9:20 AM
>> Subject: [sdiy] jfet driver
>> Hi,
>> It is time to admit: I don't really know transistors that
>> well, much to my shame. I grew up in an op-amp world, and
>> other than driving relays I just don't grok the transistor.
>> I am trying to drive an n-channel jfet with a 5 volt ttl
>> signal. I know this is easily done with a few components. Can
>> anyone point me in the right direction?
>> thanks!
>> --
>> --Tim Ressel
>> Circuit Abbey
>> timr at circuitabbey.com
>> _______________________________________________
>> Synth-diy mailing list
>> Synth-diy at synth-diy.org
>> http://synth-diy.org/mailman/listinfo/synth-diy
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy

More information about the Synth-diy mailing list