JFETs too consistent for you? Try this biasing technique.

[earthtonesaudio]

[unverified]

Here's an idea for a kind of pseudo-automatic bias for a JFET.  The two LEDs should be the same color and type, pointing at each other but otherwise shielded from other light sources.
The "backwards" one is supposed to generate a negative voltage for the N-channel JFET gate.

Oh, and pay no attention to the identifiers on the JFET and LED, those were just the default options that showed up in CircuitLab.

[J0K3RX]

Interesting!

[Gurner]

[unverified]

Here's an idea for a kind of pseudo-automatic bias for a JFET.  The two LEDs should be the same color and type, pointing at each other but otherwise shielded from other light sources.
The "backwards" one is supposed to generate a negative voltage for the N-channel JFET gate.

Can you explain how that happens?! (cos I just can't see how it can - doesn't mean it can't...I'd just like to be enlightened!)

Also, you've instantly  lost a couple of volts of headroom across D2? (& when you're only setting out with 9V in the first place ...and possibly a battery ....that doesn't sound too appealing)

[Gus]

470 ohm drain resistor?  lets guess 4VDC drop across the 470 ohm.  4VDC/470= 8.5ma drain current.

When I saw this post I looked in a bookcase.  I first  read about leds being used as source and sensor in Forrest M. Mims III "Led Circuits &Projects" ISBN 0-672-21006-1 page 77 my book is a second printing 1975.

[slacker]

Can you explain how that happens?!

LEDs are light sensitive, if you shine light at them you get a small voltage generated across them. Alex's schemo has the anode connected to ground, so in theory you get a small negative voltage at the cathode.

[Gurner]

Can you explain how that happens?!

LEDs are light sensitive, if you shine light at them you get a small voltage generated across them. Alex's schemo has the anode connected to ground, so in theory you get a small negative voltage at the cathode.

Thanks - presumably(!) the LED is going to need a certain amount of stable/dependable light before this can be relied upon - so, what about ambient light variances dark gig, light bedroom, murky rehearsal hall ...or the enclosure itself being lightproof?)

Also, I'm not sure that this eliminates biasing woes - because that's going to result in a fixed negative gate voltage depending on the LED (& different LEDs have different fwd voltages)...... the biasing required for a jfet depends upon the jfet itself ...putting something that's a fixed (& even then something that in itself depends on a certain amount of light to work), will likely be far from optimum?

[earthtonesaudio]

Headroom concerns:



Ambient light concerns:

The two LEDs should be the same color and type, pointing at each other but otherwise shielded from other light sources.

Biasing concerns:

The two LEDs should be the same color and type, pointing at each other but otherwise shielded from other light sources.

Also note the sarcasm in the post title:

JFETs too consistent for you?

Anyway, you'd be trading the standard JFET bias problem (JFETs are highly variable) for a very non-standard bias circuit which may* automatically bias the JFET at a good operating point for audio.

*If you're lucky, OR: if you're clever at choosing the LED and quiescent current AND the fact that the biasing is part of a negative feedback loop is sufficient to overcome JFET variability.  This last part is still up in the air.

[edit] Changed the schematic so the LED is turned off by the BJT instead of on.  

[alparent]

So you would be switching from FET matching to LED matching. 

[m4j0rbumm3r]

It's a pretty cool idea. 

Have you actually tried it? If so, on what and with which results?

[earthtonesaudio]

The only test I have done so far is shine a bright white LED at a couple other LEDs and measure the voltage output with a meter.

[diode description]: [forward voltage], [reverse voltage with external light source]
diffuse green: 1.4, 1.2
clear orange: 1.5, 1.3