Ash Small's Rule.....Regarding JFET's / General Chatting / Forums

Forums

4hv.org :: Forums :: General Chatting

« Previous topic | Next topic »

Ash Small's Rule.....Regarding JFET's

1 2 next

Move Thread

LAN_403

Ash Small

Wed Jan 31 2018, 02:25AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Ash Small's rule when biasing JFET's is as follows.....Bias the gate to - one third rail to rail voltage, thus leaving one third plus and one third minus headroom......And thanks to Steve Connor to turning me on to the potential of JFET's......BTW, don't use a source resistor for biasing, use diodes or LED's .....More to follow.....

Conundrum

Wed Jan 31 2018, 04:29AM

Registered Member #96 Joined: Thu Feb 09 2006, 05:37PM
Location: CI, Earth
Posts: 4062

Thanks for sharing!
There are nice small FETs in old electret microphones, they often have sub volt switch.
Incidentally has anyone hacked the "breath sensor" found in E-cigs yet?
These have other uses notably for the disabled.

I found a few the other day, its best to use new ones because of the contamination issue.
They can be found incredibly cheaply but careful disassembly is needed.
The mic unit can be found at the "business end" of e-cig, where the mouthpiece goes.

Sulaiman

Wed Jan 31 2018, 11:31AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

Ash Small wrote ...

Ash Small's rule when biasing JFET's is as follows.....Bias the gate to - one third rail to rail voltage, thus leaving one third plus and one third minus headroom......And thanks to Steve Connor to turning me on to the potential of JFET's......BTW, don't use a source resistor for biasing, use diodes or LED's .....More to follow.....

In many circumstances I may agree, but not as a general purpose rule-of-thumb.

diodes and LEDs are electrically noisy,

and have a dV/dT that may upset your biasing
and why would you need a +/- 1/3 voltage swing on the gate when the source is at an almost fixed voltage ?.

P.S. my go-to jFET is 2N3819, 100R source resistor, 10mA nominal Idq, c1V at source, gate near 0V..

Ash Small

Wed Jan 31 2018, 01:03PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I found the -6V thing using trimpots, it seems that's where everything evens out, lots of gain and plenty of headroom.

I'm using J105's, nominal current 500ma. I had 134 Ohms on the source, which turned out to be 6V @ 50ma.

I wondered about noise, I can't see anything on the 'scope, I've not tried a sound test yet, I want to play about with the drain resistor some more, and check linearity, etc, but everything is looking good so far.

JFET's are strange devices, so many differing opinions on how to use them. Most seem to want to set the gain to a pre-determined value. This doesn't make sense to me, I'm inclined to set it to it's mid-point, and let it do what it wants.

It's an incredibly simple circuit, one cap and one resistor on the drain, and a diode string on the source, with another resistor to set input impedence on the gate. No need for an input cap with this circuit, and no need for a by-pass cap on the source.

Ash Small

Wed Jan 31 2018, 06:04PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

New information:-

So I've been tinkering with this for a while, and I can confirm that the point where there is maximum gain with zero distortion, at 12.5Hz, with an input signal of 1 Volt peak to peak, is precisely 6 Volts across the drain resistor (why do I keep wanting to call it an anode resistor?).

It balances beautifully at this point, -6V on the gate, and +/-6V headroom.

Current is 24.5ma.

Ash Small

Tue Mar 13 2018, 12:07AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

So I got this all boxed up, I still need to play with it for a bit, but I'm getting some cool vibes from it

Edit: Some pretty cool asymmetric soft clipping, and a bit of that voltage sag thing, the S shaped thing. I just need to learn how to control it

Edit: I'm thinking 'wide knees'

Edit: There is no input cap in this circuit, it doesn't need it.

Hazmatt_(The Underdog)

Wed Mar 14 2018, 03:19AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

Yea, you need to be careful with high gain designs and plastic boxes, I usually pick up AM radio stations when I build this way.

A die-cast Aluminum box would offer some shielding over plastic.

Ash Small

Tue Apr 24 2018, 06:36PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I don't have a lot of gain, around 8 without the boost capacitors, but i'm constantly tinkering with it. No problems whatsoever with the plastic box and noise/interference.

I've now established that capacitors improve the sound (diminish noise) when using diodes to bias the gate, and they even increase gain, which the textbooks say isn't supposed to happen, but i'm using something like 220uF, I think.

Hazmatt_(The Underdog)

Wed Apr 25 2018, 01:16AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

Capacitors do increase gain in textbooks, if you look into it.
The reason being is you have DC current flowing through a resistor to bias an amplifier, and a capacitor bypassing that resistor. At DC the transistor "sees" normal bias conditions, up until you apply AC and the amplifier "sees" a short or a very low resistance, which is why in a common source or common emitter amplifier will have much higher gain with a bootstrap or emitter resistor shunted by a capacitor than without.
If you do a Bode plot of your amp with vs. without you can see the effect on the gain.

Ash Small

Wed Apr 25 2018, 05:12PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Hazmat yes....but I'm using diodes (LED's) on the emitter/source to bias the gate, I can fully appreciate how the capacitor reacts to a resistor in this location, but I don't understand why there is also noticable gain with a capacitor across a diode string. I do have about half an Ohm to share the current between two strings of LED's, but half an Ohm is negligible.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob