jfet preamp - does a jfet exist that'll work at just 5V supply?

[scott wurcer]

Originally Posted by peskywinnets

I did, but equally, you clearly haven't locaked in to my comment about NOT using opamps!

BF862 costs almost nothing and is widely stocked, and rarely has Vp over .5V or so.

[peskywinnets]

Originally Posted by Tekko

What is the application ? Is it an amplifier or just a inverter ? Even the single jfet amplifier needs its gate biased somewhere between GND and Vcc = coupling cap needed.

If its just a inverter, theres ttl and cmos logic which is designed for 3.3 and 5 volts.

It's an audio preamplifier.

If I use a self biased jfet, then the gate biased to ground through a high value resistor, but will actually be negative wrt to the source, therefore no input DC blocking cap needed.

Originally Posted by scott wurcer

BF862 costs almost nothing and is widely stocked, and rarely has Vp over .5V or so.

Thanks...how can I work out what gain can be coaxed out of a jfet...I realise it's related to transconductance, but I'm having diffs gleaning which part of the datasheet that's etracted & what formula to use? I'm looking for a voltage gain of about 3x ...doable?

[zinsula]

What is the voltage swing at the input?
How would you use it? as a buffer, i.e. common drain, or should it amplify, i.e. common source?

[peskywinnets]

Originally Posted by zinsula

What is the voltage swing at the input?
How would you use it? as a buffer, i.e. common drain, or should it amplify, i.e. common source?

voltage swing is about 1V max peak to peak (but averaging about 300mv)....I seek a voltage gain of about 3x out of this circuit.

Just experimenting right now to establish the pinch off & the IDS of the J110 I have here....then I can work out the biasing!

let's see if I have this right....

gate grounded
source grounded
drain 100R

VCC= 5V

voltage drop across the 100R resistor is 0.489V ...therefore an IDS of 4.89mA?

[zinsula]

2N3819 from Vishay is easily available. Check the Datasheet, if you set ID with appropriate source resistor at about 5mA, you should have enough headroom at the input.
But...if you have 1Vpp and you want 3Vpp, this will be difficult. There isn't just enough headroom.
And how do you handle the output? Midpoint will be above ground....

[zinsula]

Originally Posted by peskywinnets

[...]100R resistor is 0.489V ...therefore an IDS of 4.89mA?

Yes.
But if you are going to ground the source, you might be able to swing the negative input half wave, but not the positive....more than 0V G-S will only forward bias the G-S...no good.

[peskywinnets]

Originally Posted by zinsula

2N3819 from Vishay is easily available. Check the Datasheet, if you set ID with appropriate source resistor at about 5mA, you should have enough headroom at the input.
But...if you have 1Vpp and you want 3Vpp, this will be difficult. There isn't just enough headroom.
And how do you handle the output? Midpoint will be above ground....

You make a good point...I'll have to reign the gain expectations back a bit.

[peskywinnets]

Originally Posted by zinsula

Yes.
But if you are going to ground the source, you might be able to swing the negative input half wave, but not the positive....more than 0V G-S will only forward bias the G-S...no good.

But the source will be sitting at a DC level higher than 0V ...so it won't be forward biased on the positive signal swings. So long as the source is biased higher than the largest voltage input swing, then I should be ok?

[Tekko]

Why is DC coupling a requirement ?

Its a common misconception that dc coupling improves bass response.
Also in most audio applications you dont even want frequency response down to DC.

Besides any imaginable audio source will have a dc blocking cap on the outputs anyways.

[jcx]

"not using op amps" is not an engineering design criteria - maybe for a homework assignment

the better way to use engineering design skills is to specify the problem in performance terms, define input, output signals, levels, available supplies, area, cost , power requirements that the application demands

specifying device tech from the start is not a proper performance related constraint - you rule out too many options you may not be aware of