simple discrete jfet and mosfet headphones amp

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Gus

Well-known member
Joined
Jun 3, 2004
Messages
5,249
Location
n
look in the art of electronics for the simple power amp circuit fragment

Vr should be higher than 1/2 Vcc

First there is a 3 or more volt G to S difference needed to turn most MOSFETs on

Next you can "pull" harder to the B+ with the Mosfet (RDSon) than the resistor can pull to ground.

Think current, advailabe each way from the Q point.

You high end might have a drop because of the MOSFET gate cap (forms a RC lowpass) with the 2.2K and output Z of the jfet driver.


You might be better with a 100K linear pot and a input gate to ground R for a more true log curve. Look at geofex.

How much current do you need source and sink for what Z headphones?
 
A 4.7K will cut off more highs. The lowpass -3Db point moves down with > R and >gate C. Some mosfets have a lot of gate cap.

A fast look at the irf510
180pf gate cap, not to bad
2 to 4 volt gate threshold voltage
RDSon .54ohms

The gate cap is not bad for this mosfet so the 2.2K should be fine

I would use a scope and a temp trim pot and resistor network to adjust Vr for fine tune the clipping of the output with the load that you expect to have connected. I would then make the divider using fixed resistors
 
Maybe it's just me but shouldn't the 2200uf cap be connected to the drain of the MOSFET part (like the JFET setup) and not the source? This is an Nchannel part, it will not like being used like you have it, plus the body diode might give you trouble too but that is just from looking quickly.

IRF says 4v for the gate for full turn on but I will counter that and say that my use of mosfets in the real world tells me that non logic level parts need at least 5v of stiff voltage to be "on" at low current levels. raise this and you will need higher voltage on the gate.
 
OK, I apologize beforehand if any of this is remarkably stupid, but from my somewhat limited understanding of this solid-state stuff (I'm a toob guitar amp guy, and not really good with them), but since the MOSFET is (apparently) a source follower, wouldn't it be simpler to ground R1 and connect R4 directly to the MOSFET gate, eliminating C2 and R5, and using the JFET's drain to bias the MOSFET gate (since the JFET's drain needs to be biased pretty close to V+/2 for maximum headroom, right?)? That's what I'd do if these were triodes - what am I missing?
 
v/2 is not max symetrical output swing with this type circuit. Think about source and sink currents.

Yes one could direct couple the jfet drain to the gate of the mosfet but then you would need to have work out the output swing of the jfet and the bias point of the mosfet. Then how does the different harmonic structures add up? Then the big mess jfets are all over the place.

Look for Paul S's preamp article in an old issue of Glass Audio

I did not note the gate R going to vr. I would ground the gate R and do a split source resistors setup one to help set the gain and the other resistor with a big bypass cap to adjust the drain voltage to what is wanted or maybe something else

vishay fet apps will help

Now there are many ways to do things like this. You could even use BJTs
 
Back
Top