DIY Power supply for a Szekeres headphone amp

[halojoy]

If you put 2 resistors from V+ (+30) to V- (0V)
They wiil divide the voltage.
If both have same value, the voltage between them will be half, of total voltage. Half from 30 is 15.

If one resistor is 2x and the other 1x,
they will divide 30 volt into 2 voltages
that also is like 2x and 1x.

One voltage is over first resistor, the other voltage is
over the other resistor.
But the sum of both voltage is always 30.

For 2x and 1x resistor. The 2 voltages will be 20 and 10 volts.
In this case the voltage Between the resistors
can be +10 volts or +20 volts.
It depends if you put 2x or 1x resistor at bottom, from 0 volts

[Bricolo]

does the 220Ohm Rg resistor cause no voltage drop between the R2-R3 junction ans G?

[Bricolo]

A last question, before I can start calculating my resistor values (I hope )

At idle (and even with a Vin/=0)
Do we want the mosfet to be in the linear mode, or in the saturated mode?

[Lligior]

Rg should not cause any voltage drop. It will of course but it is very small and can be ignored. It is there mostly to keep the mosfet stable. It reduces high frequency oscilation, and if I remember correctly makes the switching smoother. The mosfet should operate in the linear mode.

[Bricolo]

Ok

Rg causes no voltage drop, since no currant flows through it

So:

Vg=(R3/(R2+R3))*Vcc
Vd=Vcc

We chose a bias current to use (say 200mA)
so, in the datasheet, on the graph of Id vs Vds (this graph is made with many curves, we look where Id=200mA

The curve must cross Id=200mA on the near vertical section (when the mosfet isn't saturated)
Now, we see that many curves, for many values of Vgs, cross the I=200mA line

For safety, we'll chose one that is still near vertical on I=300mA (so the mosfet will still not saturate when we have an input signal)

Let's choose Vgs=3V
We see, that on this curve (Vgs=3V), when Id=200mA, Vds=1.27V (you must zoom a lot on the graph to be precise enough)

To obtain Vgs=3V, with a 12V PS

R3/(R3+R2)=3/12
->we can let R2 to 100K; so R3 must be equal to 300K

If Vcc=12V
Vds=1.27V
V4 (voltage on R4) must be 12-1.27=10.73V
I through R4 is 200mA
->R4=10.73/0.2=54R



Is this correct? (have I understood everything correctly? )




PS: that's something I find strange: Vds<Vgs when the mosfet is in non saturated mode

[traderbam]

Use the Ids vs Vgs (not Vds!). Choose the 25C curve.

[Bricolo]

I've also done this, thirst

It's a mistake!
The Ids vs Vgs curve is for Vds=20V
we are far from this value

[Bricolo]

oops

just a question:



is the transistor in saturated mode on the vertical section, or on the horizontal one?

[Lligior]

None of the above. As a general rule 12V (Vgs) will drive a mosfet into saturation operation. This graph can be used as an aid to determine the bias point. If you look at the graph, think of the Vgs lines as possible bias points. For a good bias point consider the voltage swing of your input, choose one of the Vgs lines and see if your peaks will fit in the range. As you can see from the graph the Vgs=1V and Vgs=.5V have a very different distance between them. If your FET operates in this region you will get distortion. Note: the distance between all of the .5v steps is slightly different, and you will get some distortion because of this. Just try to avoid the really bad regions.

[Bricolo]

I don't understand