VAS current calculation

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
A quick explanation.

The VAS stage operates in Class A. It should be able to drive the load looking into the base ( gate) of the output stage.
The current through the VAS should be a bit higher than this to accomodate any variations in load.
A rough example. If you have one pair of (BJT) output transistors
with driver transistors and a minimum expected load of 4 ohms and a voltage swing of 20 volts rms ( 50 watts into 8 ohms) , the peak current is about 7 amps ( into 4 ohms). If the current gain is about 1000 ( including the driver transistor) , the base current will be about 7mA into the drivers.
A VAS running at 10mA would be a safe bet .
Remember that this should be vaild at HF also and should include all components connected to the VAS stage.
Better you check out a good book like one by Self ( also on the web !).
Cheers.
Ashok.
 
If you have a MosFet output stage the required bias for the
VAS depends very much on the gate charge of the used MosFets
and sometimes additional exterenal gate-drain-caps.

The available current from the VAS will limit the max possible slew rate under any condition.
But normally this is not the critical point as the VAS is usually the
stage with the dominant pole anyway...
Additional you might consider that the gate charge behaviour is strongly nonlinear. So the driving impedance (output impedance
of VAS) should comparably low.

Sorry, I do not have simple rule of thumbs or tested examples....

Bye
Markus
 
OK another attempt.

demons_wing ,
With MOSFETS the load will primarily be the capacitance you see at the gate. This can vary depending on the way the circuit is configured. Lets assume that the total capacitance seen by the VAS stage is C . You now need to drive the highest frequency (f) of interest with a peak voltage of V volts.
So to ensure you have enough slewing capability (dV/dt)at this frequency
dV/dt = I / C where I is the qiescent current of the VAS stage.
dV/dt is also = 2*pi*f*V
So I=C*2*pi*f*V.
The practical value of I should be slightly above this.
I hope this is not an over simplification !
Don't forget that the stage driving the VAS will see the Miller Capacitance at the base of the VAS transistor.

Something more to chew on !
Cheers.
 
Are those methods work also on determining the bias on Differential pairs?

Why there are designs that uses enormous bias current, up to 50mA, if much smaller than that is enough? Some said about linearity in single transistor. How important is this linearity issue?
Or is there an optimal bias, unique for different transistor?

I've experimented with changing VAS bias values, and there is difference. The smaller the bias, the sound seems tobe more detailed. The bigger bias the bandwith somehow like limited (the treble sound is more detailed with less bias current). I use bipolars for diff and VAS. Maybe other device will result differently with bias changing?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.