Help on bias multiplier

[dinstam]

Hello,
My knowledge in electronics is limited and I need help.
In Bob Cordell’s book on page 49, in the part about “Output stage bias current”, it is noted that, the 2 drivers and the 2 output transistors of Figure 2.12 require at least 4 Vbe voltage drops. In the schematic the Vbe multiplier around Q5 delivers 3.3xVbe ? What am I missing?
Thank you

(Thank you Bob Cordell for your excellent book.)

[lineup]

It is rather normal
3.3 x Vbe can be enough to set a low current in Class AB output stage.
For Class A we need more than this.

[AndrewT]

the temp comp value for the semiconductors varies with current.

If the outputs and drivers and Vbe multiplier are all operating at different currents, even if we adopted identical devices for all 5 locations, then we find that precisely 4times is not the correct temp co.
If the Vbe does not measure the full range of output or driver temperature change then it will under compensate. More likely the drivers do not change too much and thus the Vbe over compensates for the output devices and a corresponding reduction in Vbe sensitivity to temperature brings the temp co to near what is required by the circuit over a range of operation conditions.

[dinstam]

Hello lineup and AndrewT and thank you for your replies.
Bob Cordell uses a 518 Ohms resistor across the base-emitter of the multiplier transistor (R7). He must have done some calculation which I cannot follow. This a not a preferred resistor value.
Regards

[ostripper]

Originally Posted by dinstam

Hello lineup and AndrewT and thank you for your replies.
Bob Cordell uses a 518 Ohms resistor across the base-emitter of the multiplier transistor (R7). He must have done some calculation which I cannot follow. This a not a preferred resistor value.
Regards

That 518R would most likely be 330-470 fixed resistance plus a 500R trimmer (20 turn variable resistor) , in series .

Cordell uses a 10ma source for the voltage stage. below 1 is a 8.8ma source.
The variables are A. The source current , B. the gain of the Vbe device , and C. the ratio of the B-C to the B-E resistance.

2sc3503 and MJE340 will be on average - a 2.3 to 1 ratio (B-C / B-E - for 50 - 100ma op bias).

Observe what the values are between the 8.8ma and 10ma examples below.
Real life amps reflect these bias points within a few percent. Both examples set a BJT OP stage to 75-80ma per device.

PS - for a real circuit you would start with the 500R trimmer at maximum value (499R )

OS

[dinstam]

Hello ostripper,
thank you very much.
Regards