Maker Pro
Maker Pro

compond transistor and Hfe

F

f

Jan 1, 1970
0
After a long search I have found a description of the circuit in
question. It is the Sziklai configuration. The circuit in question is at

http://www5b.biglobe.ne.jp/~houshu/synth/EgMos0210a.GIF


involving Q7 and Q8. The designer has selected the two transistors where
the dominant feature appears to be that Hfe does not change with
variation in emitter current. Equivalent types are KSA733 and KSC945.
What is the advantage of using a transistor with constant Hfe in this
configuration? Is there also an advantage when used in the Darlington
configuration?


TIA, f
 
H

Helmut Sennewald

Jan 1, 1970
0
f said:
After a long search I have found a description of the circuit in
question. It is the Sziklai configuration.

Hello Frank,
is there any reference of a circuit to this name(Sziklai)?
The circuit in question is at

http://www5b.biglobe.ne.jp/~houshu/synth/EgMos0210a.GIF


involving Q7 and Q8. The designer has selected the two transistors where
the dominant feature appears to be that Hfe does not change with
variation in emitter current. Equivalent types are KSA733 and KSC945.
What is the advantage of using a transistor with constant Hfe in this
configuration? Is there also an advantage when used in the Darlington
configuration?

It is absolutely insignificant that both transistors have the same Hfe.

Both transistors should have a high Hfe, because of the 1MOhm resistors
in the base of Q7. THat's all.
There is really no more secret for choosing Q7 and Q8.

Best Regards,
Helmut




He
 
R

Robert Baer

Jan 1, 1970
0
f said:
After a long search I have found a description of the circuit in
question. It is the Sziklai configuration. The circuit in question is at

http://www5b.biglobe.ne.jp/~houshu/synth/EgMos0210a.GIF

involving Q7 and Q8. The designer has selected the two transistors where
the dominant feature appears to be that Hfe does not change with
variation in emitter current. Equivalent types are KSA733 and KSC945.
What is the advantage of using a transistor with constant Hfe in this
configuration? Is there also an advantage when used in the Darlington
configuration?

TIA, f

I notice that Q2 does absolutely nothing except take up space and
deplete the pocketbook.
Q7 and Q8 is a complimentary darlington configuration, and Q7 could be
a small signal transistor and Q8 could be a power transistor.
Absolutely nothing indicates that any transistor has a "constant"
beta, and i seriously doubt such beasties exist.
The "standard" darlington configuration cannot be saturated.
 
F

f

Jan 1, 1970
0
Robert said:
I notice that Q2 does absolutely nothing except take up space and
deplete the pocketbook.
Q7 and Q8 is a complimentary darlington configuration, and Q7 could be
a small signal transistor and Q8 could be a power transistor.
Absolutely nothing indicates that any transistor has a "constant"
beta, and i seriously doubt such beasties exist.
The "standard" darlington configuration cannot be saturated.

The mosfets are part of a cd4007ub so there is no choice except to short
it out if it is unneeded. The datasheets for Q7 and Q8 show dead nuts
flat hfe vs collector current over at least two orders of magnitude.
Since the circuit is a linear piecewise function generator and the
output stage does not have any negative feedback, I assume that the
types of Q7 and Q8 were selected for the beta characteristic, so that
the buffered output voltage would remain a linear function of the input
current of Q7. For the Darlington, the beta of the composite is the
product of the hfes, and I guess that is true for Sziklai too. If the
betas of the individuals in a Darlington vary widely over the range of
operation then I would suspect that the beta of the Darlington varies
even more widely.
 
H

Helmut Sennewald

Jan 1, 1970
0
f said:
The mosfets are part of a cd4007ub so there is no choice except to short
it out if it is unneeded. The datasheets for Q7 and Q8 show dead nuts
flat hfe vs collector current over at least two orders of magnitude.
Since the circuit is a linear piecewise function generator and the
output stage does not have any negative feedback, I assume that the
types of Q7 and Q8 were selected for the beta characteristic, so that
the buffered output voltage would remain a linear function of the input
current of Q7.

Hello F(rank),
this is a big misunderstanding of the circuit of Q7 and Q8.
Q7 and Q8 just make a buffer(voltage amplifier) with a voltage gain
of one. The base current of Q7 is about 10mA/(Hfe1*Hfe2).
As long as this base current multiplied by the 1MOhm resistor is a
low value, the buffer doesn't influence the waveform.
You can replace it with a high beta darlington transistor too.

Best Regards
Helmut
 
F

f

Jan 1, 1970
0
Helmut said:
Hello F(rank),
this is a big misunderstanding of the circuit of Q7 and Q8.
Q7 and Q8 just make a buffer(voltage amplifier) with a voltage gain
of one. The base current of Q7 is about 10mA/(Hfe1*Hfe2).
As long as this base current multiplied by the 1MOhm resistor is a
low value, the buffer doesn't influence the waveform.
You can replace it with a high beta darlington transistor too.

Best Regards
Helmut

Thanks, I forgot about the local feedback from the emitter resistor. It
makes sense now.
 
Top