# In a cascode current source why the current buffer transistor is immune to temparature fluctations

#### Himanshoo

Joined Apr 3, 2015
260
Since the text says that collector voltage of Q1 is maintained constant by Q2(buffer) in order to cancel harms of early effect...as Vce of Q1 doesn't change with load voltage....

If suppose temparature fluctuates causing a variance in Vbe of both transistor ....then how can current buffer transitor Q2 is able to make collector voltage of Q1 constant since...it should change with a change in Vbe?

#### Attachments

• 30.3 KB Views: 62

#### #12

Joined Nov 30, 2010
18,190
It doesn't. Your drawing says, "improved stability with load voltage variation", not, "perfect stability with temperature variation".

#### Himanshoo

Joined Apr 3, 2015
260
It doesn't. Your drawing says, "improved stability with load voltage variation", not, "perfect stability with temperature variation".
but load voltage variation is somewhat related to Vbe variation...

#### #12

Joined Nov 30, 2010
18,190
So what? That circuit is still, "improved" compared to only one transistor as a constant current source.
Do you want someone to explain why, "improved" doesn't mean, "perfect"?

#### crutschow

Joined Mar 14, 2008
24,362
but load voltage variation is somewhat related to Vbe variation...
Load voltage variation refers to the change in current with a change in load voltage.
It says nothing about a change with temperature -- and that circuit will change current with temperature.

#### Himanshoo

Joined Apr 3, 2015
260
So what? That circuit is still, "improved" compared to only one transistor as a constant current source.
Do you want someone to explain why, "improved" doesn't mean, "perfect"?
no matter whatever could be the cause of variation whether its load voltage or temparature..the thing which i didnt get in the circuit is that if suppose Vbe changes(whatever the reason could be)..then how can Q2 which is acting as a current buffer maintains a constant collector voltage for Q1..
In order to maintain a constant collector voltage for Q1..., Q2's Vbe should remain fairly constant ..but if with any load voltage variation or temparature ...Vbe tends to change..but the text doesn't support my inference..

#### Himanshoo

Joined Apr 3, 2015
260
Load voltage variation refers to the change in current with a change in load voltage.
It says nothing about a change with temperature -- and that circuit will change current with temperature.
pls refer the attachment.....load voltage and temparature variation are probably somewhat interelated..

#### Attachments

• 35.2 KB Views: 32

#### #12

Joined Nov 30, 2010
18,190
The text does not support your inference because the text does not intend to support your inference. The text explicitly denies your inference and demonstrates that in post #7. Why do you ask, "How can this not change with temperature?" when your text (and I and crutschow) says it DOES change with temperature?

#### crutschow

Joined Mar 14, 2008
24,362
If you want a current source that cancels much of the Vbe variation with temperature then you can use a current-mirror circuit.

#### Himanshoo

Joined Apr 3, 2015
260
If you want a current source that cancels much of the Vbe variation with temperature then you can use a current-mirror circuit.
I appreciate your input ...but i have explictly stated my query in post 6..

#### Alec_t

Joined Sep 17, 2013
10,913
if suppose Vbe changes(whatever the reason could be)..then how can Q2 which is acting as a current buffer maintains a constant collector voltage for Q1.
It can't. It provides an approximately constant collector voltage.

#### Himanshoo

Joined Apr 3, 2015
260
It can't. It provides an approximately constant collector voltage.
exactly yes .....it can't ..but then how come it tends to make collector voltage approximately constant.

#### t_n_k

Joined Mar 6, 2009
5,447
@Himanshoo
Presumably you understand that if only the load resistance itself changes and all other parameters (such as Vbe & hfe) are unchanged, then Q1 collector voltage is more stable than would be the case if Q2 was not part of the circuit.

#### Himanshoo

Joined Apr 3, 2015
260
@Himanshoo
Presumably you understand that if only the load resistance itself changes and all other parameters (such as Vbe & hfe) are unchanged, then Q1 collector voltage is more stable than would be the case if Q2 was not part of the circuit.
no i want to ask that even though Q2 is the part of the circuit how come it stablizes Q1's collector voltage to a constant level if Vbe of the transistor changes?

#### AnalogKid

Joined Aug 1, 2013
8,308
Since the text says that collector voltage of Q1 is maintained constant by Q2(buffer) in order to cancel harms of early effect...as Vce of Q1 doesn't change with load voltage....
If suppose temparature fluctuates causing a variance in Vbe of both transistor ....then how can current buffer transitor Q2 is able to make collector voltage of Q1 constant since...it should change with a change in Vbe?
First of all, there is no thing as a perfectly constant current source or sink. There just plain is not, and unless you change your thinking on this you never will be satisfied with the answers your are getting. All constant current circuits are based on some kind of feedback. Since transistors and opamps do not have infinite gain, the output current never will be perfectly constant. It's way better than a plain resistor, and the term "constant current" is perfectly true in the real world. But all current source outputs vary with load conditions, temperature, age, barometric pressure, and other variables.

The collector voltage of Q1 is ***NOT*** constant. It just isn't. Q1's collector voltage WILL CHANGE WITH TEMPERATURE for at least two different reasons. Accept this as true or continue to be frustrated. Both transistors' Vbe will decrease with increasing temperature, and this will change both collector voltages and the circuit's current sink value. Adding Q2 to the circuit reduces (but does NOT eliminate!) one minor source of error caused by variations in the load impedance, but that error has *nothing* to do with the temperatures of the two transistors.

ak

#### crutschow

Joined Mar 14, 2008
24,362
I appreciate your input ...but i have explictly stated my query in post 6..
And we have explicitly given you an answer which you seem to be unable to understand.

#12

#### #12

Joined Nov 30, 2010
18,190
Having let this work in my brain for several hours, I have come to the conclusion that, if you can not tell the difference between, "does" and, "does not" then you should try to find a forum in your own language. That will help with the translation problems.

#### Himanshoo

Joined Apr 3, 2015
260
First of all, there is no thing as a perfectly constant current source or sink. There just plain is not, and unless you change your thinking on this you never will be satisfied with the answers your are getting. All constant current circuits are based on some kind of feedback. Since transistors and opamps do not have infinite gain, the output current never will be perfectly constant. It's way better than a plain resistor, and the term "constant current" is perfectly true in the real world. But all current source outputs vary with load conditions, temperature, age, barometric pressure, and other variables.

The collector voltage of Q1 is ***NOT*** constant. It just isn't. Q1's collector voltage WILL CHANGE WITH TEMPERATURE for at least two different reasons. Accept this as true or continue to be frustrated. Both transistors' Vbe will decrease with increasing temperature, and this will change both collector voltages and the circuit's current sink value. Adding Q2 to the circuit reduces (but does NOT eliminate!) one minor source of error caused by variations in the load impedance, but that error has *nothing* to do with the temperatures of the two transistors.

ak
As we vary the load impedance (lets say decreases) more voltage gets dropped across the fixed internal resistance of power supply which causes a decrease in voltage across the load...
is this the error which you are talking about??

#### #12

Joined Nov 30, 2010
18,190
As we vary the load impedance (lets say decreases) more voltage gets dropped across the fixed internal resistance of power supply which causes a decrease in voltage across the load...
is this the error which you are talking about??
No, no, and no.
Everyone here assumed the supply voltage was perfectly constant because there was no mention of it in the circuit description. You have brought in another irrelevant factor to confuse the issue.

#### Jony130

Joined Feb 17, 2009
5,145
No, we add Q2 into the circuit to reduce the Early effect. Change in VCEQ2 (due to load voltage variation) will do not have any effect on Q1 collector current.