Maker Pro
Maker Pro

impedance of a constant current source

sabahan

Jul 14, 2013
31
Joined
Jul 14, 2013
Messages
31
I'm learning the art of amplifier design,know that replacing a collector load with a constant current source greatly improve the distortion of the amplifier,and now the problem is,how I determine the impedance of the constant current source?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
You determine the impedance of a source by looking at the ratio of voltage to current.

Assume there is a constant V1 volts across the constant current source and that the current through it is I1. Now change the voltage across it to V2 volts and determine the current again. Call this I2.

The resistance will be (V2 - V1)/(I2 - I1) ohms.

First you might like to verify it works on a resistor. Then try the same exercise for a current source and a voltage source.

You may find in one case that your answer is incalculable, but at the same time, easily expressed.
 

sabahan

Jul 14, 2013
31
Joined
Jul 14, 2013
Messages
31
do I measure the Vce of the transistor with the resistor or the resistor only or both?I'm using a led to set the reference voltage
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
If you're trying to determine the impedance of the current source it is the voltage across the current source that it important.

The impedance of a constant current source (at least an ideal one) isn't going to change if it is placed in a different circuit.

Assume an ideal current source and place a 1 ohm resistor across it. The change that for a 2 ohm resistor. In this case the voltage across the resistor IS the voltage across the constant current source (ignoring the sign).

I presume your problem is theoretical in nature. Even if it is not, you need to understand the theoretical construct first and then look at how an implementation differs from the ideal.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
... how I determine the impedance of the constant current source?
You need to clarify your question. Are you asking "How do I calculate what the correct impedance of the current source should be in a particular amplifier stage", or
"How do I calculate the actual impedance of an existing current source that I have a schematic for?"

In either case, you need to post a schematic, and probably a lot more information, before we can answer properly.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Click Go Advanced, then click on the paper clip icon at the top of the edit window. From there it's pretty obvious.
 

sabahan

Jul 14, 2013
31
Joined
Jul 14, 2013
Messages
31
I'm using a LED as reference to set the current to about 2Ma,the CC with be connected to a common emitter amplifier configuration ,it replace a resistor ,so I like to know it resistance value
CC.jpg
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
That's only constant current if you don't change the resistor value. If you want constant current you have to connect the LED's cathode and emitter resistor to 0V and use the collector as your sink pin for the current. I would not use LEDs for the reference as the voltage drop is not very well maintained. LEDs are for producing light not as a reference. Use a zenner diode or band-gap reference or some other more stable method.
Adam
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
I think he is using this circuit in place of the collector reistor in a common emitter amplifier stage. Agree that an LED is not the best voltage reference.

Bob
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
I would be less concerned about the use of the LED as a voltage reference than the fact that a current source won't work if the base bias circuit is connected in parallel with the transistor. To be specific, if you want to use that circuit to source current from the positive rail, you have to use the PNP version of the circuit, so that the base bias resistor can be connected to the 0V rail and the LED current path is independent of the path of current into (or out of) the collector of the transistor.

Have a look at the NPN version of the constant current source (sink). The current is sunk into the collector, and the base bias circuit is supplied from the positive supply rail, independently of the collector of the transistor. In other words, it's a three-terminal circuit; one terminal connects to 0V (the LED and the emitter resistor), one terminal supplies base bias, and the third terminal (the collector) cannot be connected to the base bias terminal. Compare it to your circuit, which has only two terminals. That is not a constant current source.

I think this is what Adam was trying to explain in post #10.

In answer to your question though, a current source doesn't have an equivalent resistance. It can be described as having an AC or incremental resistance, but not a DC resistance.

I guess what you want to know is how to set the transistor's collector voltage, to keep it at the optimum voltage of about half-way between the supply rails. You can't do that with fixed bias, because a current source doesn't behave like a collector load resistor - that's the whole point of your experiment! Using a current source as the collector load, if you over-bias the transistor, it will pull its collector way down towards 0V, and if you under-bias it, the current source will pull the collector voltage way up towards the positive supply.

You need some kind of voltage feedback mechanism to set the transistor bias to a "Goldilocks current" - "not too hot, not too cold, but just right" :)

I don't have time to go into detail now. Tell us more about what you're trying to do.
 

sabahan

Jul 14, 2013
31
Joined
Jul 14, 2013
Messages
31
I read the book "designing audio power amplifier" ,it do mention that replacing the collector load resistor with a constant current source will reduce the distortion cause by the early effect,I get the info of CC from the web,of course they did not mention so much of it,thank you for pointing out the problem,after all, I'm not a electrical nor a electronic engineer
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
a current source doesn't have an equivalent resistance.

Weeeeell... It does. It may be dynamic, and it certainly has an ideal resistance. However I suspect that's not relevant.

I get the info of CC from the web,of course they did not mention so much of it

Have a look for the 2 transistor version of this circuit. It uses a more stable voltage (Vbe of a transistor) to set a voltage reference. One is demonstrated in section 3 (image 3.1) of this: https://www.electronicspoint.com/got-question-driving-leds-t256849.html There is also a link to where it came from which has some more explanation.
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
I read the book "designing audio power amplifier" ,it do mention that replacing the collector load resistor with a constant current source will reduce the distortion cause by the early effect,I get the info of CC from the web,of course they did not mention so much of it,thank you for pointing out the problem,after all, I'm not a electrical nor a electronic engineer
I think (I'm not an expert) that using a constant current collector load would not reduce the distortion due to the Early effect, because the collector-emitter voltage would still be varying over a wide range. Using a current mirror as the load would work, because the collector-emitter voltage would not change significantly as the current through the transistor varied. But then the output transistor in the current mirror would suffer from the Early effect instead, if you ran it into a resistive load.

As I understand it, the reason for running a common emitter amplifier into a constant current load is to increase the stage's voltage gain.

Steve, comments?

a current source doesn't have an equivalent resistance.
Weeeeell... It does. It may be dynamic, and it certainly has an ideal resistance. However I suspect that's not relevant.
I should have said that it's not meaningful to describe a current source as having a DC resistance. It is only meaningful to describe it as having an AC, incremental, or dynamic resistance. Would you agree with that?
Have a look for the 2 transistor version of this circuit. It uses a more stable voltage (Vbe of a transistor) to set a voltage reference. One is demonstrated in section 3 (image 3.1) of this: https://www.electronicspoint.com/got-question-driving-leds-t256849.html There is also a link to where it came from which has some more explanation.
In this case, wouldn't a higher incremental resistance make it more difficult to stabilise the collector voltage? I was considering suggesting that he put a resistor in parallel with the current source, specifically to reduce the incremental resistance, for this reason. But anyway, what's the proper way to stabilise the collector voltage when the collector load is a current source?

Perhaps I've got this all wrong!
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
As I understand it, the reason for running a common emitter amplifier into a constant current load is to increase the stage's voltage gain.

Yes.

I should have said that it's not meaningful to describe a current source as having a DC resistance. It is only meaningful to describe it as having an AC, incremental, or dynamic resistance. Would you agree with that?
Yes, and in fact the original question was of its impedance and the way of measuring it was by means of calculating the small signal response. And that's exactly what you're saying here.

In this case, wouldn't a higher incremental resistance make it more difficult to stabilise the collector voltage? I was considering suggesting that he put a resistor in parallel with the current source, specifically to reduce the incremental resistance, for this reason. But anyway, what's the proper way to stabilise the collector voltage when the collector load is a current source?
In the absence of feedback, that's true. But greater gain allows the feedback loop to exert more control.

The input of an op-amp often employs a long tailed pair. The word "long" in this refers to what is essentially a constant current source. This is how differential amplifiers achieve very high gain in a single stage. See here).
 
Last edited:

sabahan

Jul 14, 2013
31
Joined
Jul 14, 2013
Messages
31
Common emitter amplifier with active load

I'm learning how to design a amplifier,I know that with a constant current source,I can reduce the harmonic distortion,now the problem is,do I place the CCS on the emitter side or the collector side?if I place it on the emitter side,the effective resistance of the emitter will be very high,which make the amplification factor very low,if I place it on the collector side,the resistance will be too great which make the amplification factor too large,if I place a CCS on the emitter and a active load on the collector,their resistance is about the same,which make the amplification factor less than two,I'm I right,or I have miss out something?

All transistor used are MPSA42 and MPpSA92
 

sabahan

Jul 14, 2013
31
Joined
Jul 14, 2013
Messages
31
If I use the circuit on the VAS of the amplifier,since both the early voltage of the active load and constant current source is the same,assuming all transistor are matched,the ro of both of the transistor with be the same sine their current is about the same,ignore the base current,so my question is how can I increase the resistance of the active load,or reduce the resistance of the CCS,?if I need a magnification factor of 10
/?
ignore the load first,or have I miss out something on the calculation?
 

Attachments

  • Cir1.jpg
    Cir1.jpg
    30.7 KB · Views: 92
Top