Author Topic: Gain in differential current mirror  (Read 1134 times)

0 Members and 1 Guest are viewing this topic.

Offline sahko123Topic starter

  • Frequent Contributor
  • **
  • Posts: 324
  • Country: ie
Gain in differential current mirror
« on: June 23, 2020, 09:16:52 am »
I'm designing a class a amplifier which requires identical curgent source and sink mirror. Where one element sources current and the other element sinks the same amount of current. This second element will then by biased differentially to drive a load. Essentially creating a dc coupled amplifier in class a configuration. I have 2 questions is it possible to use one bias transistor to bias an npn and pnp transistor to draw and source identical (to within a few uA) and also add gain to the bias transistor so that I don't need to use the same current to bias the transistor as I need to be sourced and sunk. I know this might be phrased in a bit of a weird way but I can't think of a better way to phrase it.

This is what im thinking of trying.
« Last Edit: June 23, 2020, 09:20:45 am by sahko123 »
Asking for a friend
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: Gain in differential current mirror
« Reply #1 on: June 23, 2020, 01:24:05 pm »
You have your emitters and collectors swapped in your schematic.

What you are suggesting is more simply done with separate NPN and PNP current mirrors.

If you want a differential current output, then it is possible to drive an NPN and PNP pair at their connected emitters while adjusting the voltage between their bases to maintain class-a operation.  I designed an audio power amplifier based on this idea which worked fine but the input buffer stage of a complementary current feedback amplifier does the same thing better.
 

Offline sahko123Topic starter

  • Frequent Contributor
  • **
  • Posts: 324
  • Country: ie
Re: Gain in differential current mirror
« Reply #2 on: June 23, 2020, 08:02:50 pm »
Roight this is exactly what i was looking for thanks just didnt have a name. But my other question still stands is there a way to get programmable current gain so that the bias transistor of the current mirror takes alot less current so that i wouldnt need to have 8 power transistors for just 2 audio channels? (I know this would ultimately not be a current mirror but at least a current multiplier)
Asking for a friend
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: Gain in differential current mirror
« Reply #3 on: June 24, 2020, 10:29:46 pm »
Right this is exactly what i was looking for thanks just didnt have a name.

The configuration of 4 transistors as a complementary buffer is called a "diamond buffer".  Sometimes the two input transistors are replaced with diodes.

Two diamond buffers can be used with a resistor between them setting the transconductance.  See below.

Quote
But my other question still stands is there a way to get programmable current gain so that the bias transistor of the current mirror takes alot less current so that i wouldnt need to have 8 power transistors for just 2 audio channels? (I know this would ultimately not be a current mirror but at least a current multiplier)

Usually the current mirrors have a fixed ratio determined by their emitter areas on an IC or by there emitter series resistors in a discrete design which is what I did in my audio amplifier.

So you want a current mirror with a variable ratio?  That would be of limited use because negative feedback from the output should still be used to control distortion.  In any event, there is no inherent way to make the ratio variable.

A resistor in series with the output of the diamond buffer or between two diamond buffers can set the transconductance or voltage to current ratio.  If the input is grounded so that the output is a virtual ground, then a JFET or MOSFET can be used for gain control.  The two diamond buffer configuration is used by some very fast voltage feedback operational amplifiers in place of a differential pair.


 

Offline sahko123Topic starter

  • Frequent Contributor
  • **
  • Posts: 324
  • Country: ie
Re: Gain in differential current mirror
« Reply #4 on: June 25, 2020, 03:45:50 pm »
a current mirror with variable ration would have reduced the space parts take so i get to use smaller parts but im not even sure if current mirrors will be needed
Asking for a friend
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: Gain in differential current mirror
« Reply #5 on: June 25, 2020, 04:49:25 pm »
There actually is a way to make a variable current mirror but I doubt it will save space.

Start with the typical current mirror using emitter series resistors to control the ratio.  Replace one of the resistors with an FET used as a voltage controlled resistor.  Then an operational amplifier combined with a matched FET can control the channel resistance accurately.  As a bonus, the FET only operates in one quadrant and is referred to the positive or negative supply.
 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15801
  • Country: fr
Re: Gain in differential current mirror
« Reply #6 on: June 25, 2020, 04:54:57 pm »
You may want to take a look at this: http://www.freepatentsonline.com/4994730.html
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf