Current Mirror Discussion

Glen
you once posted this Stochino very fast power amplifier.

Now I ask you,
what are those BAX16 diodes doing in those mirrors?
Are they there to make worse or better ....

I can not tell.
But my guess, following the discussion, they will make worse ....

Lineup ;) Regards
 

Attachments

  • stochino_mirrors.jpg
    stochino_mirrors.jpg
    29.2 KB · Views: 1,240
I know where to look for CM ideas..
OP-AMPS !! some have CMRR and PSRR of 100db or greater.

the 3 transistor version seen earlier if hooked between the rails
like this:
An externally hosted image should be here but it was not working when we last tested it.

(Collector of third CM tranny goes to V+)

Maybe I'm wrong , but would this not cancel out PS
ripple through the CM increasing PSRR , too.
I always wondered how a dual opamp could maintain
>100db crosstalk margins..
This same technique is used on the STK-xxx multi channel
discrete OP driver chips to allow a 7 channel HT receiver
to run off an anemic power supply and still have acceptible
performance.
OS
 
And here is the mirror used in NE5534. It is one 'Glen Miller' type.

This mirror is of course close to perfect choice,
as in NE5534, when the output is at 2 x Vbe voltage.
But can of course be used with good result,
even if output is at 1 x Vbe, as in AKSA 55.
 

Attachments

  • ne5534_mirror.png
    ne5534_mirror.png
    4.5 KB · Views: 1,036
yes, op amp techniques work in discrete designs (after all they are basically the same circuits). 2 limitations though....
1) monolithic transistors are very closely matched which improves performance far better than is possible in discrete designs. this matching comes from the transistors being in the same piece of silicon and also because they almost always match in temperature.

2) you can't get any transistors with multiple emitters or multiple collectors.
 
1) monolithic transistors are very closely matched which improves performance far better than is possible in discrete designs. this matching comes from the transistors being in the same piece of silicon and also because they almost always match in temperature.
Then we can start using dual devices (2n3807-11 etc.)
From what I see most duals are lower voltage(60v Vceo) so
no problem for a CM . Might have to cascode the LTP to use them
as the input differential.
OS
 
Disabled Account
Joined 2006
Japanese have plenty dual types, the most well known is 2sa1349 and compliment but there are other higher performance ones.

Unclejed, just the other day I was browsing the net and found a audio manufacturer that said he uses multiple emitter high gain transistors in the inputs. I tried finding some but I could only find some Rf higher powered ones, yet there is a lot of literature available on this subject which claims these to be available in discrete form. With more time Ill try finding some in the thousands of pages of google.
 
some years ago (why is it always in the past?) there was a series of japanese parts (2saxxx, 2scxxx) that were internally wired to be current mirrors or diff amps, depending on which part you bought. no emitter degeneration was possible because of the internal wiring. i don't know if those parts are still manufactured - i'd guess not.
someone will have to tell us the part numbers; i don't remember and can't get to my notes with part numbers at the moment.

remember the japanese receivers from the mid 1980's that used a current mirror load for the diff amp stage? a lot of those receivers used these parts, so you might get part numbers from looking at some old (technics and pioneer?) schematics ...

mlloyd1


myhrrhleine said:
anyone know of a chipmaker who makes chip mirrors?
 
mlloyd1 said:
some years ago (why is it always in the past?) there was a series of japanese parts (2saxxx, 2scxxx) that were internally wired to be current mirrors or diff amps, depending on which part you bought. no emitter degeneration was possible because of the internal wiring. i don't know if those parts are still manufactured - i'd guess not.
someone will have to tell us the part numbers; i don't remember and can't get to my notes with part numbers at the moment.

remember the japanese receivers from the mid 1980's that used a current mirror load for the diff amp stage? a lot of those receivers used these parts, so you might get part numbers from looking at some old (technics and pioneer?) schematics ...

mlloyd1


i was thinking about the precision of chips :)
 
ostripper

I have previously used the now obsolete 2N3811A ,2N2920 with excellent results.
Have a look at the Linear Systems LS312,LS313 (NPN) and LS351,LS352 (PNP) They are a bit hard to get at the moment, although I know someone in Austria who was able to get a few from his local distributor. The PNP is a direct replacement for the 2N3811.
If you visit the Linear Systems site, they have a PDF called
"Linear Systems-Product Selection"

SandyK