Best Input differential configuration in amplifiers.

kASD

Member
2004-12-25 12:24 pm
N/A
Hi everyone,

Just fascinated by the various topologies used at input stages of an amplifier, such as,
1 Single Ended Differential
2 Dual Differential
3 Diamond Differential
4 Cascode loaded Differential [dual and single ofcourse]
5 Cascaded Differential


What is the best configuration for an amplifier input stage according to your experience.

So , guys lets start another thread dedicated to our differential stage topologies.



cheers,
Kanwar
;)
 

kASD

Member
2004-12-25 12:24 pm
N/A
The diamond differential............
 

Attachments

  • dd.jpg
    dd.jpg
    10.1 KB · Views: 1,486

zinsula

Member
2004-03-12 8:42 am
Replace the bjt's with J-Fets in your diamond differential, get rid of the biasing diodes, and so you have my favourite.

You might cascode those to use higher than 15-20V rails though.

Cheers, Tino

BTW, something similar to your diamond differential is used in LM6172 and LT1364, afaik.
 
Just my opinion: A-Standard Configuration is "good enough" up to 150W-200W. By this I mean that the improvements afforded by other configurations are incrementally small and likely disappear under XO-distortion and the noise floor.

However, higher power amplifiers generally mean higher CL gain which could be enough to make the improvements more easily audible. It has been remarked on, and I've leard it myself, that some very high power amps have enough gain to make even thermal (Johnson) noise audible when idle.

Going back to the standard configuration -- For self constructed amplifiers, fewer components means fewer oportunitities to screw up, fewer component to fail, fewer potential bad solder joints, fewer things to investigate when the &%$$^&*=% doesn't work, fewer component pairs needeing matching. It means a smaller PCB footprint which makes it easier to do a good enclusure layout and wiring scheme. If the topology is single sided rather than mirror image, a current mirror can be added which may well provide benefits virtually equal to the more complex circuits.

If I had the time I would probably try building all the other configurations just because they look cool. I must remind myself that I should allocate some (MORE) of my time to listening to what I build rather than let it all be totally absorbed by the next great idea.