Discrete Line Reciever/Line Driver/Diff Opamp

Hullo.

Was just having a tinker and saw there was a nice way to create a differential opamp with something I was already playing with. It's certainly nothing new but could be used as a reciever, line driver or diff opamp - and certainly can be improved on as well.

edit: You'll probably want to pull both feedback loops around 0V, as I said there are probably many improvements. Happy to explore it further...

Habibi

Why? About the only reason (this century) to make discrete circuit blocks is to tolerate higher PS rail voltages, or provide heavier output drive capability, but it will be a struggle to get close to even inexpensive IC performance.

JR

Well you answered your own question there didn't you. To add to "tolerate higher PS rail voltages, or provide heavier output drive capability", education, experimentation, curiosity...

Forums used to be fun, I thought...

JohnRoberts said:

Why? About the only reason (this century) to make discrete circuit blocks is to tolerate higher PS rail voltages, or provide heavier output drive capability, but it will be a struggle to get close to even inexpensive IC performance.

JR

Click to expand...

I agree with the two examples you gave. 
Discrete parts on the front end can also give better noise performance.  And as Habibi said, education, curiosity etc. 
I'd bet more than half the projects on this forum are built with discrete devices, either in hollow or solid state.
Are these folks persuing mostly pointless endeavours?

The idea you posted Habibi can be expanded further and, again, done right, will perform pretty darn nicely. 
Maybe an integrated ends up having a better ppm THD number, maybe that won't matter though. 

I am not trying to chill your buzz,,,, For educational purposes go for it. Do not expect improved performance.  ICs can throw tens of extra devices for almost no incremental cost.  I used to manage an engineering group so one full time task was to keep my junior engineers from reinventing wheels.
=======
OK for educational purposes here is something completely different, a discrete differential amp I scribbled out years ago using a novel long tail triple (three input devices).



This was a special function circuit to accept two DC coupled mic preamp outputs floating up at phantom voltage (as high as 50V), combining the two differential audio signals and level shifting down to nominal 0V, to scrub off tens of volts. The third input transistor is fed with a combined DC voltage of the other two input transistors scaled down to 1/2.

I never melted solder for this, while I did buy some high voltage capacitors well over a decade ago (since used on different projects.) 

The last time I designed a DOA was back in the 80s, when we couldn't purchase low noise devices, relatively inexpensively, off the shelf like now.

JR

JohnRoberts said:

....This was a special function circuit to accept two DC coupled mic preamp outputs floating up at phantom voltage (as high as 50V), combining the two differential audio signals and level shifting down to nominal 0V, to scrub off tens of volts. The third input transistor is fed with a combined DC voltage of the other two input transistors scaled down to 1/2. ....

Click to expand...

Hmm, 2N5771 has Vceo of only 15V  :

moamps said:

Hmm, 2N5771 has Vceo of only 15V  :

Click to expand...

: :  Do not take the schematic literally (there is probably more wrong with it than that). That just happened to be the first handy pnp in my schematic capture program...  I make no claims for this circuit other than it is novel (AFAIK) and "might" have some utility for something, but I never melted solder for a good reason...  Once you get the audio signal pre-amplified to line level, you can cap couple with decent film capacitors relatively inexpensively.

The three device long tail "pair" has some interesting characteristics, some good, some bad, so far unexplored by me because discrete circuits are like so last century.  8)

JR

Hi JR,

All good. Be careful though, it did come across like you were (as a moderator) suggesting members shouldn't encourage DIY electronics... on a DIY forum 

Can you tell us more about the three device LTP? Did you ever use it in other products?

Habibi

Habibi said:

Hi JR,

All good. Be careful though, it did come across like you were (as a moderator) suggesting members shouldn't encourage DIY electronics... on a DIY forum 

Click to expand...

Don't start listening to me now...  :

Can you tell us more about the three device LTP? Did you ever use it in other products?

Habibi

Click to expand...

No, I never used this but i soaked it in beer many times... Without getting too esoteric  a three device LTP (need a new name for that..) will have a different transconductance relationship than 2 device LTP that can influence stability compensation, slew rate, etc in NF closed loop design.. 

If this idea really had merit some ancient probably would have stolen it a long ago.  8)

JR