Why not recreate an ancient representative of the first integrated operational amplifiers and track down the developers?
Have a discrete op-amp at hand as a universal genius - and all this with exactly the components that you yourself consider to be the best at this point.
What could you /we conjure up from this? From an audio preamplifier to a power amplifier, to an active reconstruction filter for our countless D to A converters or even a sensitive, low-noise thingamajig for RIAA equalization - anything is possible!
Why don't we start a small dimensioning challenge,
at the end of which perhaps this year's DIYAUDIO (i.e. 2024) operational amplifier will emerge?
Here's a suggestion:
Fire away, but don't look in vintage data books and just copy - the topology is very well known and has been frequently integrated and sold as a standard, worldwide.
Kind regards and good luck,
HBt.
Have a discrete op-amp at hand as a universal genius - and all this with exactly the components that you yourself consider to be the best at this point.
What could you /we conjure up from this? From an audio preamplifier to a power amplifier, to an active reconstruction filter for our countless D to A converters or even a sensitive, low-noise thingamajig for RIAA equalization - anything is possible!
Why don't we start a small dimensioning challenge,
at the end of which perhaps this year's DIYAUDIO (i.e. 2024) operational amplifier will emerge?
Here's a suggestion:
Fire away, but don't look in vintage data books and just copy - the topology is very well known and has been frequently integrated and sold as a standard, worldwide.
Kind regards and good luck,
HBt.
Attachments
If you are building a circuit using discrete components, i.e. not on a monolithic IC chip, then feel free to use any and every discrete component available.
Use both PNP bipolars and NPN bipolars. Both are widely available and circuit designers can take advantage of this. To name one example, the lowest voltage noise bipolars are PNP and the lowest current noise bipolars are NPN.
Use JFETs. Their strengths and weaknesses are different than bipolar transistors. Exploit this.
Use MOSFETs.
Use Schottky diodes. Use Light Emitting diodes. Use Current Regulating diodes. They're discrete semiconductor devices, they're in stock today, and they're useful.
Use ferrite beads. Use three terminal "feedthrough" capacitors. Use inductors and transformers. Use Zener diodes. Use precision resistors. Use any discrete device that offers an advantage.
Use both PNP bipolars and NPN bipolars. Both are widely available and circuit designers can take advantage of this. To name one example, the lowest voltage noise bipolars are PNP and the lowest current noise bipolars are NPN.
Use JFETs. Their strengths and weaknesses are different than bipolar transistors. Exploit this.
Use MOSFETs.
Use Schottky diodes. Use Light Emitting diodes. Use Current Regulating diodes. They're discrete semiconductor devices, they're in stock today, and they're useful.
Use ferrite beads. Use three terminal "feedthrough" capacitors. Use inductors and transformers. Use Zener diodes. Use precision resistors. Use any discrete device that offers an advantage.
Thank you Mark,
everything you say reflects the motivation and fascination (of such an undertaking). Among other things, it answers the question of why it might be worth having your /our very own universal (perhaps limited to audio) circuit available.
In addition to the fact that such a learning medium simply generates joy and fun, you learn a lot about circuit technology and the associated components.
Regards,
HBt.
everything you say reflects the motivation and fascination (of such an undertaking). Among other things, it answers the question of why it might be worth having your /our very own universal (perhaps limited to audio) circuit available.
In addition to the fact that such a learning medium simply generates joy and fun, you learn a lot about circuit technology and the associated components.
Regards,
HBt.
I have included the missing push-pull output stage (of the original) so that “the open collector” need not lead to further questions or fixations on the part of seafaring nations ..!
#
This is an interesting and alternative idea (solution) to the well-known LM741, in my opinion.
#
This is an interesting and alternative idea (solution) to the well-known LM741, in my opinion.
Attachments
For me most interesting circuits for discrete clones of monolithic IC op amps (in order to sonic character) are such with only one gain stage within the NFB loop like AD817, AD825, OPA2604 etc.
and such based on a current-conveyor topology like the for I/U converters good known AD844
https://www.diyaudio.com/community/threads/current-conveyors.163039/
https://www.diyaudio.com/community/threads/current-conveyor-as-a-voltage-amplifier.223835/
and such based on a current-conveyor topology like the for I/U converters good known AD844
https://www.diyaudio.com/community/threads/current-conveyors.163039/
https://www.diyaudio.com/community/threads/current-conveyor-as-a-voltage-amplifier.223835/
How is this discrete op-amp so different than just a SS power amplifier, with the NFB parts pulled out and the inverting input brought out instead to an input pin? The transistors are smaller in the "push-pull output stage"? The transistors are smaller in the driver to the "push-pull output stage"? The quiescent current is less?
To be honest, I don't understand your question (right away).
The schematic is basically nothing more than a map, a roadmap - in other words, a basic schematic that can be adapted or designed to meet individual needs.
Only the component selection and the dimensioning ultimately determine the application, it could be a logarithmizer or a power amplifier ... or a low-power module, an operational amplifier, an analog calculator...
The basic structure is the same or similar to that of an ordinary power amplifier. But such a module is much more (valuable) as a learning medium.
Only from this point of view does such an undertaking make sense, otherwise I might as well buy a bag of leftover A109, µA709 ... at auction. But none of us would want to use this ancient OP as an integrated component.
But who knows what else you can get out of the actual core?
The schematic is basically nothing more than a map, a roadmap - in other words, a basic schematic that can be adapted or designed to meet individual needs.
Only the component selection and the dimensioning ultimately determine the application, it could be a logarithmizer or a power amplifier ... or a low-power module, an operational amplifier, an analog calculator...
The basic structure is the same or similar to that of an ordinary power amplifier. But such a module is much more (valuable) as a learning medium.
Only from this point of view does such an undertaking make sense, otherwise I might as well buy a bag of leftover A109, µA709 ... at auction. But none of us would want to use this ancient OP as an integrated component.
But who knows what else you can get out of the actual core?
both an integrated and discrete op amp so as a power amp are identical except for the selected quiescent currents of the individual stages and the deliverable peak current of the output stage.
This term I don't understand:
with the NFB parts pulled out and the inverting input brought out instead to an input pin?
according the wanted conditions one can always choice both non inverted mode and inverted mode - thus sometimes both the non inverted input pin and the inverted input pin can serve for the signal input and NFB input.
Got it.
Just thinking of an op-amp as a 3 terminal device (sans power pins), while a power amp is two - the feedback mechanism embedded somewhere in the whole circuit. Rather than brought out to a pin you can connect that up to yourself. The closest is the old tube power amps with 4, 8, 16 ohm taps and a wire for the feedback that you're supposed to connect to the tap used.
There's a long thread that went on for years that I presume has lots of good info, will this new thread have improvements?
https://www.diyaudio.com/community/threads/discrete-opamp-open-design.218373/
https://www.diyaudio.com/community/threads/discrete-opamp-open-design.218373/
I can't tell you that.
If the 175 pages of the thread, which has been running for over a decade, deal with exactly this topology, then perhaps a layout and dimensioning will be found there, perhaps, perhaps not.
Hybrid circuits use individual discrete transistors to construct opamps, but the transistors are bare die rather than packaged parts. The finished opamp units are small; bigger than an IC but much smaller than a daughter card PCB. They allow designers to pick and choose best-in-class components from many different fabrication processes, without the compromises of monolithic construction.
National Semiconductor's LH0032 (L for linear ; H for hybrid) was a hybrid opamp sold in the 1970s. Its datasheet gain-bandwidth product was 0.7 Gigahertz and its datasheet slew rate was 500 volts per microsecond. Today in 2024, Apex Microtechnology builds and sells the PA96 hybrid opamp, with a gain-bandwidth product of 0.17 Gigahertz, a slew rate of 250 volts per microsecond, a maximum supply voltage of 300 volts, and a maximum output current of 1.5 amperes.
_
National Semiconductor's LH0032 (L for linear ; H for hybrid) was a hybrid opamp sold in the 1970s. Its datasheet gain-bandwidth product was 0.7 Gigahertz and its datasheet slew rate was 500 volts per microsecond. Today in 2024, Apex Microtechnology builds and sells the PA96 hybrid opamp, with a gain-bandwidth product of 0.17 Gigahertz, a slew rate of 250 volts per microsecond, a maximum supply voltage of 300 volts, and a maximum output current of 1.5 amperes.
_