Best practises for scaling voltages and buffering outputs?

[Sound]

[hexagon5un]

[Sound]

Hi again! I will check the Art of Electronics regarding this matter.

I understand the 1K resistor as a protection, but! the subject here as suggested before by Blue_hell, is for example on an inverting amplifier, tap the feedback resistor after the 1K output resistor, in order to track perfectly the voltage input.
The output tracks perfectly the input, without drop.
The output is protected. Isnt it?
But the impedance is almost 0

The truth is that I (still ) don't understand exactly how works the impedance and maybe the question does not have sense.
But my question was if theoretically there is a need of have determined impedance in the output. Is it needed? Maybe is good have same output/input ratio impedance in all the modules?

[blue hell]

An output impedance of 1k may not seem much, but when loaded with 100 k about 1% of your signal will fall over the 1 k resistor so the circuit following sees 99% instead of 100% .. now you could compensate for this, but then when you connect two inputs to that output about 2% will drop instead ... so there is another 1% that would need to be compensated away. Now this 1% may not seem much but its enough to mess up your tuning (if it happens to be a VCO frequency control signal).

Still it is a good idea to have an output resistor after an opamp, if not for short circuit protection then for stabilty ... not all opamps will need this, as most are short circuit proof and a lot can deal with odd loads as well.

So IMO when you include the resistor it's best to get the feedback from after that resistor to still have the resistor and also have a very low output impedance.

But ... when you never work with multiple inputs connected to an output and you always use an input impdance of 100k you could compensate for an output impedance of 1k ... and you can make a choice whthrer to compensate the inputs or the outputs ... but this is messy IMO.

[Sound]

Good! I didn't thought in that! So , one output with the feedback resistor taped after the 1K output resistor you could feed as many inputs as you want: You will have no voltage drop. So, sending whatever signal to a passive multiplier, you could split that signal and always will track OK the input!

This should be the standard. isn't it?

[Sound]

Thinking in this:

About the inverting amplifier with the feedback resistor after the 1K output resistor:

If we short the output of an inverting opamp with the feedback resistor after the 1K output resistor, the opamp, looking for its equilibrium, would provide all the current as possible to the output.

So It will be the same than a simple opamp without the 1K protection output, Isnt it?

And, If an opamp provide all the possible current to its output, could drop the current in other points of the circuit?

About the inverting amplifier wuth a trimmable resistor as a feedback:

Could be any inconvenience using a trimmable resistor as a feedback? maybe we compromise stability?

[blue hell]

[blue hell]

[Sound]

Thanks again Blue!
Interesting thread this one.

[Sound]

More thoughts about:

In a standard inverting amplifier output with its 1KΩ output resistor, it's observed and measured that the voltage out drops after the 1KΩ when it is connected to an input of other inverting amplifier.
If nothing is connected to its output there is NO voltage drop after the 1KΩ resistor.

Actually, I think we can understand this like a voltage divider to the virtual ground generated in the inverting input of the second inverting amplifier.
Also notice that for example using 100KΩ resistors for IN and feedback, the second inverting amplifier or input inverting amplifier will have 101K as a resistor IN and 100K as a feedback resistor.
And it causes a drop in its output, regarding the previous module.
It could be solved with resistor feedback 100K+1K in series and matched 0.1% with Rin. But a trimming feedback resistor will solve it and also will compensate all the unmatched resistors of all the opamps involved in the chain of the processor.

See that besides the need of match the gain, 1% resistors are not appropriate , see for example:
Inverting amplifier with Rin=99K and Rf=101K
Input 5V will output ~5,1 V Its a difference of 0.1V
Input 10V, will output ~10.2V Its a difference of 0.2V

Consider that one semitone is equal at 0.083V.

Notice also that now we can see track 1/Voctave signals its more dificult because it depends of the configuration output of the source module and the configuration input of the destination module.

So regarding 1V/oct tracking we should organize the modules in the following categories:
*V/octave generators (midi to CV interfaces, VC keyboards, etc...)
*Voltage processors (mixers, spliters, inverters etc...)
*Destination modules (VCOs...)

*So for V/Oct sources best seems voltage followers without resistor output.
If needed inverting outputs Rf and Rin should be matched and if needed limit the current could use output resistor but the feedback resistor should be tapped after it.
For split that signals the better solution is a passive multiplier. There will not be drop. with the limitation of the amount of current that can provide the opamp output.

*For V/Octave processors its more complicated because the correct tracking will depend of the output configuration of the source and the input configuration of the destination module.
A good option could be an inverting output with trimming feedback resistor with the feedback resistor after the limiting current resistor output.
But if its trimmed fed with a module source that has the standard configuration output with its 1K output resistor it will not track perfectly sources without the 1kΩ resistor output and vice-versa.

*For V/Oct destination modules, I mean Oscillators, same than before, depending with the configuration output of the module that the oscillator was trimmed, it will track properly or not, other sources.

I would like add that all will track V/Oct signals if ALL the modules had same standard + trimmed Rf output.

How do you see it?

For me was useful I noticed for example that my Midi to V converter (Cwejman S1) seems that there is not resistor output and the best way for split that signal is with a simple passive multiplier. Will be very useful for my Teezers!