Non-Inverting Unity Gain Buffer

[Unlikekurt]

Hello

For a non-inverting unity gain buffer, figure TL072, with a bipolar supply.  Is there any necessity to have input/output resistors?  If not, is there a strong argument for their addition?

Thanks
James

[antonis]

For input, absolutely none..

For output, no other than current limiting..

(we're talking about signal series resistors, of course..)

[Unlikekurt]

Thanks Antonis,
That's sort of what I was thinking, but at input and output.  Good to know I can leave the input open.
That said, on the breadboard I don't seem to have any issue omitting both in application but wanted to correct as necessary.
Is there a simple formula for the calculation for the output R; should I simply measure the AC current in application?
Best
James

[antonis]

You can roughly determine that resistance by Power Supply voltage (considering output hitting supply rails) & op-amp maximun allowed working current..

In fact, the above is overkill (due to there should be some kind of output shunting resistance) but you can use it as a starting point..

[Unlikekurt]

Antonis,

Ahh gotcha.  Make the resistor sink whatever the limiting maximum of the IC is?

James

[Mark Hammer]

Is this intended o be a buffer before any other pedals, or a buffer between pedalboard and amplifier?

[Unlikekurt]

Mark,
It's located at the output of a preamp stage, before a reverb stage.
It's a buffer before the signal splits between the dry line and the reverb driver stage.

James

[amptramp]

When running from bipolar supplies, there should be a high-value resistor from the non-inverting input to ground unless you are DC-coupled to the previous stage.  This keeps leakage currents from shifting the bias all over the place.  There is no need for an output shunt resistor.

[Unlikekurt]

Amptramp,
Ok, so 1M or so pull down at input, but would it require a series R at the output?

[PRR]

> Is there any necessity to have input/output resistors?

Resistors are SO cheap, it is not worth your mental strain to figure if you can omit them. At least in DIY.

You need "some" DC path at the input so the opamp "knows" where to sit. In a bare-bones g-amp, this "can" be through the guitar. Unless it is an active guitar. Which might have a DC-bleed resistor, or might not. In general you build your circuit with a minimum of assumptions about what the User might connect it to.

Likewise a few-K series resistor protects the delicate chip inputs from rude external voltages. (RG may go into detail here. http://www.geofex.com/circuits/what_are_all_those_parts_for.htm )

Output to ground resistor is not essential on any of the usual opamps. However all opamps have DC error, the small stray DC at the output may distress the next box in the chain (if it was not defensively designed), so it is conventional to have an output cap. Which leaks, so you add a bleed resistor.

When an output cable is shorted, the opamp tries to deliver infinite current. A '709 opamp is easily killed this way. '301 and later chips are normally "protected", but a small series resistor reduces strain.

When an opamp drives capacitance, notably a loooong piece of cable, it oscillates. A few hundred Ohms in series lets the opamp see a benign load.

When in doubt, STEAL all the bits of a known-good plan. Much easier than thinking.

[ElectricDruid]

When in doubt, STEAL all the bits of a known-good plan. Much easier than thinking.

Plagiarize! Let no-one else's work evade your eyes! Remember why the good Lord made your eyes, so don't shade your eyes, but plagiarise, plagiarise, plagiarise!

Only be sure always to call it please "Research"!

It turns out to be a very good method for electronics as well as mathematics, and certainly describes most of the history of stomp boxes, both commercial and DIY.

[Gus]

From 2011
https://www.diystompboxes.com/smfforum/index.php?topic=93699.0

read reply 2 for the buffer.
Some good posts in that thread.
As PRR posted resistors do not cost much.

If you look in books etc. sometimes you will also see diode protection on the input

[PRR]

> Only be sure always to call it please "Research"!

OMG! Research says Tom is still alive!!
https://en.wikipedia.org/wiki/Tom_Lehrer
"Born   April 9, 1928 (age 91)"
also: https://www.nature.com/articles/d41586-018-03922-x

[Unlikekurt]

Thanks to everyone on their feedback on this.
I wasn't having much luck finding a schematic with a unity buffer with +/- rails.  But I guess I can just look at any non-inverting gain stage with a bipolar supply.

[antonis]

I wasn't having much luck finding a schematic with a unity buffer with +/- rails.  But I guess I can just look at any non-inverting gain stage with a bipolar supply.

I'm not sure exactly what you're looking for so I'll violate a bit R.G.'s draw ( ) on http://www.geofex.com/circuits/what_are_all_those_parts_for.htm, already posted by Paul.. 
(Rf is left there for perfectionists..) 

[Unlikekurt]

Thanks Antonis,
That's basically the ticket.  For instance, let's say I went with a 100K ground reference at the non-inverting input, is the purpose of adding a 100K in the feedback loop for compensation?  And my initial query was whether or not I needed a series resistor at the output of the op amp?

[antonis]

Feedback resistor of the same value with bias one is only needed for op-amps of relatively high input bias current..

Output resistor is only needed when op-amp drives low enough value loads (e.g. a pair of hard clipping diodes shunted to GND) or a capacitive load, as Paul already told you..
>When an opamp drives capacitance, notably a loooong piece of cable, it oscillates. A few hundred Ohms in series lets the opamp see a benign load. <

[Unlikekurt]

Thanks Antonis!