[sdiy] Non-inverting output buffer

[Andre Majorel]

On 2010-04-27 12:51 -0700, David G. Dixon wrote:

> One thing about putting the short-circuit protection resistor
> between the opamp output and the feedback junction of a follower
> is that it doesn't afford as much short-circuit protection as
> the same resistor beyond the junction.
> 
> For example, a 1k resistor inside the junction will only limit
> the short-circuit current to about +/-13.5V/1k = +/-13.5mA
> regardless of the input , while that same resistor outside the
> junction will limit the current to V_in/1k.
> 
> Indeed, 75R does not really provide any short-circuit
> protection, since the output current would be +/-13.5V/75R =
> +/-180mA, which is well beyond the current which most opamps can
> generate.

OK. I've made some tests and the circuit performs very well. With
V_in = 7.4 V and R2 = 82 ohms, the voltage between the
non-inverting input and the far side of R2 is 0.5 mV. This is true
for no load, R_L = 100 k, R_L = 10 k, and even R_L = 1 k ! If my
so-so voltmeter is to be trusted, that's a gain of 7.3995 / 7.4 =
0.99993 or less than 70 PPM of gain error.

All the above with a TL072, R1 = 3.9 k and no cap.

Tomorrow I'll look how high R2 can go. It would probably be futile
to raise it above R1, though.

-- 
André Majorel http://www.teaser.fr/~amajorel/