As I still had the prototype board set up from looking at the open-loop gain, I thought I'd try another quick
experiment. This time I'm going to attempt to measure the output resistance. I'll do this by back-driving the output
of the DUT [device under test] with a function generator.
What I'm measuring here isn't the effective resistance you'd see with feedback operating around the op amp, where the
excess gain would usually reduce the apparent output resistance to something well under an Ohm, rather it's the
actual resistance that's sitting there in the IC. Although it's often not a concern - to the point many datasheets
don't even give you a typical value for it - it has a couple of minor consequences in a real circuit, so it's worth
being aware that it's there.
This is the circuit I'm using.
I've taken what I had for the open-loop gain test, removed the 1V-step part, and added a connection to a function
generator via a resistor. The 1k resistor is the one being used for the measurement, the 10k is just there because
the generator isn't all that good and works better outputting a large signal. Think of the 1k resistor and the
internal resistance as forming a potential divider. Measuring both ends of the measurement resistor I can calculate
the current and then use that with the voltage measurement, at the output of the DUT, to calculate the output
resistance.
My test assumes that I can model the output of the op-amp simply as a voltage source with a resistor. The real device
will be more complex than that but I'm only after a rough approximation, so this will do.
Here's the board ready to test.
Here is the result at 1kHz. The blue trace is the drive signal to the 1k resistor. The yellow trace is the other end
of the resistor at the output of the DUT.
The integrator doesn't do a wonderful job of holding the average output of the DUT at zero, but it's just about good
enough for me to calculate a resistance value. If my calculations are correct, that gives a resistance of 46 Ohms. I
wouldn't have too much confidence in that, though the OP27 datasheet gives a typical figure of 70, so it's not too
much out of the way.
It doesn't seem to vary much with frequency - I'm seeing a similar sort of figure all the way from 500Hz up to 10kHz.
At low frequencies it goes off, but that's because the integrator increasingly starts to notice it and tries to
correct for the change.
References
[2] Simple Op Amp Measurements. James M. Bryant.
Analog Dialog, Volume 45, Number 2, 2011.
https://www.analog.com/media/en/analog-dialogue/volume-45/number-2/articles/volume45-number2.pdf
If you found this interesting and would like to see more blogs I've written, a list can be found here:
/members-area/b/blog/posts/jc2048-blog-index-new-version
