The simulated inductance value is the product, L = R1*R2*C. 1H = 1 ohm *
1 Meg * 1 uF. If you have megohm resistors for R2 you definitely need picoamp input bias/offset currents since this thing is DC coupled, but generally op-amps that have picoamp input bias currents are the last ones that you'd want to put resistances in series with their outputs as in this circuit, they have enough trouble driving heavy loads as it is.
The whole point of using a large value series floating inductor, simulated or otherwise, as opposed to a large value floating resistor is you want to push low-frequency current through that "inductor." A floating simulated inductor that is not actually good at pushing low-frequency current into a heavy load is t*ts on a bull! That's what I think!
Give this circuit a light load to work into on either side with signal coming in from the other and it performs OK in a sim, but that's just easing up on its job requirements to make it look better.
ge
urn
Even
a 1it
or
son
not
as
m *
Oh, Duh, thanks... I didn't even do the dimensional analysis.
George H.
If the op amp supply rail connections went to any place except the supply rails, this might be true. All that the buffer does is deliver more current than the original output (and adds delay around the feedback loop, but not much).
It's difficult to imagine that this is going to make any network "non-reciprocal", whatever that might mean.
-- Bill Sloman, Sydney
It's only 1 ohm from each terminal to an op amp output.
Another cool thing about the circuit as shown is that when connected to load resistor in the fashion of an RL lowpass filter it has gain, depending on op-amp and input frequency up to 30dB worth.
I took a little time yesterday experimenting with the circuit in Spice as he drew it using similar type op amps to the picoamp input bias current device he suggests using. I think it's worth checking out what the sim shows for that even with input swings of just a couple volts on one side from a source working into a 1k load.
I still contend Bob came up with this one off the cuff and it looked clever so he showed it, but that there's a reason it appears that after this episode of the show he never brought it up again, building floating gyrator-inductors that have good performance isn't a trivial design problem and there's a fair amount of literature/journal papers about it on the 'net.
I guess what I'm arguing is that a grounded inductance gyrator simulating a large-value inductance doesn't have to actually sink appreciable currents to be a useful device, it just needs to present and inductive-looking impedance, but that a large-value floating gyrator that can't drive a load very well for anything but small signals is not a particularly useful device.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.