CFA question

Hey, click on "data sheet"

Analog Devices has really trashed their web site.

Il 31/05/2015 19.02, John Larkin ha scritto:

That's strange. Anyway here it is the datasheet

Bye Marco

Yes but now I can design my car on my iphone. Yay.

Oh look actually it's changed now, instead of a car advert it is a giant robot hand.

And social network instead of designing circuits. Or buying ICs.

CFB op-amps, we like to call them. As you know this -12dB/octave characteristic is common for any op-amp configuration with a second gain pole close to its rolloff frequency. CFB op-amps can have that just like VFB types. Some peaking in the gain response is also common.

If it's excess high-frequency phase shift that's bothering you, there's a clever solution, called "active compensation", see AoE-III, Section 5.8.7, page 314. This involves adding an identical opamp in the feedback loop in a specific fashion, and it can extend the phase response by nearly a decade.

I recall seeing that in an ancient AD app note (AD-107, 1987). It's not on AD's web site any more, but I snagged a mirror long ago before they trashed their site:

It's more of a frequency-domain trick, because its behaviour near the corner frequency isn't as nice.

Cheers

Phil Hobbs

OK. Now it is clearer. I've suspected that, but I've found no confirmation.

Oh yes. I've encounterd some stability issues in my application and suspected that the culprit was the CFB amplifier. Indeed, now I've learned that my first modeling of the circuit was a little bit wrong. if the suggested solution,"active compensation"is the same pointed out by Phil Hobbs, I've used it in the past in one project and learnt from the same AN.

Anyway, my copy of AoE-III sholud knock at my door next days. Thanks for the help. Bye

Marco

Yes, I referenced James Wong's app note in our book.

I found a small capacitor across R2 tamed the beast in Spice modeling. I haven't tried it on the bench. James Wong says the technique is best for gains > 5.

Interesting, thanks. I've never used it either, though I was intrigued enou gh to remember it all this time--there's usually more than one source of ph ase funnies in an instrument, so I never seem to be able to avoid a calibra tion step of some sort.

Cheers

Phil Hobbs