[SOLVED] Linearity of a Fully differential Wideband amplifier

[harpv]

Hi,

I'm designing a fully differential wideband amplifier and one of the requirements is a THD of -42.5dB. How do I set the fundamental frequency for THD calculation in Cadence if my unity gain bandwidth specification is 82.5 Mhz, DC gain is 12 dB, VDD = 3V and O/P swing 2.1V - 0.9V?

The architecture uses a source degenerated folded cascode input stage to improve the linearity. Are there other ways to improve the linearity of the circuit?

Thanks,
Hari

 

[erikl]

Usually the THD fundamental frequency is given by the main operating frequency of the amplifier. If it is actually operating as wide band amplifier, there isn't such a main frequency - but the center of your wide band application could be used as such.

If you wish an honest THD info about distortion of your amp, you should use a fundamental frequency such that at least its 3^rd harmonic (THD "total" then means "third") still fits into the full gain bandwidth of your amp, i.e. 6..7 MHz in your case.

Linearity can be improved by larger gain plus larger feedback. Non-linearity decreases proportional to the ratio of (closed-loop-gain/open-loop-gain).

 

[harpv]

Thanks for the reply erikl.

So since the source degeneration is a feedback mechanism any circuit which employs this technique can be said to be in a closed loop? In my case the folded cascode amplifier by itself will have a large gain and we are source degenerating it and reducing the gain thus increasing the linearity. Please correct me if I'm wrong.

I've taken a fundamental frequency of 10Mhz, and the 3rd harmonic will fit into the full GBW of the amp, or will there be a problem with this or how accurate will the THD measurement be?

Regards
Hari

 

[erikl]

So since the source degeneration is a feedback mechanism any circuit which employs this technique can be said to be in a closed loop? In my case the folded cascode amplifier by itself will have a large gain and we are source degenerating it and reducing the gain thus increasing the linearity. Please correct me if I'm wrong.

No, I think you are right: degeneration is a feedback mechanism, indeed a local one, i.e. working inside one stage only.

I've taken a fundamental frequency of 10Mhz, and the 3rd harmonic will fit into the full GBW of the amp, or will there be a problem with this or how accurate will the THD measurement be?

Well, here you consider mainly the 3^rd harmonic, which certainly is the most important one. The second-important, the 5^th harmonic is in a frequency range where the gain is already about 6dB down.

For comparison you could try with a much lower fundamental frequency - say 1 MHz - but I guess there won't be a big THD difference.