Problem increasing GBW of two stage fully differential amplifier

[Sudeep Banerji]

Hi guys

I have made a two stage opamp with pmos differential input and CS stage for the second stage in .18u technology. I got a GBW of 800MHz and I am trying to get near 3GHz. Equation says GBW=gm/Cc and despite increasing gm to even 2mA/V the GBW does not increase further. Any advice on how to increase the GBW. There is no constraint on current and right now the tail current is 200uA. Any help is appreciated!

 

[Dominik Przyborowski]

What about second stage load? Mentioned equation work quite well for very well poles separation.

 

[csolis_GT]

If you increase the gm by increasing the transistor sizes, keep in mind that they will add parasitic capacitances and they might become comparable to CC. This means that if you set the dominant pole with CC, the equation GBW = gm/CC assumes that secondaries pole all beyond (higher frequencies) than GBW frequency, in this case 3GHz. It seems that to operate at such high frequencies, your biasing current should be in the order of mA.

 

[D.A.(Tony)Stewart]

Did you use 100 Ohm differential Amp on both input and output or single ended?

Did you test unity gain or gain >4? Or so....

is your test set lossless to 3GHz and calibrated for RL>15dB

 

[Sudeep Banerji]

Thanks for your replies guys! Oh I did not pay attention to the assumption behind the equation. I guess the phase margin is not good enough. Yes I did get better results with higher current. Thanks!

 

[BigBoss]

Cascode differential amplifier can provide this GBW product.