有没有比较懂全差分放大器的，我的trans response出了问题

fuweidavid

在edaboard上也发了帖子，就没有翻译过来了，多谢了

I designed a fully-differential two-stage OTA. It is a telescopic pmos input one(cascoded) with a nmos common source amplifier. The CMFB feedback to the bottom nmos transistor of the first stage.

I have done the biasing and CMFB circuit, I compensated the circuit with lead compensation. The openloop small signal gain is 96dB and unity gain frequency is 100Mhz. Phase Margin is 49 degree. The CMFB Loop gain is 75dB and unity gain frequency is 24Mhz and Phase Margin is 70 degree. The input common mode range is from -2.5 to 0.5. The first stage current is 100uA for each branch. The second stage current is 700uA for each branch. The load cap is 2pf and the compensation cap is 0.9pf

When I test the trans response of the circuit. I added a positive step and negative step 100mV at each input(the common mode input is always working at its range). The feedback circuit is -1 feedback with resistors. The output should be -200mV. However, the trans response has a huge ringing. The peak ringing is higher than -350mV and it keep 250ns. The Slew Rate of the circuit is about 100V/us.(This is calculated by the I/C at each stage). I first thought this is due to my CMFB circuit do not have enough differential input range(the differential input range of the CMFB circuit is about -0.4 to +0.4). However, when I adjust the step to very small like 10 mV. I still get 150% peaking at the trans response. Can any experienced engineer point out what's wrong with my circuit. Thank you very much!!

nool

把共模反馈断开看看，仿真中并没有共模波动吧？
可能是线性度的问题

sejimmu

差分运放裕度小了点啊，GBW内只有一个主极点么？共模放大电路过补偿了嘛，共模GBW可以弄大点吧。。