How to analyse this circuit?

[wjxcom]

Hi, all: there has a strange circuit in the attachment. In this circuit, I think M10, M12, M37 and M43 can be a OPA, but what is the role of R0-R4?

help me, please!!

 

[mists]

r0~r4 give a input dc level, maybe the input signal is ac coupling

 

[wjxcom]

Hi, mists: yes, you are right. the input single AIP and AIN is differencial single.
but how and why r0-r4 can give a input dc level?

 

[jordan76]

Hi wjxcom,

R0 and R1 forms a voltage divider to provide a dc bias voltage for AIP. R3 and R4 perform similarly for AIN.

I also think the input signal AIP and AIN are probably ac coupled.
BTW, M43 and M47 are just source followers which are biased by M10 and M12 respectively.

Hope it helps you.

regards,
jordan76

 

[wjxcom]

Hi, jordan76: apologize you! but what's the role of R2 and R5? I think these two resistor is unuseless because there have no voltage or current on these two resistors.

 

[venkateshr]

Jordan is correct. The circuit is a typical Pmos version of source follower with loads as current sources and the resistances(except R2 and R5) are used for biasing for ac coupling.
I dont find any usage for resistors R2 and R5 except for prevention againt back gate effect . i.e changing output may affect the input due to the gate drain capacitance .But again here it is voltage follower with gain close to unity so not much effect will be there due to this.
Frequency response will affect due to this addition of resistance. it will make the occurence of the first zero earlier by the factor of 1/R .

 

[jordan76]

Hi, jordan76: apologize you! but what's the role of R2 and R5? I think these two resistor is unuseless because there have no voltage or current on these two resistors.

Hi wjxcom,

Indeed,R2 and R5 will not be of any effect on dc bias.
From small signal analysis(if I did it right), there will be pole-zero doublet around gm43/Cgs w/o R1; if R1 added, assume gm43*R1<<1,a second pole around 1/(Cgd*R1) will be introduced besides the pole-zero doublet around gm43/Cgs which will filter out higher frequency noise coupled from the input.

Hope it helps you!

regards,
jordan76

 

[venkateshr]

Jordan I didnt get u properly . With the resistances the first zero(in transfer fn ) would be preponed .

since TF(with resistance) = A/(1+RCs) without that it would be approx. A .

one effect without the Resistors is the effect on ip due to changing op due to the Cgd cap. am i right ? kindly answer my doubt.

 

[sabu]

i tnk r0,r1,r4 and r5 are for providing the dc biasing for the inputs to the ampr.While i tnk that r2 and r3 are for improving the closed loop stability of the diff ampr. please check it out.
regards
sabu