Need idea on the circuit design of ac amplifier but in dc it is unity gain

[analogmind]

Hi I would like to get circuit ideas for the following:

Dc gain = 1 (unity gain)
Ac gain = 100

That is if I give 1.8v DC + 1mV sine signal , the output should be 1.8v DC + 100mV sine signal. The output response should be fast ( fast settlinfg time)

 

[kak111]

You need two parallel "subcircuits" one for DC-part , one for AC-amplification.
What is frequency range for this circuit and settling time value.

 

[analogmind]

the input signal is in the audio frequency range. Settling time is max 1millisec. How can I have two subcircuits? AC or small signal 100mv should ride on dc 1.8v.

Thanks.

 

[LvW]

You need two parallel "subcircuits" one for DC-part , one for AC-amplification.

I don't think so. It's much easier.
You can use a standard opamp in a non-inverting configuration with a lowpass in the feedback path .
A resistor R1 between the output and the negative input provides 100% dc feedback (for dc unity gain) and an R2-C2 series combination between the inv. terminal and ground provides a frequency dependent feedback.
With R1=99*R2 you have a an ac gain of 100 (and a dc gain of unity). The capacitor must be chosen as large as necessary to provide a lowpass corner that is sufficiently small (dependinmg on the ac frequency)

 

[analogmind]

How will it be unity gain in DC? In DC capacitor is open and then dc gain is (1+ R1/R2) ? Am I wrong?

---------- Post added at 15:18 ---------- Previous post was at 15:07 ----------

Sorry i missed the point that R2 and C2 are in series. looks like the capacitor value is large which means large area on chip. Any ideas to reduce this ?

 

[LvW]

looks like the capacitor value is large which means large area on chip. Any ideas to reduce this?

Depends on your requirements (corner frequency).
For an ac gain of 100 the corner is mainly determined by the series combination R2-C2: wo=1/R2C2.
Therefore, R2 should be selected as large as possible - consistent with R1=99*R2.

 

[FvM]

looks like the capacitor value is large which means large area on chip. Any ideas to reduce this ?

Noone would expect a IC design problem up to this point. Of course it doesn't change the basic circuit theory, but I doubt that it's reasonable to integrate audio frequency range coupling or filter capacitors on chip. You also won't use any silicone resistors for amplifier gain definition, because their linearity and drift behaviour isn't acceptable. Only thin film would be an option.