The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> a basic question about SC filter
https://designers-guide.org/forum/YaBB.pl?num=1246555498

Message started by subgold on Jul 2^nd, 2009, 10:24am

Title: a basic question about SC filter
Post by subgold on Jul 2^nd, 2009, 10:24am

i have a simple switched cap filter as in the attachment. since the information is only captured at the end of phase 1, do i need to take care about if the opamp settles at the end of phase 2?

Title: Re: a basic question about SC filter
Post by vivkr on Jul 2^nd, 2009, 11:48pm

As you have a capacitance which is permanently connected in the feedback loop (you seem to be realizing a lossy integrator with this), it does matter that the opamp settles well in both phases. By the way, is this diagram complete, or are there some more switches which you have omitted for sake of simplicity?

Regards,

Vivek

Title: Re: a basic question about SC filter
Post by raja.cedt on Jul 3^rd, 2009, 12:46am

hi,
whats this ckt actually, because its like variable gain amplifier not filter....

Thanks,
Rajasekhar.

Title: Re: a basic question about SC filter
Post by buddypoor on Jul 3^rd, 2009, 1:05am

As already mentioned by VIVEK, I don´t know what the task of the shown circuit could be. A filter ? An integrator ? I don´t think so.

Title: Re: a basic question about SC filter
Post by manodipan on Jul 3^rd, 2009, 2:43am

Hi,
I think it's a high pass filter with cutoff defined by feedback components and gain at infinity defined by unswitched capacitors.now it's important that opamp settles at the end of charge transfer phase otherwise output is erroneaos and will lead to several problems like nonlinearity related distortion etc.you can start with a behavioral model and check filter performance with nonidealities in opamp like finite gain and UGB.

Title: Re: a basic question about SC filter
Post by thechopper on Jul 5^th, 2009, 4:56pm

manodipan wrote on Jul 3^rd, 2009, 2:43am:

Hi,
I think it's a high pass filter with cutoff defined by feedback components and gain at infinity defined by unswitched capacitors.

In order for this ckt to be a HPF, should not the input capacitance have a switch in series with the opamp input clocked with CK1?

Tosei

Title: Re: a basic question about SC filter
Post by buddypoor on Jul 5^th, 2009, 10:33pm

Hello SUBGOLD !

Some days ago you have asked a question concerning the function of a specific circuitry - and five community members made up their mind to find the task of the circuit.
Don´t you think now it´s your turn to tell us something about the purpose or the origin of the circuit ?
Or are you not interested anymore ?
Regards

Title: Re: a basic question about SC filter
Post by subgold on Jul 6^th, 2009, 5:48am

buddypoor wrote on Jul 5^th, 2009, 10:33pm:

hi,
sorry for the late reply because i didn't get online during the weekend.

as already pointed out by manodipan, i don't see why this circuit cannot be used as a HPF.

if the input cap is also a switched cap as Tosei suggested, then it seems to be a LPF.

Title: Re: a basic question about SC filter
Post by thechopper on Jul 7^th, 2009, 2:32pm

Hi subgold,

Actually I was not suggesting the input cap was a switched cap. I was only suggesting a switch in series with it and the input of the opamp. that would provide sampling on your input cap and still perform a HPF action. Since there would be no reset during ck2 on such cap then it would not act as a resistor in average (in which case then you would get a LPF).

I agree that not switching the input cap performs HPF action, but without a sampled input (as opposed to what I was suggesting). Therefore you have a continuous-time output and I do not see why the opamp needs special settling requirements during ck2 (reset phase).

Regards
Tosei

Title: Re: a basic question about SC filter
Post by buddypoor on Jul 8^th, 2009, 12:29am

Besides the question of transfer characteristics, I am afraid the circuit as shown will exhibit severe bias point problems due to the 2 non-switched capacitors.
Regards

Title: Re: a basic question about SC filter
Post by subgold on Jul 9^th, 2009, 2:45am

HdrChopper wrote on Jul 7^th, 2009, 2:32pm:

Hi Tosei,

that's exactly my question.

do you mean for the shown schematic (no sampling on the input cap), it is not necessary to settle in ck2? other replies of this topic suggested it should, which is also what i thought at first. because although the information is only sampled in the end of ck1, the capacitor should also get the correct charge in ck2, in order to further process the information in ck1.

Title: Re: a basic question about SC filter
Post by subgold on Jul 9^th, 2009, 2:59am

buddypoor wrote on Jul 8^th, 2009, 12:29am:

Besides the question of transfer characteristics, I am afraid the circuit as shown will exhibit severe bias point problems due to the 2 non-switched capacitors.
Regards

hi,

could you please explain your doubt about the transfer function? i think it is a very basic building block for active filters. of course it is a little simplified in order to illustrated the question. in many other cases (e.g. biquad or higher oder filters), there might be feedback caps coming from the following stage, but i thought the conceptual background concerning the settling requirement should be the same.

as for the bias problem, i think we always need some strategies to guarantee the correct biasing for a pure capacitive feedback network, there is nothing special about this circuit compared with other switched cap circuits.

Title: Re: a basic question about SC filter
Post by thechopper on Jul 9^th, 2009, 2:32pm

Hi subgold,

What I meant is that: since you are supposed to fulfil the settling requirements for clock phase 1, then settling requirements for clock phase 2 will be already covered since there is no sampling nor charge redistribution during that phase. Therefore the requirements are not as stringent as during clock phase 1.

Hope this clarify my viewpoint.
Tosei