The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> Frequency respond in the filter
https://designers-guide.org/forum/YaBB.pl?num=1192821863

Message started by James Bond on Oct 19^th, 2007, 12:24pm

Title: Frequency respond in the filter
Post by James Bond on Oct 19^th, 2007, 12:24pm

Hi,

I met a strange problem recently. I built a 5 order chebyshev low pass filter with a 2db ripple in the passband. The opamp used in the filter is a fully differential op with open loop gain 58dB, 3dB bandwidth 220kHz. At 1.2MHz, it still has 45dB gain.

I test the filter with 1Vpp input swing, supply is 1.2V and sweep the input frequency from 10k to 1MHz, the cutoff frequency is 1.2MHz. The filter works fine with low frequency from 0-100KHz, however, the amplitude of the output signal is attenuated when the frequency increases from 200KHz and the attenuation is more than 3dB for some frequency. The transient simulation shows that at frequency around 300KHz, the signal is distorted very much even higher than higher frequency. I could n't find the problem, does anybody know what might be the reason for this attenuation? Any suggestion would be appreciated. Thank you very much

Best Regards
james

Title: Re: Frequency respond in the filter
Post by tosei on Oct 19^th, 2007, 4:59pm

Hi,

Did you check for slew rate conditions? you mentioned some distortion showed up which might suggest some slew rate problems are present.

Cheers
tosei

Title: Re: Frequency respond in the filter
Post by jimwest on Oct 19^th, 2007, 6:43pm

To James,
According to your large input swing, the distortion and the extra attenuation may be introduced by the nonlinearity of the mos switch, if adopted.
Another reason might be the deviation of f-3dB which caused by the change of the effect resistance or capacitance.

K/R,

Jim

Title: Re: Frequency respond in the filter
Post by James Bond on Oct 20^th, 2007, 2:45am

Hi,

I tried to sweep the small input signal with amplitude 50mVpk @600KHz, there is very few attenuation at the output this time. I think it might be that the THD is too high at high input swing (500mV @600khz, thd =1.4%), all my resistors and capacitors are ideal , so the values would be very accurate. Do you think it's the thd problem ?

james

Title: Re: Frequency respond in the filter
Post by buddypoor on Oct 23^rd, 2007, 5:48am

Hello James B.

Which topology was chosen for your filter ?
Did you use a suitable opamp macro to simulate the response ?
What is the opamp slew rate ?

Title: Re: Frequency respond in the filter
Post by sheldon on Oct 23^rd, 2007, 6:24am

James,

Could also provide more information about the filter structure. It sounds
like it is an active-RC filter which should have a large swing range. Is it
a gm-C filter architecture? Also how did you order the stages, that is, did
you consider the Q of the stage when placing it in the design?

Best Regards,

Sheldon

Title: Re: Frequency respond in the filter
Post by James Bond on Oct 23^rd, 2007, 11:55am

Hi,

I have solved the problem, the signal is clipped in some of the stage that's why the thd is very bad.

jimmy