| Author |
Message |
hebu
Joined: 15 Nov 2004
Posts: 195
|
 28 Jun 2005 15:20 sc filter |
|
|
|
|
In IC implementation and audio frequency application. SC or GmC has the
better dynamic range?
Consider the flicker noise, thermal noise, linearity. And, the filters are same
type.
|
|
| Back to top |
|
 |
eda4you
Joined: 17 Sep 2002
Posts: 283
Helped: 17
|
28 Jun 2005 15:41 gm-c filter |
|
|
|
|
| hebu wrote: |
In IC implementation and audio frequency application. SC or GmC has the
better dynamic range?
Consider the flicker noise, thermal noise, linearity. And, the filters are same
type. |
Hallo Hebu!
SC is the much better choice! SC has better linearity, the flicker and low frequ. noise can be cancelled using correlated double sampling or choppering. With SC you can just get DR as high as 120dB, while gmC has mabye the half!
|
|
| Back to top |
|
|
hebu
Joined: 15 Nov 2004
Posts: 195
|
29 Jun 2005 14:38 gm-c filters |
|
|
|
|
Usually, what's the clock frequency in audio application?
How I define the requirement of clock frequency, anti-aliasing filter, and smoothing filter?
Thanks,
|
|
| Back to top |
|
|
mists
Joined: 14 May 2002
Posts: 380
Helped: 33
Location: China
|
30 Jun 2005 4:46 noise in sc filter |
|
|
|
|
| the audio frequency is not larger than 192k( in general it is 48k), so the clock frequency just rely on the oversampling ratio in you system. 12.288M is usually used frequency.
|
|
| Back to top |
|
|
hebu
Joined: 15 Nov 2004
Posts: 195
|
30 Jun 2005 5:20 sc filter |
|
|
|
|
If we want to make a Req such as 1M ohms high, Req=T/C.
And the clock freq is as high as 10MHz, the C should be 0.1pF.
Is this realiable size in 0.35um IC implementation?
Or, anything I missed?
|
|
| Back to top |
|
|
eda4you
Joined: 17 Sep 2002
Posts: 283
Helped: 17
|
01 Jul 2005 9:38 gm-c filter latest application |
|
|
|
|
| hebu wrote: |
If we want to make a Req such as 1M ohms high, Req=T/C.
And the clock freq is as high as 10MHz, the C should be 0.1pF.
Is this realiable size in 0.35um IC implementation?
Or, anything I missed? |
Yes, 100fF is no problem. BUT the noise (= k*Temp/C) may be to large.
I would say caps like 1pF is more suitable.
|
|
| Back to top |
|
|
sgperzoid
Joined: 29 Mar 2005
Posts: 58
Helped: 5
|
01 Jul 2005 10:36 filter sc gmc |
|
|
|
|
| hebu wrote: |
If we want to make a Req such as 1M ohms high, Req=T/C.
And the clock freq is as high as 10MHz, the C should be 0.1pF.
Is this realiable size in 0.35um IC implementation?
Or, anything I missed? |
100fF is already the minimum value of mimcap for 0.18um CMOS
I think for 0.35, maybe it is higher, you may check it easily
for smaller value, you can put capacitor in serial or using MOS dummy cap
but the accuracy degraded, I think this is fatal for analog filter...
|
|
| Back to top |
|
|
Google
AdSense
|
01 Jul 2005 10:36 Ads |
|
|
|
|
|
|
| Back to top |
|
|
hebu
Joined: 15 Nov 2004
Posts: 195
|
04 Jul 2005 15:28 what is gmc filter |
|
|
|
|
Two questions,
1. how about the 100fF cap to tolerate the charge injection?
2. If a switch cap ckt needs to be made a precise matching between 0.5fF and
50pF, is it possible?
|
|
| Back to top |
|
|
