The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> three questions need your help
https://designers-guide.org/forum/YaBB.pl?num=1285676864

Message started by lhlbluesky_lhl on Sep 28^th, 2010, 5:27am

Title: three questions need your help
Post by lhlbluesky_lhl on Sep 28^th, 2010, 5:27am

1, for ac noise simulation of opamp (single-ended and fully differential), which frequency bandwidth should be used? i usually use frequency from 0Hz to GBW Hz to calculate the integration noise and then average it, and the noise unit is V^2/Hz or V/sqrt(Hz), is that OK? or any other advice?
besides, how to simulate the transcient noise for opamp? for example: sc circuit.

2, in pcb board, for a wire of 2cm long, what is its equivalent capacitance and resistance? someone tells me, the resistance can be very small, less than 0.1 ohm, because the wire can be thick enough, that is, the cross area can be large enough, then the resistance can be very small. however, for a thick wire in pcb board, the capacitance may be large, so, i want to know, what is the capacitance of a 2cm long wire, if its resistance is less than 1 ohm? can it be 1pF~3pF? or larger? or smaller? please give me some advice.

3, what is the pad capacitance in 0.18 um process in general? someone says, it is about 4pF, but for our chip, we tested the pin capacitance,and it is in the range of 9pF ~ 12pF, it is so big, is it usual? the chip package type is PGA.
besides, what is the relationship between pad capacitance and package type?
which package type has a smaller pad capacitance? and what is the pad capacitance for LCC package type in 0.18 um process? can it be smaller than the PGA package type? please help me, and explain to me more about this.

thanks all for reply, thanks all.

Title: Re: three questions need your help
Post by sheldon on Sep 28^th, 2010, 6:52am

With regard to item #1,

1) You probably should use a wider bandwidth. Assuming a one pole
roll-off, then the equivalent noise bandwidth is (pi/2)*gbw
so noise at higher frequencies, even at 10*gbw can contribute
to the total noise of the amplifier

Not sure why you average the noise. Seems like you should
provide the total integrated noise and noise spectral density
Averaging can be deceptive if there is a significant amount of
1/f noise.

a switched-capacitor circuit is not an amplifier so not sure what
the question is

--> Why simulate the transient noise of an op amp if can can
simulate it with the small signal noise?

--> For Switched Capacitor circuit, use periodic noise analysis

--> If the op-amp is a dynamic circuit, for example, switched
capacitor common feedback, then use pnoise

Best Regards,

Sheldon