Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

bandpass filter with bandwith 1khz and fr=140khz

Status
Not open for further replies.

hamid.abbaszadeh

Member level 3
Joined
Jan 26, 2012
Messages
63
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Location
iran
Activity points
1,666
how can i made it.as i search about it for this filter we need to an opamp with big gain band .how can i use from digital filter with atmega128a1
 

Yes, that's true. According to the filter Q of 140, you need a large gain margin at the center frequency, e.g. an OP with several 100 MHz GBW. Or use an OP with moderate 20 to 50 MHz GBW and adjust the center frequency. Both is well feasible with recent OPs. A practical problem is however initial frequency accuracy and frequency drift.

I doubt that implementing a digital filter for 140 kHz center frequency (and respective high sampling frequency) is feasible with an AVR provessor.
 

i need to a bandwith near 1khz its not feasible with passive one

- - - Updated - - -

as my friend said i need to a op with 100M gainband. but if i want to have 1khz bandwith its not feasible.this is my problem
 

A series LC acts as a narrow bandpass filter.



The ratio of L:C affects bandwidth. The load affects bandwidth.

You will need to adjust values to suit your spec.
 

A passive filter with Q=140 isn't easy to make, but feasible with good ferrite inductors. With modern OPs, there's no principle problem to design it as active filter, but as said, achieving a stable centre frequency is a practical issue. Well considered component selection and tuning are necessary.

The filter can be made as digital filter, but you didn't yet chose a suitable hardware. The mentioned ATmega isn't it. In case of a digital implementation, you have also to design AD and DA interfaces.

The filter specification might be elaborated. Are you intending a simple bandpass with one pole-pair, or a more complex one?
 

im used to filter pro for design filter.i think q=140 need to op with near 1Ghz bandwith
 

im used to filter pro for design filter.i think q=140 need to op with near 1Ghz bandwith

Hamid - as far as I know FILTER PRO (from TI) contains only two filter topologies: Multi-feedback (MFG) and unity-gain Sallen-Key. Perhaps there is a newer version with more alternatives?
I think, in your case (fo=140kHz, Q=140) these two structure alternatives are not suitable.
According to my experience, you should try other filter topologies (for example GIC-based filters).
Do you have access to "Filter Free" (from Nuhertz, www.filter-solutions.com) or "Filter Wiz Pro" (from Schematica Software, www.schematica.com) ?
Both programs allow much more different filter srtuctures.
 

tank u man but im not hope so topology cant make miracle it need a big gainband
 

I agree with LvW that high frequency active filters ask for other topologies than Sallen-Key or MFB.

For the discussed 140 kHz, standard topologies are still within reach, using high speed general purpose OP with a some 100 MHz GBW. You'll need to adjust gain and frequency according to real OP parameters. If your filter design tool doesn't support real OP parameters, you can e.g. use LTSpice to trim the filter components in a simulation. Fine tuning of center frequency should be provided by a variable resistor.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top