# Band-Pass Filter Design - Help

Discussion in 'Homework Help' started by Sgt.Incontro, Feb 7, 2013.

1. ### Sgt.Incontro Thread Starter Member

Dec 5, 2012
50
1
Hi, all:

For my coursework, I have been given the task of designing a band-pass filter. The specification requires that:

1) The lower cut-off frequency should be approximately 200Hz.
2) The higher cut-off frequency should be approximately 4800Hz.
3) The filter design should be comprised of both a low and high pass filter. (Incorporated together.)
4) The design should be kept simple. (No OP-AMP filters etc haha)

Now, the way I can see it, there are only two main ways of doing this:
- Using LR filters
- Using CR filters

IMO, using CR filters for these frequencies will severely attenuate the frequencies within the selected range (as I found out from a computer generated bode plot)

SO I have therefore come up with this as a supposedly "better" alternative:

FINALLY, here come my questions:
1) Can the chosen values (in the above diagram) be improved?
2) Will the small value of the resistors in the above design be viewed negatively - as they may only be able to handle SMALL input voltages?

Any other comments, guidance appreciated.

Thank you in advance.

2. ### #12 Expert

Nov 30, 2010
16,337
6,821
All filters can be scaled.
All passive filters have voltage loss.
Did anyone tell you how much loss is acceptable? the impedance of the load? the impedance of the source? if it has to be "practical"?

3. ### tubeguy Well-Known Member

Nov 3, 2012
1,157
197
When building a bandpass filter using a low pass and a high pass you should put the high pass filter first.
Yes, it might be counter-intuitive

Edit: Don't forget about capacitors...

4. ### Sgt.Incontro Thread Starter Member

Dec 5, 2012
50
1
Hi, thanks for the reply.

1) What do you mean by scaled?
2) No, we weren't told how much loss is acceptable. (However, as I said earlier, creating a bode plot for the RC version showed that the loss was incredible - the peak of the curve started around -30dB. If anyone wishes, I can upload the schematics for this too.)
3) Impedance of the source is unknown. ( I think this can be somewhat ignored in this case.)

5. ### Sgt.Incontro Thread Starter Member

Dec 5, 2012
50
1
Why may I ask? Just interested to know.

There won't be any capacitors as IMO only a simple LR design is needed, unless I am wrong.

6. ### tubeguy Well-Known Member

Nov 3, 2012
1,157
197
Using a CR high pass first is one way to include L, C, and R's in your design.

7. ### #12 Expert

Nov 30, 2010
16,337
6,821
Scaled: V x (Xl/R) is the same as long as the relationship betweem Xl and R stays the same. R can be 1 ohm, 10 ohms, or 100 ohms and you'll get the same frequency response as long a Xl increases proportionately (and, in the real world, the current isn't high enough to load the source down significantly).

8. ### Sgt.Incontro Thread Starter Member

Dec 5, 2012
50
1
Quick question: for an RL low-pass filter is the phase (PHI) given by arctan(Zcomplex/Zreal) giving arctan(WL/R)?

(Wikipedia seems to say -WL/R, although I am not sure where they got the - sign from.)