I have a project (for a math class strangely enough...) to simulate an RLC bandpass filter (series RLC circuit, if that's important) using LTSpice IV that will pass frequencies between 10 Hz and 1.0KHz with a max 3dB damping, where at least -15 dB is reached at 10kHz. It's a pretty simple project, I think, but it's been a while since I've worked with RLC circuits and I'm finding that I'm getting a little more confused than I thought I would.

Before I began calculating I worked out the transfer function of this circuit to be (R/L)*s/(s^2+(R/L)s+1/(LC)) (using laplace transforms--I haven't inversed it yet. It's the transfer function that I am analyzing in spice). From w=1/LC and bandwidth=R/L I calculated that with a 1k resistor I'd need a .642uF capacitor and a .161 H inductor.

When I simulated in spice, I had to adjust these values, as expected, but I I ended up adjusting the capacitor value by quite a lot. I figured that I should leave the inductor value alone since every time I adjusted that I'd be changing both the natural frequency and the bandwidth, so I instead got the bandwidth to where I wanted it to be and adjusted the capacitor. In the end I have values for R=1.01k, L=.161 H and C=15.60 uF.

My basic question is whether I am going about this in a very inefficient way and what process might an engineer use to solve this. The natural frequency of the circuit is now quite different and I'm wondering if in practical situations engineers will sacrifice the natural frequency for the correct bandwidth values or if perhaps there's a way of conserving both that I'm not thinking about.

My questions stem from the fact that while the project is developed for the electrical engineers in the class, I have not myself studied any sort of electrical engineering and so am unfamiliar with the correct process and any additional factors I should be trying to think about.

I do not know how to post pictures but if someone tells me how I can show what my graphs look like. Another issue I'm having is that the graph of the phase angle seems to have a slight hump in the middle--not at all as smooth as it looked like with the bandwidth was narrower.

Sorry if this is a long winded and confusing question--but any general advice on this would be very much appreciated.

cheers,

finley