The filter topology here is the Twin T topology. I personally find this topology to be too cricital to be useful in a large scale manufacturing environment. To counteract that, I modify the traditional Twin T discussion to use all equal resistors and all equal capacitors. In this way, you can take advantage of the fact that components manufactured at the same time, by the same manufacturer, using the same process will probably be pretty close in value, and will probably exhibit similar temperature characteristics. There will be a sharp peak - somewhere - although component tolerance may put it off a bit.
I've also added two resistors that allow some control over the Q and gain. They are varied in a 2 to 1 ratio - take Ro and multiply it by 10 to get Rq1, divide it by 20 t get Rq2 for example. Q's down to 2.5 are pretty easy to realize, but below that resistor values affect center frequency. Amplitude tracks as well, although neither effect is completely linear. If you need more control, you need to go back to my modified Deliyannis filter configuration. To get the highest peaks, simply make Rq1 open and Rq2 zero.
The Twin T band pass filter's ultimate stop band rejection is unity gain, so it is technically not a bandpass filter, but a resonator. This may limit its usefulness in your application. Unity gain will extend down to zero frequency and up to the bandwidth of the op amp, which may allow unwanted harmonics to pass.
There is a curious variation on this topology. It is possible to put two or more networks inside the feedback loop. For more on this subject, look at the stagger-tuned filter.