Hey all. I am fairly new here and I am hoping you may help me in designing a narrow BPF along with an AMP for the Intermediate Frequency stage of a basic superhet AM radio (IF center freq at 455kHz and bandwidth of 10kHz).

I have done a fair amount of research and found many circuits that contain crystal components. I would rather not use such components for now as I am not familiar with them at all.

I was wondering if I can use a Chebyshev or Cauer filters (do Cauer BPF exist? I have tried googling them many times with no results) for the IF narrow band pass filter?

As for the IF AMP, I have no idea what to do. How can one amplify only a certain band of frequencies? Would any amp with a magnitude response that contains the frequencies required work (regardless of other frequencies)? Any help or references will be greatly appreciated.

Thank you all for reading.

 

Google 455kHz SAW filter.

 

An amplifier by itself will amplify a wide band of frequencies as determined by the active element frequency limits and parasitics in the amplifier. In an amplifier used for IF it is designed to have a frequency response larger than required with the actual bandwidth determined by the IF filters.

Here are some basic IF amplifier references?

 

Thank you for the replies. I googled SAW filters as MrChips suggested but it seems those you have to buy rather than build. I am trying to BUILD my own narrow bpf and my own IF amp. They don't need to be perfect, but they need to work. Crutschow, I have googled "IF Amp" so many times. For some reason when I opened your link I decided to go to images this time, and it seems I did find a couple, so thank you!

I am still wondering about the narrow BPF??? Can I use Chebyshev or Cauer filters?

 

there is an older circuit for narrowing bandwidth, google "Q multiplyer" an add on circuit that adds regeneration to improve selectivity.

 

I am still wondering about the narrow BPF??? Can I use Chebyshev or Cauer filters?

Old radios used tuned transformers. The primary and secondary windings had capacitors in parallel with them. No fancy active filters were involved.

 

..........................

I am still wondering about the narrow BPF??? Can I use Chebyshev or Cauer filters?

Certainly you can use Chebyshev or Cauer filters. How steep a rolloff do you want at the band edges (in dB/octave or dB/decade)?

 

there is an older circuit for narrowing bandwidth, google "Q multiplyer" an add on circuit that adds regeneration to improve selectivity.

Hmmm, never heard of these. Just googled "Q-Multiplier" and found some circuits. These look fairly easy to implement. So I guess I will have to decide between these and either a Chebyshev or Cauer. Quick question, if you don't mind (what are R1 and R2 in this link: http://www.seekic.com/circuit_diagram/Basic_Circuit/Q_multiplier_filter_circuit.html ? Are they two POTs with their 2nd pins connected to each other?)

Old radios used tuned transformers. The primary and secondary windings had capacitors in parallel with them. No fancy active filters were involved.

I actually just recently learned about these, when looking at the IF AMP images that I got from Crutschow, I saw these transformers you speak of.

Certainly you can use Chebyshev or Cauer filters. How steep a rolloff do you want at the band edges (in dB/octave or dB/decade)?

I don't know what a good answer is I don't want the filter complexity to go through the roof but I want it to be an alright filter... Do you have any recommendations. I haven't seen Cauer BPFs, I have only seen Chebyshev BPFs. Or is Chebyshev the way to go? That is probably why you ask me this question...

 

I assume you are working on a solid state design. As stated by an earlier post you need to determine the slope of the bandwidth. Ceramic filters are one way of narrowing the band width. A lot of if transformers for solid state circuits
are adjustable. On way is to tune each of the transformers to a slightly different center frequency. The more tuned circuits you have the narrower the bandwidth can be controlled. The attached pdf file is of IF transformers stocked by mouser electronics. Just a rough calculation of the first one has a bandwidth of about 6KHZ

 

..........................
I don't know what a good answer is I don't want the filter complexity to go through the roof but I want it to be an alright filter... Do you have any recommendations. I haven't seen Cauer BPFs, I have only seen Chebyshev BPFs. Or is Chebyshev the way to go? That is probably why you ask me this question...

Well there's your problem. You can't design a circuit without knowing the requirements. The type of filter isn't the question, it's how good does the filter need to be (the steepness of the rolloff). Once you determine that, then you can proceed with the design.

 

I assume you are working on a solid state design. As stated by an earlier post you need to determine the slope of the bandwidth. Ceramic filters are one way of narrowing the band width. A lot of if transformers for solid state circuits
are adjustable. On way is to tune each of the transformers to a slightly different center frequency. The more tuned circuits you have the narrower the bandwidth can be controlled. The attached pdf file is of IF transformers stocked by mouser electronics. Just a rough calculation of the first one has a bandwidth of about 6KHZ

Thank you, I will check out your link now.

Well there's your problem. You can't design a circuit without knowing the requirements. The type of filter isn't the question, it's how good does the filter need to be (the steepness of the rolloff). Once you determine that, then you can proceed with the design.

I agree with you but the problem with giving you a roll off spec is that I have never designed a narrow band pass filter before (haven't even used one) so I do not have the expertise to give you specs like that. I do not know what a "standard" roll off for a narrow bandpass filter is which is feasible to build with discrete components (and not have a really complex circuit).

I may have to resort to purchasing a narrow BPF instead

 

A typical active filter can generate a 2nd-order response (12dB/octave rolloff) for each op amp. So it's a trade-off between how much roll-off you want at the band edge to suppress the signal from a strong adjacent channel compared to how many opamps and other parts you want to use.