Just breadboarded a simple 40106 oscillator and added an RC/CR. Worked a treat! So, what do I do next?
The funny thing about RC filters is they also depend upon input and output circuits that they attach to. The source feeding into the filter should have a relatively low impedance so as not to affect the chosen filter response (which likely applies to your application).
Another challenge is the input impedance of the stage following the filter. It should feed into a high input impedance for best results. Feeding into a circuit with lower input impedance will affect the response. Exactly how and what ranges to watch out for is something I want to investigate. I hope to run some LTspice simulations on that purely out of curiosity.
I only mention these things because what seems simple doesn't always work as expected. So for best results, low impedance feeds into the RC filter and then the filter feeds into a high impedance following stage.
I heard that 'proper' filters incorporate op-amps (is an LM386 ok?). Is this so the quieter tones are louder? The next step would be great to see!!!!
Many thanks again!!!!
The RC filter is a
passive filter. You can't beat it for simplicity. But it does suffer from what is known as "insertion loss". The signal coming out will be reduced in amplitude.
With an opamp (or transistor) circuit, you make what is known as an "
active filter". It is "active" because it includes active components (that amplify).
So what is best depends upon your needs (as in most things in life).
If your RC filter is going into another amplifying stage, you really don't need to worry about insertion loss (you simply make it up the lost gain in the following stage(s)). If you want your filter to amplify a band of frequencies above the current level, then an active filter will be required.
Here's another deciding factor:
One RC filter drops power at a rate of 3 dB per octave beyond the cutoff frequency. With signals, you are usually only concerned with voltage, which drops at 6 dB / octave. [Apologies for going into decibels here... -3 dB means a drop of 50%]
If you wanted a sharper cutoff than that, you need to run multiple RC filters in series. The problem is that this doesn't work well unless you go to an active filter design. The reason is that the output impedance of a passive RC filter is relatively high. Feeding into the next RC filter means that both filters now become corrupted in filter response because of this (both filters will interact and change the frequency response).
A simple fix is to run one RC filter into a buffer (opamp), which then feeds another RC filter. The buffer provides the high input impedance that the first filter needs, while providing a very low output impedance to feed the next filter with. So now we have gone full circle in this post.
Of course, you can also have opamp circuits that are active filters in themselves, rather than just a buffer, but that gets more difficult to describe in a brief discussion.
Opamp ChoiceThe LM386 is really meant to be an output amplifier, driving a speaker or similar type of load. If the output in your application is simply an audio output signal to be plugged into a stereo (for example), I would not use that.
Use a voltage amplifier opamp like the TL071, TL072, or TL074 for single, dual or quad opamps, depending upon your needs. These will be easier to work with than the LM386 (fewer connections) and provide better quality results.
Sometimes I have seen LM386's used in pedal circuits. This is normally done when it is doing some extreme "driving" into another distortion type of stage. I still suspect that this application of the LM386 is less than optimal, but I've not checked it out in detail.
I've overly simplified the opamp selection here, since there are lots of other things to consider. It requires a book to treat the subject properly. But the TL07x opamp is a great jellybean part for low cost - high performance. If you need lowest noise, lowest distortion, the NE5532 or NE5534 is suitable for most applications.
Update:I added some notes about effects of inter-stage impedance effects on filters here:
http://www.experimentalistsanonymous.com/ve3wwg/doku.php?id=rc_filters_coupling
