[sdiy] Wanted: non mathematical description of the functionofRC-filters

[Richard Wentk]

Coming to this late:

Thinking about this, I wouldn't use voltage divider man - I'd use a spring or reservoir model. The time domain differences *matter*, and it's going to be much easier to describe phase effects later if you start with a tactile physical-ish model of what C & L actually do. 

Once you have that, you can move on to explaining superposition and Fourier, and explaining *why* the waveshape changes how it does. Although in fact that part is kind of optional, because you can get an intuitive understanding of waveshape changes just from the time domain. 

If the model is too simple - and I think vdm is - you're not really explaining the AC case properly, or modelling what happens to step events in the time domain. This is bad, and will probably cause more confusion than enlightenment later. 

I'm surprised you didn't do RC filters until just before the final exam. On my course we did Bode plots of frequency and phase in the first term - the good thing about Bode plots being they're handy rules of thumb, and you don't even need to understand all that s=jw stuff to get useful answers with them.

As for maths - I'm certainly not a fan of gratuitous maths, because I'm not all that good at it, and also because a lot of DSP models of audio circuits, which are based on analog design maths, seem to be drastic oversimplifications of what happens in real hardware. 

But I've seen it suggested that the difference between engineers and technicians is that engineers really understand the math, while technicians copy stuff out of books - and while that's unflattering, it may well be true. 

Richard

On 7 Aug 2013, at 10:09, Thomas Strathmann wrote:

> On 06.08.13 21:14, David G Dixon wrote:
>> I agree with you, to a degree.  The math needs to come after the concept.
>> However, in the case of an RC filter, the concept is of a voltage divider
>> where one of the "resistors" has a frequency-dependent resistance.  None of
>> this is going to make sense unless a current balance is drawn around the
>> central node of the divider.  This, quite simply, requires some math!
>> However, it is not much more than arithmetic.
> 
> Agreed. In the electronics lecture I was subjected to at university the
> RC filter was something we did as an example in preparation for the
> final exam. It's quite basic if you know just a little bit about
> electronics.
> 
>> If someone wants to understand RC filters, but isn't even prepared to do the
>> simple math around a voltage divider, then I would say that they simply
>> aren't invited to the club.  Sorry, call it elitism if you will, but a
>> modicum of effort must be made with these things.
> 
> The good thing about almost all matters of learning is that it's not a
> club. If it were I wouldn't be interested in learning things and
> improving my understanding. But I'm also a firm believer that it's not
> the math that is hard but the way it is taught sometimes. I'm not a
> believer in dumbing things down to a cookbook approach (although that is
> in my experience exactly what engineers do), I just don't think that
> everybody needs to have the same level of understanding. Sometimes a
> glimpse at the truth is just enough, just like you to KVL and KCL
> instead of Maxwell for example. For me the voltage divider "model" is
> good enough to work with, for someone else it might be an analogy to a
> suitable mechanical system. Sorry, if I come across as a little
> self-centered and arrogant in what and how I write. Higher education is
> a topic I care deeply about -- it's nearly impossible for me to resist.
> 
> 	Thomas
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy