wiml@blake.acs.washington.edu (William Lewis) (10/05/89)
Hello, I need some help from those people out there who understand
analog circuitry (in depth, I mean...) I'm trying to build a simple
circuit to divide the frequency of in incoming (audio) signal down
an octave or several. What I was thinking of doing was to digitize the
signal with a comparator of some sort, feed that through a flip-flop
or two, and then put capacitors across and in series with the output
to re-"analog"-ize it. (ie, chop off the aliasing frequencies with a
capacitor across the outputs, and remove any dc using a capacitor
in series. Plus a volume control and such of course.)
Now the question is... Would this work? Would it work reasonably well?
The signal it's to be dividing is a normal audio signal (speech, music,
etc.) which I'd expect to have a pretty complex waveform. Would this
1-bit digitization introduce huge amounts of distortion? (A little is
OK, but I want to be able to understand what's coming out). Is there a
better way to do this? (if so, what?) I feel more at home with digital
stuff (and low speed. All I want to see are pullup resistors and
bypass capacitors on the power supply =8) ) so if you reply,
"Oh, that's easy, just use a Boyglesthibben circuit" explain to me
what a Boyglesthibben circuit is...
Anyway, thanks in advance for any help / suggetions... if you feel
this is too unusual a request to be broadcast, email and I'll
summarize...
--- phelliax
"To Forbid is to Suggest" (... a guy I know, and others too)
--
wiml@blake.acs.washington.edu (206)526-5885 Seattle, Washingtonmyers@hpfcdj.HP.COM (Bob Myers) (10/07/89)
> Hello, I need some help from those people out there who understand >analog circuitry (in depth, I mean...) I'm trying to build a simple >circuit to divide the frequency of in incoming (audio) signal down >an octave or several. What I was thinking of doing was to digitize the >signal with a comparator of some sort, feed that through a flip-flop >or two, and then put capacitors across and in series with the output >to re-"analog"-ize it. (ie, chop off the aliasing frequencies with a >capacitor across the outputs, and remove any dc using a capacitor Well, one-bit sampling at a sufficiently high rate is certainly possible, but I don't think that it will either (a) be as simple as you'd like or (b) provide the desired results with acceptable signal quality. You have asked to be able to "divide" the frequencies of an audio signal down "by an octave or so"; this, though, sounds like the same thing you'd get (and very easily, I might add) by simply recording the signal on tape and then playing it back at half speed. Presto, everything drops to 1/2 the original frequency, at the expense of taking twice as long to play! (An example of the "There ain't no such things as a free lunch!", or TANSTAAFL, Theorem!) Doing this sort of thing "on the fly" (real time) could be done digitally, by sampling the signal at the appropriate rate and number of bits, and then dumping the results into a memory or onto a tape for later playback (which is really the same as the analog recording) or performing some mathematical operations to translate the signals down to the desired frequencies - not an especially simple task. (You're basically asking for the complement of the operation performed by those nifty little high-tech cassette recorders, which can speed up voice recordings without changing the pitch. You want to keep the original SPEED, but LOWER the pitch. I think.) Some further information, please - why do you want to do this? What effect are yoo really trying to acheive? How concerned are you over the quality (fidelity) of the final result? What other limitations (power, cost, size of the unit, etc.) are you faced with? What frequencies need to be handled? Bob Myers KC0EW HP Graphics Tech. Div.| Opinions expressed here are not Ft. Collins, Colorado | those of my employer or any other myers%hpfcla@hplabs.hp.com | sentient life-form on this planet.