MIKE ROSS wrote in a message to WAYNE CHIRNSIDE:

 MR> "WAYNE CHIRNSIDE" bravely wrote to "ROY J.
TELLASON" (10 Jan 04 
 MR> 11:40:00)  --- on the heady topic of "Re: bag of chips"

 WC> Built a Hi-Fi audio link that used a 4046 PLL at either end. Got 
 WC> well over 1,000 feet range with just a LED and a 3" glass lens at 
 WC> one end and far smaller lens at transmitter end.

 MR> When I was in school learning about pulse techniques, I 
 MR> experimented on phase modulation using pin 5 or control voltage 
 MR> input on a lowly 555 timer. It was amazingly clear sounding 
 MR> knowing it was being chopped up into square bits. That is also 
 MR> when I learned about "aliasing". I kept seeing the output waveform 
 MR> split up on the scope into a low frequency component and couldn't
 MR> understand why it did it until I learned about Nyquist. Fun 
 MR> stuff...

Speaking of chunky waveforms...

I wanted a function generator for a long time,  and although I kept hearing
about there being ways to cover the whole audio range without
bandswitching, the circuits I kept running across still used it.  Worse
yet,  they didn't even switch by decades or anything.  So finally I ran
across a chip which would do it (the number escapes me at the moment).  It
wanted a bipolar supply,  which I gave it in the form of four diodes using
a tapped transformer...

There's an op amp to buffer the output of the thing,  and also allow me to
apply some DC offset if I want it -- you have to pull that knob out to
activate it,  though.  And there are choices of sine, triangle, and square
wave outputs.  And I even added a BNC connector with a TTL-compatible logic
level output,  though I'm not so sure that the 4050 chip I have in there is
the best choice.

Here's the thing,  though -- you get real low in frequency,  and all of a
sudden the output waveform gets *real* distorted.  Sine or triangle waves
have a chunk out of the middle of the top part.  I can't recall offhand
what squarewaes do.  I showed this to one guy,  and he said to me "get
that power supply out of the box",  but it being a transformer and the
rectifier and filters being on-board,  that's a little difficult to do.  I
did add more filter capacitance at one point,  and that didn't help much. 
I think that what I'm likely to try next is removing the zener diodes that
are in there currently and sticking a couple of 3-terminal regulators in
there instead,  since I have them on hand.  Along with associated caps and
such.  Other than that,  I don't know...

You guys have any ideas on this?  

---