vc wave select

[Gene Zumchak]

May I add my own two cents (or whatever the relevant currency is).  I agree
with Tom G. that this is an interesting design approach and worthy of
consideration.  I agree with Jorgen that the resulting design is a little
"wordy" and needs some work.  For analog switches, I much prefer the
74HCT4051/2/3 series.  The logic runs between 0 and the + voltage, and the
signals can range between the + ad - voltage, typically +/- 5 to 7.5 volts.
Since the control voltages draw zero current the +/- supplies can just be
resistive dividers.

    The use of 74X30's for decoding purposes is a little unfortunate.  For
ultimate versatility, some kind of PLD is called for.  Twenty-pin EPLD's
(GAL16V8) are available for less than $2 US and can have up to eight outputs
and up to 16 inputs.  The 20V10's are still grossly overpriced at about $5.
Both require programmers.  What has caught my eye is the new Xilinx CPLD XC9530
family.  The smallest, the XC9536 has 36 macro cells and comes in a 44-pin PLCC
package.  It costs less than $3, requires no $1000 programmer ( it uses ISP (In
Circuit Programming).  Xilinx points you to a development kit for $100 with all
the compiler software you need plus a demo board with switches and LEDs and a
cable that connects to the printer port for programming.  I think this is a
super bargain.  The slowest parts are 15 ns.  For about $2 more you could get
an XC9572 with 72 macrocells.  Digikey sells the parts.  A distributor called
Insight has the $99 kit.  Check it out.

Gene Z


tomg wrote:

> > >There were low-pass filters before the Moog made his ladder. I
> > >guess we're all thankful he built his 16 transistor monster I'm
> > >sure there were others that were easier to build maybe even "easier to
> make a >board layout for". duh!
> >
> > Thats not a proper analogy! Digital control stuff doesn't get more
> interesting just because you do it with six ICs instead of one. It won't
> sound better, because it doesn't make sounds. It just gets more tedious.
> Why do you even bother with this? It's just misleading people into wasting
> their time. I do agree that the waveform selector version could have some
> merit, but the ADC-version is really a joke.
> > Go ahead designing a board layout for it. Then you might realize that
> even the waveform selector probably isn't worth the trouble. Been there --
> done that. Combining different waveforms from the same oscillator sounds
> rather dull.
> >
> > /Jorgen
> >
> > P.S. The Moog ladder is actually quite nice to do a board layout for. I'm
> >currently working on just that.
>
> Sure it is. I admit I'm not Dr. Moog. But I think I built a better flash
> converter (fc). I tried to use the paia fc but it has it's problems. Like
> it's linear and runs on +15V.
>
> There is another version floating around by B.P. Vandenburg that uses (7)
> ca3130s a bunch of precision resistors and a log scale that works well
> and runs off +5V so it can speak directly to the digital world.
>
> I'm not misleading people into wasting their time. This is a good design
> that deserves consideration. There are only 5 resistors and they are not
> critical. I admit the pc board won't be a lot fun, but not really that much
> of a problem.
>
> The cost is way low 3915's are under $3 in places 7430s are around
> $0.25 That makes the whole thing cost about $5. CA3130s are $1.30 each
> thats over $9. and you still need all those precision resistors pushing it
> to $10-12 or twice the price.
>
> I take it from your comments and criticism that you have nothing in your
> setup that has a +5 supply, has jumpers, or doublesided boards as you
> consider these to be "bad designs". Certainly nothing that mixes waves
> from the same oscillator like the P5, JP, MS...ect as that is useless.
>
> -tg