[sdiy] Four bit to serial translator
Scott Gravenhorst
music.maker at gte.net
Sun Nov 16 20:42:08 CET 2003
You want to know if any bit or bits of a 4 bit word changes. Well, A
differentiator will create a pulse, but for a 1 to 0 transistion, the pulse
will be negative and for a 0 to 1 transition, the pulse is positive. So the
differentiator for each bit should have two diodes, one for the positive
pulse and one for the negative.
Ok, here is my possibly daft idea:
to other bits
|
From |
Other ->|-+--R-+
Bit | |
R |
| |
gnd |
| |\ +-->|--Vdd (CMOS)
+-->|--+---+--R----|+\ |
| | | >-+----+----o Pulse Out
| R +--|-/ |
| | | |/ |
| gnd | |
| +---+--R----+
| |
| |
| |
BIT ----||--+--|<--+-R--+
| |
R |
| |
gnd |
|
|
From -|<-+-R---+
Other | |
Bit R |
| |
gnd to other bits
Probably good to use a CMOS buffer to drive this.
The R represents a resistor, they are not necessarily
all the same value. I think this would use 2 ICs at most,
No transistors, 4 caps for the differentiators and 8
diodes. Fewer resistors than 52... looks like 4 per
bit plus 2 more for the opamp gain control stuff.
Something like 18 to do this for a 4 bit word. The caps
and resistors to ground after the diodes would control
the width of the pulse.
Peter Grenader <peter at buzzclick-music.com> wrote:
>Guys and gal,
>
>I'm in a bit of a pickle and it's getting sour here - thought you cold throw
>me a life raft.
>
>I've added an asynchronous VC to Milton. Quite the bomb. It doesn't give
>two sheets what the incoming pulse is doing, or even if there IS an incoming
>pulse for that matter - when the VC input's A to D changes state, the
>counter immediately reacts. With this you can now set the start and end
>step to wherever you want, plus achieve some amazing non-linear
>progressions. Driving it with a sine wave being particularly interesting as
>the sequence accelerates at both ends.
>
>Problem is, as the name implies - it's asynchronous. It functionally
>remains a bit of a lame duck as there no way of syncing anything to it. So,
>unless you're planning on a legato glissandi, you're screwed.
>
>I've worked out a way to extract a timing pulse, but its involving more
>parts than I like, and I'm certain there's an easier way.
>
>What I need is a short pulse each time the 4 bit word changes. I tried just
>monitoring the LSB, that works until it jumps to another of the same parity
>and then it's lost. I tired doubling the speed of the LSB to synthesize it
>and that was totally the wrong direction. I've thought through parity
>comparators, but again, that leaves holes.
>
>So what I've ended up doing is taking the outputs of the 4050 (all sixteen
>of them) and shaping those into pulses rather than gates and summing them
>into one continuos stream, each with a 50ms on time.
>
>Although it took only one sentence to describe it, that function requires
>sixteen transistors, 52 resistors, 22caps and 5 ICs. In short - forget it.
>
>Any competent PIC programmers out there who are up for a challenge?
>
>If nay - anyone have a CMOS logic level solution???
>
>thanks in advance,
>
>Peter
>
=========================================================
- Where merit is not rewarded, excellence fades.
- Hydrogen is pointless without solar.
- What good are laws that only lawyers understand?
- The media's credibility should always be questioned.
- Governments do nothing well, save collect taxes.
-- Scott Gravenhorst | LegoManiac / Lego Trains / RIS 1.5
-- Linux Rex | RedWebMail by RedStarWare
-- FatMan: home1.gte.net/res0658s/FatMan/
-- NonFatMan: home1.gte.net/res0658s/electronics/
-- Master of Arcane and Hidden Knowledge.
More information about the Synth-diy
mailing list