[sdiy] strange sequencer behavior

[Peter Grenader]

Welcome to my nightmare.

When I was working on Milton - for the longest time it would count fine at
higher speeds, but at lower speeds it would go 1,3,5,7,9,11,13,15 and skip
all the even steps.  It was knicking the crap out of me, because the first
sequncer i made, whose engine was of the exact same componentry, never
exhibited any of these attributes.

More than likely, the problem is noise related.  Some suggestions:

1) .01's a plenty. 

2) If it's breadboarded, check where the wires coming from and going to the
main counter directly are routed.  In the future, make these leads as short
as possible.  But if that's impossible now,  Make sure they aren't laying
over anything that may radiating to the counter.

3) Look at the isolated wires on a scope.  If yours doesn't go into the high
megahertz range, make the horizontal sweep stationary and turn up the
brightness.  You will probably see dots appearing, indicating a change of
logic state, when you don't want it.  This is either noise - or it's very
possible you've got blown CMOS.

4)  How is your clock and reset signals conditioned?  Are you intentionally
controlling their on-time?  If not, you should.  Even though most counters
respond to the high-rising side of the incoming clock, having them stay that
way any longer than required could be sending your counter misinformation
that it's responding to.  I put all of my incoming clock signals going to
the 4516 (all accept hold) through an r/c at the base of a tranny which was
then routed through a CMOS inverter before going into the counter.  I was
having a world of problems before i did this.

hope this helps.

Peter




Sven Windisch wrote:

> hi.
> 
> i got some strange behavior from my sequencer last days. i added a feature
> shown in [0]. the 4024 counter and the 74hc85 comparator. this way i'm
> able to encode the length of the sequence as a binary word. i.e. if i set
> the switches to 0100 the sequence resets after the 4th stage. this works
> well for slow speed. but if i tune the speed up it does the following: it
> counts 1, 2, 3, 4, as it should do, then leaves out 5 and 6 (and somtimes
> 7) and ounts on from 7 (or 8) through 16. this occurs only at a higher
> speed, if i slow it down again it works as it should do.
> 
> any hints?
> 
> greetings, sven.