4094 based shift register nightmare...

[Snaper]

Hi there guys,

I wasn't really sure where to post this, but this one drives me crazy.
It is a 4094 based shift register / random rhythm generator.
(Schematic attached.)

Normally, it should work like this:

- clock in
- turn the pot of the comparator
- push the stream/lock switch
- it enables random bits to go through it and fills the shift register
- if you release the switch (it is spring loaded) it feeds back the last bit to the data input, essentially loops an 8 step rhythm
- outputs are fed to AND gates to form pulses

It pretty much does this. The pot is more towards to CW the more random bits go through, the more complex the rhythms are.

The problem is that even if the switch is closed, if the pot is close to CW, random bits go through…it is like the pot has affect on the data input, which is kinda weird.
I’m looking at the board layout, just to see if there is a possibility of “crosstalk” between pins.

I've cut the above mentioned, re-routed with a flying wire, no luck.
I even cut the clock trace, just to confirm that there no crosstalk - still no luck.

The problem must be somewhere around the 4081, since I've checked the 4094s output with a scope and it seems stable - no random bits there, so in theory, the 4081 got those from somewhere, but still can't determine where.

Any tips?

---

Screenshot 2023-09-07 at 19.47.43.png
Description:
Filesize:	54.47 KB
Viewed:	125 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[blue hell]

The schematic misses some things ..

Does the 4081 have power and ground? also, does the 4091 have ground?

[Snaper]

[Snaper]

So, the mighty Dave Brown suggested to check things with a scope again so I did.
I've checked the output of the shift register and also patched something to the outputs of the 4081s.

For sure, the shift register output is stable, but there are extra random bits at the outputs of the 4081.
I don't know why, both chips are clocked from the same source, but, either the 4081 picks up something that forces it to generate extra bits or there is something else here.

Important, musically it is usable, since the random bits are in sync with the clock.
If you add bits via the switch to the shift register, you can turn down the pot and it is stable and ok and with the pot you can even control the randomness.
...just drives me mad that I can't understand why this happening...

[Snaper]

So, did some more investigation.
I was all wrong earlier.

As a tip, I've used a simple audio to "hear" the first AND gate pins.
As soon as I've touched the pot, the data comes from the 4094 gets "wonky" - out of rhythm/sync.
Why is that?

Also, the data output of the 4094 is more like a square.
More like the this.

---

IMG_7681.jpg
Description:
Filesize:	1.18 MB
Viewed:	103 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[electrotech]

Hi Snaper,
I've just seen this thread.... and had a look at your circuit.
You mention the 'comparator' pot a few times... I'm wondering if this pot has an intermittent connection on it's wiper ; this would cause the output of the op-amp/comparator to go high randomly.
I'd suggest you try adding a decoupling capacitor (say 10nF) between the wiper and ground. It's value will depend on the value of the potentiometer but I'm assuming it will be 100k ?
This will remove any noise at the comparator's input, especially when moving the control.
As for your 'scope photo - this looks like what you'd get if you're using AC coupling.... there's certainly a capacitor in-circuit there somewhere.

Hope this helps.

Andy

[Snaper]

[Snaper]

Tried with a 10nf ceramic, then with a 100n.
Neither of those helped - the damn issue is still there.
And it is still weird, since it works with some clocks…