16-Step sequencer external start problem?

[elmegil]

I am working on an integrated panel + sequencer MOTM PCB for Synthcube, to eliminate the wiring.... (Synthcube arranged with Ray's estate to keep the MFOS PCBs available for the DIY community). See the attached picture for the current prototype.

Everything works and this version lines up nicely with the panel, but I have one weird problem that I'm looking for input on. I'm hoping someone else with one of these can help me figure out if something is wrong or if this is just some weird corner case behavior.

I'm using an Oakley Triple LFO to generate a 20ms pulse into the External Start. I have the mode switch set to "stop", and the stop step set to 15. I'm running the rate at its highest value (which gives me about a 40Hz clock), and of course the pulse from the LFO is coming in less frequently than the time it takes to go through 15 steps of the sequencer.

The behavior I expect to see is: pulse comes in, resets and starts the sequencer, which then gets to step 15 and stops again.

What I'm actually seeing is *almost* that -- but the sequence isn't consistently starting at step 1. Sometimes it starts at step 3, sometimes 4, sometimes 5. Right now the settings I'm using it's pretty consistently starting at 3 or 4, or else starting at 1.

I'm looking at the clock output with my scope and it is clear that sometimes it does fewer steps than others, so this isn't a case of buzzing through the first couple of steps so quickly I can't see the LEDs light up. It really is skipping the start of the steps.

I've verified the resistor values in the area -- I had earlier found that R42 & R43 were incorrect values in my BOM but I've corrected it and the behavior persists.

Looking at the reset pin of the 40193 counter, when the external start comes in it just generates an RC spike there, not a square pulse. However, the pulse is high and wide enough that per the datasheet it should be having no problems correctly resetting the counter to 0. (Datasheet says 150ns is the worst case minimum time for the pulse, and the pulse itself is on the order of hundreds of microseconds, well beyond the 150ns spec.

But it appears for all intents and purposes that the counter isn't getting reset all the way to zero consistently. Sometimes, even most of the time, it does, but sometimes it's not.

If I mess around with the rates, and the pulse width from the LFO an so on I can get it to a state where it is more consistently resetting, but it is not hard to find the states where it doesn't behave.

The only other major difference I can see here is that I'm using a 15V supply, not the 12V originally spec'd but according to Ray's notes it should work from a 9V to a 15V supply, and of course this is hardly the first one of these to use a 15V supply in an MOTM system.

So the question is: has anyone else seen this behavior? Is it just a "quirk" where some combination of LFO width, rate, and clock rate create glitchiness, or do I need to keep digging into why it's behaving this way? LIke I say, everything else appears to be working fine, including the reset button (though the reset button gets a bit more processing to give a nice square pulse of 4 to 5ms).

---

IMG_3371.JPG
Description:
Filesize:	1.35 MB
Viewed:	19 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---