One shot pulse generator problem

[armdnrdy]

I have bread boarded a small section of a circuit, and can't seem to get it to go with the "stock" values.
This section allows the sync pin to accommodate either a positive going pulse, or a negative going pulse. This filter/diode section is straight out of the CEM3340 data sheet, and is included in synth VCOs such as the Digisound 80-2.





The problem that I'm encountering is...with the monostable one shot that I've incorporated...I cannot get voltage to pass through C2 and C29.
If I change C2 and C29 from .001µf to .1µf, change R16, R31 from 10K to 1M, and connect R31 to ground instead of -V...I can get this circuit section to pass the one shot pulse.



The original filter values set the cutoff at 15kHz. The adjusted filter is 1.6hz. I tried adjusting the one shot resistor/capacitor values for a shorter pulse time....to no avail. It is noted that the pulse width on the sync pin doesn't matter because it reacts to the rising and falling edge.

I've tried pull ups and pull downs of different values....and different locations, still....no go.                   
The only thing that seems to work is adjusting the sync input C&R as outlined above.

I'm curious what a Digisound 80-2 user would do if he wanted to add manual sync pushbuttons to the module with stock components.
Any ideas on this?

[PRR]

Turn D11 around?

As shown, it is slow drop. D11 doesn't really do much. The CMOS will tend to speed the transition, maybe not enough.

D11 the other way, it will DROP when button is pressed. The CMOS will output a sharp rise, which will pass through the C-R differentiator.

[armdnrdy]

Thanks for the reply Paul.

I pulled the one shot from "A guide to debouncing part 2"
http://www.ganssle.com/debouncing-pt2.htm

It states, "The diode is an optional part needed only when the math goes haywire."

I tried pulling the diode completely....no difference. I still can't get the pulse to pass the .001µf cap.

[anotherjim]

Could the coupling caps be err... constipated? Like a charge pump cap with the reverse bleed diode missing does. In this instance, you may need a reverse diode from the .001u cap/steering diode junctions to the most negative supply rail.

[armdnrdy]

Could the coupling caps be err... constipated? Like a charge pump cap with the reverse bleed diode missing does. In this instance, you may need a reverse diode from the .001u cap/steering diode junctions to the most negative supply rail.

Thanks for the reply Jim.

Adding this diode wouldn't explain why the data sheet circuit section which was used in commercially available synth modules doesn't work.

I believe that the problem lies in the "pulse generator" that I am using.

I have used this 4093 circuit to control other CMOS ICs without incident but...as I stated above...this is actually a debouncing circuit rather than a "one shot"
Yesterday I had the thought to incorporate a traditional 4093 one shot after the debouncing circuit to see if it makes a difference to the circuit section that follows.

I'm surprised how little information there is on the net when I type in...synth trigger circuit, Manual trigger button synth, etc.

[ElectricDruid]

The pulse generator is set up with 18K and 1u. That gives about 0.5 secs to drop. The schmitt trigger should give sharp edges again on the output (does it? if not, there's your problem). You need those sharp edges to get past the HPF.

What was the need to increase the HPF resistors on the sync inputs so much? 10K to 1M is a big jump up. Ok, you're not going to need to reset an LFO at audio rates, but at the same time, the original value would have worked for slower sync too. Maybe try 100K as a compromise?

HTH,
TOm

[anotherjim]

I was just coming back to say pretty much what Tom said!
When I've had trouble getting a pulse through the C it's been the input rise time too slow.

This is coming from buffered CMOS, so you'd think it was fast enough, but you may be suffering from breadboard capacitance.
And you can't just change to 74HC logic because of the 10v supply.

Try adding simple bjt inverters as a speed-up.
If the + pulse used a pnp, it'd get the + edge fast and likewise npn for the - edge. That would reverse the meaning of your switches because of the inversion.