Capacitor Discharge Effect Troubleshooting

[mattoverse]

I am using a 470uf to ground capacitor on the power input to a 4093 oscillator circuit.  By having a momentary normally closed switch after the power switch I am able to achieve a cool effect as the capacitor is draining when the circuit is opened, really does some fun things depending on how many oscillators you have running and their speed etc..  The problem I am having is sometimes repeated pushing of the momentary switch sort of locks up the 4093 or something and it does not always restart even with main power switch being toggled.  I do have a 100k pullup resistor in series from power to pin 1 as an attempt to avoid any lockup but it does not seem to be cutting it.  Ideas? 

[mattoverse]

Below is a demo of the unit and from about 1:30 - 1:55 a good example of the issue I am running into it.  I think the momentary push button switch may be faulty, I rewired it and it seems to work a bit better but the issue is still there.

[anotherjim]

What you have is an example of just how power supply dependent a Schmitt Trigger oscillator is. It does make a cool space drum effect (if space drums were ever cool). 2 or more in the same chip can interact sounding something like oscillator sync in a synth.
Somewhere on the web, there's a scheme for a VCO. The Schmitt Trigger oscillator chip is powered by the control voltage.



With a little re-design, the individual gates could all be tuned to different frequencies.

A danger with CMOS is in having input voltages that can be higher or lower than the supply pins. Despite having internal protection diodes which should clamp excess voltage, it can cause lockup of the internal logic due to the so-called SCR (Thyristor) effect formed by successive P-N junctions through the chip layers. It doesn't destroy the chip, but it does cause it to freeze until all power has been removed.

Can you post your exact circuit, because I'm sure lockup can be avoided.

[mattoverse]

I'm running the oscillators into each other.  Here is the schematic, it's not that neat but just needed to get the schematic done quick so I could move onto board layout.  The extra grounds are so that I have enough ground pads for input/output etc..  SW1 (not listed) is also ground.  P1N2 is just Pot 1 / Pin 2(Center) etc..

[anotherjim]

I can't see what you are doing with the rest of the timing components. Presumably, you have potentiometers as variable resistors between the gate outputs and the timing capacitors? If so, you should also have fixed resistors in series with the variables so you cannot have excessive currents in the gate outputs. With 9volt supply, 10k fixed resistors should be good.

I think the values of your timing capacitors are too large. When you drop the power supply, they can be holding too much charge for too long providing input voltages greater than the power supply. This often leads to CMOS latch-up as I described in my previous post.

Because of the high input impedance of CMOS gates, you can have much smaller capacitors together with larger value timing resistors. For this type of project, you can use cheap ceramic capacitors in values up to 100nF which combined with 1Meg ohm potentiometers will give you all of the audio range and some. For modulation low frequency range, film capacitors up to 680nF will get you into the sub-1Hz range.

Note the size of the timing components in the VCO scheme I linked. 220pF and 2.2Mohm!

[mattoverse]

Thanks! I appreciate the feedback and insight.  Yeah, the capacitor values/potentiometer values were selected to provide the particular frequency sweep I wanted for each oscillator, but it sounds like using small caps with larger value potentiometers may be a better solution along with the series resistors you mentioned.

[Tony Forestiere]

Matt: Very neat sounding toy.
Not meaning to derail, but listening to the lockup at 1:30-1:50 cracked me up. Seems analogous to "flooding" a carburetor.
*best Southern drawl* "Yep. Too much juice in the carb. Ya gotta mash the gas all the way down an' hold it so ya get more air in her".
Never mind me. Carry on.

[mattoverse]

I am thinking maybe a momentary switch that connects pin 1 to ground might function as a simple reset feature for any potential lockup.  Maybe not surprisingly, since rewiring the switch the lockup problem has been greatly reduced.

I am also wondering if I could reduce the size of the power drain cap but maintain the discharge time I am after by putting a resistor in parallel with it?

[duck_arse]

what about if you wired a large value resistor and a big reservoir cap in series. connect the IC supply to the cap +. the resitor goes to supply. you could have a shorting switch across the resistor, the supply will "dump" into the cap (for an upswell?), and then open the resistor connection, so the only current is supplied by the reservoir cap. maybe needs a discharge limiting resistor between cap and IC supply pin, I haven't really thought this through.