CMOS + Negative voltage = ???

Repeater · Joined: Mar 19, 2011 Posts: 13 Location: Rochester

So lets settle the score... CMOS can or can't work with negative voltages without comparators? Is this something that is just 'not advised' because the chips don't respond to negative voltages or is it dangerous because it damages the chip/circuit? If anyone has a technical explanation I would love to hear it.

Perhaps a way to ask this that would be even more relevant to this forum... has anyone experienced any unique musical result by messing with V- and CMOS?

blue hell · Posted: Sat Aug 06, 2011 7:55 am Post subject:

Negative voltages is not a problem in its own, it's all voltages that go outside the supply voltage range that the CMOS is powered with. So it cant go over Vss and not below Vdd or .. usually it can exceed the margin by about 0.3 V (to be found in data sheets).

CMOS inputs are protected with diodes that will come into action when you go out of the specified range, what happens then depends on the actual current flowing - it may destroy the diodes causing the input to be no longer up to specs.

So you as a designer need to take care that this situation does either not occur or that when it occurs the current is low enough for the input protection diodes to survive it.

DGTom · Posted: Mon Aug 08, 2011 7:39 pm Post subject:

check out Ken Stones Psyco LFOs & Rob Hordijks Benjolin schematics for good practical examples of alternate CMOS power set-ups for synth use.

Rene Schmitz has a good article on switching full range signals with CD405X ICs & there is a good description in Mr. Stones notes about his Sequential Switch PCB to do with the same.

I like running CMOS off +/-5V

it won't save you any parts, its still easy to kill stuff, but,
it can make a cct. abit more straightforward (sometimes not)
to imagine / follow / understand for a lumpen like me Very Happy

ezekiel · Joined: Oct 17, 2009 Posts: 30 Location: Columbus Ohio

By +/- 5v, DGTom, I assume you mean you connect CMOS chips' Vcc to your +5v supply and CMOS chip GND to your -5v supply. Thus, the CMOS chip is running off 10 volts and sees inputs that swing 10 volts peak-to-peak.

Neat idea. I'll have to consider that.

Inventor · Posted: Tue Aug 09, 2011 2:54 pm Post subject:

Interesting indeed, I had not thought of doing that. I kind of like it for many circuits, especially linear CMOS. I can also see it causing some confusion but only in rare cases. I use +9V supplies (battery in the past and wall-transformer in the future) so I may not use this approach, but it is quite curious. Hey, whatever works for you right?

Les