Advantages and disadvantages of 4049 oscillators?

[midwayfair]

Like in the Anderton tremolo.

It looks to me like it wouldn't actually take up much room compared to a dual op amp oscillator, and I've seen posts and videos that imply that you can get an almost pure sine wave out of it, as well as a true square wave and triangle wave, and even out-of-phase oscillation for stereo etc.

But I don't see them used very often. Is there a reason? Do they create noise issues? Is there some sort of difficulty getting them to work properly? Is it just that it's been done? Lack of mojo? Not enough depth or rate range?

[Mark Hammer]

My first hunch is that, while one can use a TL022 or LM358 or other low-current dual op-amp for an LFO, there is no lower-current alternative for the 4049.  I honestly don't know if they require more, less, or the same current when operated as oscillators....but it's worth asking as a question.

Of course, that really only matters if there are batteries involved.  If the assumption is that an external supply will power the circuit, then current consumption is only a concern insomuch as it bears on the production of any voltage spikes on the supply line.

[anotherjim]

Current consumption is the main problem, at least for battery operation. You can reduce current if you lower the supply voltage, but that gives less sweep range.
Sometimes I use 4093 if I need several oscillators and a triangle or square will do - the timing cap has the triangle and it's only 3 components - the cap, a resistor and the pot - plus an op-amp voltage follower so as not to load the cap.

[duck_arse]

the 4049 has a wierd pinout, w/ 2 N/C's. the 4069 looks more righter, and is smaller, with supply pins where you expect w/ cmos. have you looked on the Nat Semi "linear cmos" and "cmos gate oscillator" app notes?

[R.G.]

The technology that generates an LFO waveform has little to do with depth, range, and so on. Those are determined by how well the LFO waveform  - in both AC and DC characteristics - matches what the modulator needs, which varies hugely. There really isn't any such thing as a general application LFO. Every one of them has to be both generated, then adapted to the modulator.

CMOS inverters were designed as digital elements. They accidentally, as a side effect, happen to do some analog functions. How well they do that depends on the side effects of how they're used. CMOS inverters can be used as inverting amplifiers with poor to very poor DC accuracy and low gain. This makes them do integrators OK but not great. That may be enough for an LFO. CMOS amplifiers are noisy compared to other amplifiers, which again may not matter to an LFO. They are tricky to get working with a specific voltage supply, as they may pull lots of current when biased linear from a larger supply. So they have quirks to be taken care of. Compared to opamps, which are designed for flexibility, you may have to adapt your circuit to the CMOS, not make the opamp do what you wanted it to.

So - no particular reason, except that there are more side effects of using them, more things to consider, and possibly more adaptation to do when you finally get to matching them to a modulator. Since this is a real design problem, not just a put-the-blocks-together process, they don't get used as much. It's more work to get right.

[midwayfair]

Thanks for the thoughts, folks. I'll look at them a little closer and also check out the 4069.

[anotherjim]

The one thing that may be a "must have" is that it's so simple to turn a triangle into a sine with a CMOS inverter. It's such a shame that it takes a whole 14pin to do it.
There are ranges of logic, including 4000 series equivalents (Rohm), that are single elements in a 5 pin SMD package.
http://rohmfs.rohm.com/en/products/databook/datasheet/ic/logic_switch/standard_logic/bu4s01g2-e.pdf
...but the inverter is a buffered one. Damn!

[amptramp]

I did some work at one time trying to find out if CMOS gates operated as amplifiers in liquid nitrogen would have acceptable noise characteristics.  We got them to operate, but the drain current operating one amplifier and leaving the rest of the gates tied either low or high gave a current drain of 8 milliamps in air and more in nitrogen.  It worked, but every once in a while, there was a tremendous transient crack sound then it would go back to normal operation.  The noise level was still worse than other amplifiers we were using at the time.  Since CMOS is not rated to -196°C / 77 K, we figured maybe we were overstressing the IC (and we were using the ceramic body military temp range unit rated from -55°C to +125°C).