Extra simple signal generator

[MohiZ]

Hello to one and all! I just got my first used oscilloscope today and already made a nice find with it. I took the bottom schematic from http://www.play-hookey.com/analog/sine_wave_generator.html and tweaked the values for a bit and I got a really nice looking sine wave with adjustable frequency. You can get "pretty low" and "pretty high" frequencies. The amplitude changes a little with the frequency, but not too much, and the waveform looks like a perfect sine wave on the screen!

The zener diodes are 5.1 volts. Without them the amplitude seems to fluctuate slowly. The 16.8k and 18.39k resistances are pretty precise - I needed to use two series resistors to get these values. Other combinations might work better or worse!



I didn't draw the power supply but you can use a normal voltage divider to get the 4.5V on the ground.
Just thought you guys might find it useful as well

I'm not sure if I'm allowed to post this picture, but if I'm not I'll take it off.

Edit: the schematic

[MohiZ]

The circuit is definitely "dirty" and could be improved in many ways. It works the best in the middle half of the pot rotation, and with small or high frequency the output drops from several volts to maybe a hundred millivolts. But it's simple and only uses one pot!

Here's some screens (they have been scaled and there's actually more difference between the frequencies).

[aron]

Thanks a lot for this!

Aron

[~arph]

Thanx, I will definately try this.

[MohiZ]

You might want to make the 8.39 resistor a pot to tweak the right amount of attenuation - make it too high and the signal'll be low, make it too low and the signal will clip.

[MohiZ]

I made a proper schematic of the whole thing and put it in the general gallery. I finally settled on a 100k lin pot, although a 250k lin pot would get ya a lower lowest frequency. The cosine output can be omitted. Also, the circuit would probably work better with a 12V supply (I surmise that you could then replace the 10k pot with a 10k resistor without clipping), but I haven't tried it.

Operation instructions: The 100k lin pot changes the frequency. The 10k pot is just for initial setting up (it attenuates the signal [it also changes the frequency but only slightly, so never mind that]). First turn the 10k lin pot all the way up, and put the 100k pot in the middle. This should create a slightly clipped wave. Turn the 10k knob down until you get a clean wave on all of the frequencyes (roll the 100k knob up & down to see that the wave remains clean). Now you can leave the 10k pot in place or replace it with a fixed resistor!

Should it have an output loading resistor? I didn't use one, but it seems to work without it both plugged into an amp or an oscilloscope.

[waky]

Neat circuit! im going to build it, one question though, what are those 2 zeners there for?... im guessing some kind of amplitude regulation? but i dont know 

[cloudscapes]

that's the simplest I've seen that makes sine waves! thanks!

[MohiZ]

Neat circuit! im going to build it, one question though, what are those 2 zeners there for?... im guessing some kind of amplitude regulation? but i dont know

Yeah, they're supposed to be there for amplitude regulation, although if you run it at 9 volts like I do they don't do much of anything and I think you might as well leave them out. At 12 volts they are supposed to make the circuit more stable. Feel free to experiment.

Btw, if you build this and come up with something to improve this with, you're welcome to post it! I'm sure a better combination of resistors could be found, but this works for me and I already boxed it.

Edit: the exact circuit in the schematic produces sine wave from circa 330 hz @ several volts to 16 khz @ 100 mV peak to peak. The voltage really starts to drop only after about 10khz maybe. Try a 470k pot instead of the 100k one to get below a hundred hertz.

[~arph]

Works like a charm. I've omitted the zeners (seemed to make no difference on the scope) and replaced the 10nF with .47uF, I also used a 500k pot for speed. Now have a nice sine LFO. great I was looking for this for a long time!!

[Thomeeque]

I've omitted the zeners (seemed to make no difference on the scope)..

Funny, I'd expect that exactly those two zeners are responsible for the sine curve shape (without them I'd expect triangles on the output)..

[~arph]

Yes, that is what I thought as well.. But my guess is that the cap to ground(Vref)  is doing most of the work now.

[MohiZ]

Funny, I'd expect that exactly those two zeners are responsible for the sine curve shape (without them I'd expect triangles on the output)..

According to the page I found the original schematic from, the zeners are just for amplitude modulation, to create some feedback in the cosine generator opamp. I didn't see them contributing to much of anything either, but I left them in the circuit "just in case".

[MohiZ]

Works like a charm. I've omitted the zeners (seemed to make no difference on the scope) and replaced the 10nF with .47uF, I also used a 500k pot for speed. Now have a nice sine LFO. great I was looking for this for a long time!!

That's great! What's the lowest frequency that you get with this setup? I was thinking of implementing this into a tremolo.

[~arph]

Less then 1 Hz    You should be able to tailor it to your needs. I have to do a lot more experimenting on it.