Oscilator proboem

[bzblues]

Hello everyone!

I have a little project on the workbench that I'm having trouble with.

I must say that I'm a beginner, so be patient if I'm not using the right terms.

The project consist of six simple oscilatores (using the SS9018 transistor) conected to a simple "ruby amp" the one with the LM386 chip. Everything works with 9V.
I also added a guitar jack so I can play guitar while the oscilators are playing as well.

So, I group all the oscilator outputs into one, conected them to a 100k potentiometer so I can control the volume of the oscillators independently from the guitar signal, and conected the oscilator out signal along with the guitar signal to the input of the Ruby amp.

Now, obviously I had some issues because the guitar signal affected the oscillator signals and viceversa. So I did some reseach and I found out that I must add two "buffers" using the TL072 op-amp only. I followed the instructions on how to connect everything but all I'm getting is a squeaky noise, so obviosly something is not working out propperly. I checked and re-checked every connection, and everything seems ok, but still doesn't work out.

I'm guessing that it has something to do with the 9V surelly. Anothe option was a "voltage divider" (two 10k resistors, one to 9V and the other to GND and from both resistors to feed the buffers, but that also didn't work out.

Again, sorry if I'm not describing it the right way, I'm learning.

Thank you!

 

[dick_freebird]

May be simple "conducted emissions" / "conducted susceptibiity"
trying to mode-lock your oscillators. Give a try to some decent
local supply filtering, RC for low power or LC for higher power.
Or maybe even (say) 78Lxx regulator per oscillator to break things
up. But common signal + power ground (?) would leave a lot of
icky potential.

You might have a better time too, using a split supply if the input
signal from the guitar is ground referred. A "single supply op amp"
may "work down to ground" but when you get real close your output
gets real weak (CMOS) or just plain wrong (saturating BJT). This
might get into somethig.

 

[KlausST]

Hi,

schematic?

Klaus

 

[betwixt]

Pretty sure you are seeing either instability or "injection locking" between the oscillators.

Instability comes from having some signal conduct where it isn't wanted and getting back to an earlier stage where it causes a loop. Usually this comes down to poor layout, especially the ground connections or poor power supply decoupling.

Injection locking is the situation where the output of one oscillator upsets another. If you consider an oscillator, especially one using a simple single transistor, is an amplifier with some kind of frequency selective feedback, if another signal close on frequency enters the feedback loop it mis-triggers the oscillation cycle and tries to pull the frequencies together. By combining the output of several oscillators you are very prone to that kind of problem. The idea of using buffer stages is to isolate the oscillators outputs from each other by adding an amplifier, even if it has zero gain, between the oscillator output and the summing point. It stops anything at the summing point getting backward to the oscillator itself.

Ideally, you circuit should be:
Oscillator --> Buffer --> level control --> Buffer --> summing point

That method isolates the oscillator frequency from being pulled by adjustment of the level control and also presents the summing point with a constant impedance.

The amplifiers must work at the 9V supply you have to hand but for audio work they do not need rail-to-rail output capability or any special bandwidth. You do need to bias their outputs to somewhere close to half the supply voltage but you can do that with a single pair of equal value resistors across supply and ground connected in parallel to all the buffers if done carefully.

Brian.