Square-wave synth

[pq]

Hello,
I wanted to begin making my own synth, so I decided to follow this tutorial, I bought all the components (except the DC jack, I bought a 9V battery instead, and the perfboard I bought is a bigger one) plus a switch so I could turn it on and off, however, the schematic was really confusing and I couldn't make it work. I would like to know if there are any other guides on making a simple square wave synth which can be made using the components I bought for that other one.

Thank you in advance.

[andrewF]

Hi
to e-m

Probably best to try troubleshooting your circuit rather than jumping to a new one. It is pretty normal to find circuits do not work straight away, one tiny mistake can stop everything.

What part(s) of the schematic are you not sure about?

[pq]

There are 10k resistors on the circuit which aren't on the parts list, and also three 1M resistors, so I believe there is an error (either on the parts list or the circuit). I don't understand what to do with the +4.5V and -4.5V parts.

I'm new to electronics, though, so this circuit might just be too advanced for me...

[-minus-]

I think part of the problem here is the project has not been documented well and is unclear. There are much easier ways to make an oscillator. I'm not sure if you are able to source parts locally but getting hold of a 40106 IC is perhaps a better way to do this.

If you are a beginner I think it's important to get a result quickly with minimum components or you'll be turned off this electronics thing. I've attached a simple oscillator schematic. It's from the Beavis Audio site. The pot size and capacitor value are not all that important at this point. Pin 14 of the 40106 goes to +V, pin7 to ground.

It would be very worthwhile getting a small breadboard to experiment on. That way you could try different component values and see/learn what happens. Here's a link to the Beavis Audio site too:

http://www.beavisaudio.com/Projects/CMOS_Synthesizers/


EDIT
Just incase you want to have a go at trying to get what you might have built so far working.... The 9V battery + and - are connected together with two 10K resistors in series. This is creating a ground point where the two resistors join. This must be taken to all the parts on the schematic where you see those ground symbols (the three little lines). The + terminal of the battery has to go to pin 7 of your 741 op amp. The - terminal of your battery must go to pin 4. Maybe try this first and see if you can get this working.

---

Schematic-40106-Simple-Oscillator.png
Description:
Filesize:	6.71 KB
Viewed:	1414 Time(s)

---

[defog]

For that Beavis audio oscillator schematic, you'll need to add the decoupling cap to block DC voltage from hitting your speaker/amp. You don't need the other resistors after it or the volume control necessarily though:

[pq]

Sorry for the late reply - I kind of forgot about this thread.

I checked out the Beavis Audio site, and bought a breadboard. The projects are really easy and good for starting.
I didn't try making the 741 oscillator again, found out there was a thread in this same forum talking about it and I found an easier schematic if I ever want to try that, but I made a circuit using the Beavis Audio CMOS oscillator and a 4040 which was mentioned in the site. Since I didn't get my breadboard yet (I wasn't able to buy one locally), I figured out it would be clever to post it here to be sure if it could work.

[analog_backlash]

Hi pq.

This looks OK to me, but you should also connect the reset pin of the 4040 (pin 11) to 0V (ground) as you are not using this. The reset can be connected to one of the counter outputs to reset it (believe it or not ) after a certain number of input pulses at pin 10. It can also be reset by an external trigger pulse. If you want to play with this, you should connect pin 11 to ground via a pull-down resistor (typically 100K). This ensures that the input is never left floating, even if you have nothing connected to it.

I'm not sure about the order of your outputs. You have 9 (divide by 2), 7 (divide by 4), 6 (divide by eight), 5 (divide by 16) then 4 (divide by 128). Perhaps the fifth one should have been 3 (divide by 32)? Also, I'm unsure what you are going to do with the switches (but there may be a perfectly good reason why you've put them there...).

Anyway, it looks like you're on the right track

Gary

[pq]

[analog_backlash]

Ah, so you're building a top-octave generator - I understand the switches now! In that case, the reset pin probably won't be needed (but still keep it at 0V, with or without the pull-down resistor).

Gary

[pq]

Is there any formula for determining the frequency of a 40106 oscillator? I tried 1/4.7(µF)R(Ohm) = 1046.5(hertz), but I'm not sure about that.

[analog_backlash]

I've seen 1/RC or 1.2/RC quoted, but by experiment it was a bit out (but usually in the right ball-park). The NatSemi datasheet (a bit attached) shows a more complicated formula, but you need an oscilloscope to tell you some of the values required . If you've got a scope, you have the frequency anyway!

Gary

---

NatSemi 40106.gif
Description:
Filesize:	13.98 KB
Viewed:	15 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[Cynosure]

4.7uF caps and a 10k resistor probably won't give you the range you want. With a topic octave divider you need to atart with a high frequency. Try 0.1uF instead.

[JovianPyx]

Looking at this, I spot something that can often be a real pain in the butt.

That is a virtual ground. This is the bit of circuit you created with two 10K resistors to create an effective +4.5, ground (or zero volts) and -4.5.

This technique can work for circuits that are very simple and small and which do not require putting much current into or drawing much current out of ground. When larger and/or more demanding circuits are used, it begins to become a point of failure. One way around this is to "buffer" the virtual ground. This is done by adding a single opamp wired as a voltage follower. Feed the current "gnd" point into the opamp non-invert input and then put the opamp output to the "gnd" connection. This will allow the ground to deal with more current and will be somewhat self correcting should the ground try to drift.

That being said - my suggestion - don't use virtual grounds. You can make a nice +/- 9volt dual supply using two 9volt batteries. Such a supply is far more immune to noise propogation and floaty ground problems than a virtual ground. I know it requires another battery, but the advantage is "stability".