Hi guys,

I am a relative beginner trying to make a simple triangle wave (audible frequency) oscillator using TL082 op amps. My efforts on this front have been driving me a bit nuts.

This is the schematic I decided to use: http://chemelec.com/Projects/Osc-2/Oscillator-a.png

I know that op amps like this one use both + and - voltage and I have never dealt with a bipolar power supply before. Based on what I have read recently, however, it looks like just by following this schematic the two 10k resistors on the left will be dividing the voltage to set up a virtual ground and provide the + and - voltage the op amp needs.

So I built the circuit on my breadboard and it did not oscillate at all. I used a 12v regulated ac/dc adapter and just connected + tip to where it says "12 Volts DC" on the schematic and the ground symbols to the -, as I normally would. It looks to me that this schematic was designed so that a person with zero understanding of the concept of bipolar power supplies could just build it like any other circuit and it would still work, so that is what I did... am I missing something though? Is the negative supposed to be hooked up to something/somewhere else?

The only thing I did differently from the schematic was that I didn't use a tantalum capacitor because I didn't have one. I used a 1 uf electrolytic at first, then tried it with a few ceramics in parallel to add up to 1 uf, with no change. However, I figured this couldn't be the problem because it seems like it is just a decoupling capacitor, and either way it was still the right value.

I checked the voltage at the point that is supposed to be the triangle wave output and it was steady at 10.52 V. I checked the squarewave output and it was exactly the same. I didn't do the calculations but I'm guessing that is nothing more than the result of the resistors that it has passed through up to those points. I checked all my connections on the breadboard and had a friend (with about as much experience as me, but who also has a basic ability to read schematics) check them too. I even looked at the PCB image on the site where I got the schematic and made sure my layout matched up with the traces on there. I concluded the problem must be one of two things:

1) my op amp chip is messed up, in which case, I am sorry for wasting your time.
or
2) I am doing the power supply setup wrong. Shouldn't the voltage always be less than 6 V after being divided? The fact that I saw a voltage over 10 makes me think that somehow I am failing to divide the voltage.

Sorry for the long post! I was trying to make it detailed to minimize the need for follow-up questions.

Thanks for your help!

 

What is the voltage at pin 2 and pin 5 of the op amp? It should be 6V. Measure the voltage at all the other pins also.

 

I simulated your circuit on proteus and it was working fine.

BTW how do you know that your circuit is not working? Did you check the outputs with a scope or a logic probe. I used 0.1uF in place of 1uF and got a period of 65mS which is about 15Hz. Are you using LED to see the outputs?

Take a picture of your breadboard and show us. May be we can spot your mistakes or advise you what logical steps to take next.

Allen

 

Hey, thanks so much for your responses.

I have been testing the circuit by hooking it up to an oscilloscope (analog) and testing at both the square wave and triangle wave output points. At both points, the scope has consistently shown the same thing as my multimeter: a steady voltage a little over 10 volts.

It actually occurred to me to test the voltage at pins 2 and 5 right after I posted originally because I realized it should be 6 volts and that would be a good way to test if the voltage was being divided. For some reason it is actually 9.77 volts there. Based on that, it seems like there is no way the op amp is getting the bipolar power it needs. So now I know that I am not dividing the voltage to create the virtual ground, but I don't know why. I know how to read a schematic and it seems pretty straightforward just to follow it and set it up.

I've been looking at schematics for other circuits like this and they all seem to show the same thing and it seems to be the same thing that I did! I am stumped. I will try to take a picture of the breadboard and post it though. I might also try to redo it on another breadboard with a new set of components... maybe that will help me notice something I didn't before.

Thanks for all your help though!

 

Somehow I forgot to ground pin 4 of the op amp! I can't believe it. I had checked it so many times without noticing. After making that connection, the voltage measured around 6 volts at pins 2 and 5, as expected, but there were still no oscillations, with the square wave and triangle wave outputs showing steady voltages. I just went to Radioshack and got another TL082 so I'm going to replace that. I'm thinking that maybe having it all plugged in to the 12V without being grounded, thus exposing it to 12V at so many points, damaged the chip. I'm not sure if that is likely/possible but I am going to just try switching it out and see if it works.

 

Its impossible to tell if a circuit is oscillating or not with just a DMM, unless the oscillation is so slow that you see the voltage go up and down slowly.

However above just a Hz or two, your gonna need oscilloscope to see if the cir is oscillating.

 

....................

However above just a Hz or two, your gonna need oscilloscope to see if the cir is oscillating.

Read post #4.

 

Finally got it to work!

After grounding pin 4 of the chip and THEN replacing the chip, I finally got it to work... although, the so-called square wave is definitely not a square wave (looks like somewhere between a sawtooth and triangle wave on the scope, sounds like a sawtooth) and the triangle wave is not really a triangle wave (smooth on tops, sounds a bit more like a sine wave).

That's actually OK for my application, but I am kind of curious about why this is happening.

In any case, thank you all for your help! It turns out it was (of course) all just because of a stupid/careless mistake I made.

 

Finally got it to work!

After grounding pin 4 of the chip and THEN replacing the chip, I finally got it to work... although, the so-called square wave is definitely not a square wave (looks like somewhere between a sawtooth and triangle wave on the scope, sounds like a sawtooth) and the triangle wave is not really a triangle wave (smooth on tops, sounds a bit more like a sine wave).

That's actually OK for my application, but I am kind of curious about why this is happening.

In any case, thank you all for your help! It turns out it was (of course) all just because of a stupid/careless mistake I made.

Those descriptions sound like the signal is bandwidth limited. This is either because you're trying to run the oscillator too fast for your op amp, or the layout is limiting the bandwidth, or the scope and PROBE combination is limiting the bandwidth.

 

Or the cap is too small for the low audio frequency he is testing at.