hi, I am building a circuit with op-amp to generate a square wave. The circuit diagram is shown here: http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/square.html#c4

The op-amp I use is TL074 and when I run it, there is a problem: I get the square wave in the first few minutes (like 20min or so) and then the output becomes a constant voltage. All my resistors are 10k and my capacitor is 4.7uF.

Can anyone help me fix this problem? I want the circuit to generate the square wave constantly for up to 12hr.

Thank you!

 

What are the supply voltage(s)?

 

If you restart it, does it start running again?

 

Seems like there should be a R3 from + supply to + input. Difficult to discharge C to below ground when discharge path is ground or near ground.

 

the supply voltage is 12V

 

Yes, it runs again, but the same thing happens

 

the supply voltage is 12V

+/- 12VDC? Or just 12VDC? Note that the circuit you linked to requires a +/- power supply.

 

The oscillation runs fine for about 20 minutes, then stops? Maybe it's the JFET-input opamp latching up. Do you have a different opamp model you can try?

 

+/- 12VDC? Or just 12VDC? Note that the circuit you linked to requires a +/- power supply.

It's + and - 12V DC

 

The oscillation runs fine for about 20 minutes, then stops? Maybe it's the JFET-input opamp latching up. Do you have a different opamp model you can try?

This is the only op-amp I've got... I think the op-amp is still working though, because when I unplug the power supply and put it back on, it works again

 

It's + and - 12V DC

And "ground" is at 0V, right? In other words, you haven't confused V- and ground?

Is there a load on Vout that might be causing the op-amp to overheat and shut down?

 

And "ground" is at 0V, right? In other words, you haven't confused V- and ground?

Is there a load on Vout that might be causing the op-amp to overheat and shut down?

GND is 0V and when I run the circuit, the op-amp doesn't get hot, so I guess there's no overload

 

OK, just wanted to check the "obvious" stuff.

How about putting a 0.1µF decoupling cap on the op-amp's power pins? Maybe it's falling into an oscillation of some kind.

 

OK, just wanted to check the "obvious" stuff.

How about putting a 0.1µF decoupling cap on the op-amp's power pins? Maybe it's falling into an oscillation of some kind.

for a decoupling capacitor, do you mean that I can simply connect Vcc+ and Vcc- with a capacitor?

 

Run a cap from each power pin to ground...reduces the chances of hf oscillation swamping out your square wave. Also, tell us about your main timing cap. What value? What type? Have you tried swapping it for another?

 

for a decoupling capacitor, do you mean that I can simply connect Vcc+ and Vcc- with a capacitor?

No. Use two caps. One from Vcc+ to ground and the second from Vcc- to ground.

 

I'm guessing your 4.7uF is electrolytic...they're not always great for oscillators. If bypassing doesn't help, try increasing the value of R and using a smaller C...a value you can get as a film or ceramic.

 

Or just try a different cap! See if you can get oscillation at any frequency.

 

the output becomes a constant voltage.

What voltage? Also, where are you connecting the ground?