I have some XR-2206. When I have some time I can wire it and try to reproduce your circuit and behaviour.
Post the circuit diagram you are currently using.

 

Thanks,

I removed all optional components from the circuit to make it easier:

 

Try adjusting the "trigger level" on your scope. Knob may have other names. Post model number or photo of front face of your scope.

I am talking about the voltage on screen that causes the waveform to lock (synch).

You may be just outside the range on your lower voltage.

 

I reassure you the trigger level in the scope is in sync. If I move, left or right, I will get it out of sync.

This is another picture that I took, at lower frequencies, and which shows the oscilloscope is not out of sync:

This one is at +-6V:



And this one at +-7V:



And this one is at +-6V and out of sync:

 

Small chance that this is the problem, but are you sure that your power supply isn't going out of regulation when you turn up the voltage?

 

I also thought about that, specially since I'm using an external circuit that I made to split the current, but I just tested the supply with the oscilloscope and it all looks good; and it also does the same with a single supply, just that it fails at over 12V instead of +-6V.

 

Tie pin #9 high (greater than 2 volts above negative supply). Your waveform looks like an FSK event is happening. You may have some static electricity or other issues causing the timing to jump between pin 8 RC network (non-existent) and pin 7 RC network.

See Figure 13 on the datasheet.

 

I think you might be in the right track. Setting pin 9 2V over the negative didn't work, but when I touched the pin it reproduced the failure under the +-6V threshold.

 

Hello,

This is stated in the datasheet about PIN 9:

Bertus

 

You can try putting your Timing resistor on pin 8 and pulling pin 9 to ground (negative power rail). Just in case that works.

 

I thought there was something wrong or interferring in the breadboard, so I changed everything to a new location. Mounted the first example on the datasheet (figure 2) for the 3rd time -exactly as it is-. Then tested the circuit with the two resistors on pins 7 and 8, with and without the FSK (on pin 9), on both triangular/sine and square outputs, and everything with single supply...

No change. Everything works perfectly as expected, every single mode and option; but only as long as I'm between 10 and 12V. As soon as I pass the 12V, the output turns into this:

 

You can try putting your Timing resistor on pin 8 and pulling pin 9 to ground (negative power rail). Just in case that works.

I just tried it: it does nothing when connected to ground, and results in a flat line when connected to the negative in a split supply.

There has to be something wrong with it. I guess I can still use it anyway, but only with 12V or +-6V.

 

did you move your timing resistor from pin 7 to pin 8 when you pulled pin 9 to ground, (or in your case, negative supply)?

 

Just checked it, and F2 does the same as the output from F1; it works well only under +-6V.

So far I checked everything, every option, every mode, three different circuit configurations, and both in single and split supply voltage. They all behave the same way: everything works as expected, but only under 12V (or +-6V)... unless for very high frequencies, which allows you to increase the voltage to the maximum allowed.

 

OK. It seems you have checked everything in the function generating circuit.
Here are some things that you may not have tested:

Have you looked at your power supply voltages when the circuit is misbehaving? Maybe the power supply is oscillating or going out of regulation.

Another possibility is that your timing cap or a power supply bypass capacitor is breaking down at the higher voltages.
It just occurred to me... maybe you are installing a polarized cap in the wrong polarity -- although I would expect smoke if the cap were on the power supply.

... unless for very high frequencies, which allows you to increase the voltage to the maximum allowed.

I don't understand what you mean here. Are you saying that the circuit works correctly at very high frequencies and not at low frequencies? If so, at what frequencies does the circuit work correctly and incorrectly. Maybe the low frequency timing cap is bad or is a polarized cap installed in the wrong polarity?

 

I put an XR-2206 on a breadboard and tested it with the circuit shown in the datasheet (as in post #5) as well as the circuit you have shown on post #22.
The oscilloscope is being triggered on CH1 from the square wave output on pin-11 with a 10k-ohm pull-up resistor.
Triangular output on pin-2 is observed on CH2.
The chip works ok with Vcc from 9V to 28V.
I could not replicate your observations.

 

After rebuilding various circuits more than a dozen times, there's a very small chance that I'm doing something wrong; also checked the oscilloscope, individual components, and power supply. It has to be a malfunctioning chip; which is not surprising since I got it on ebay from China.

Thanks for all the help...

 

I am having the exact same problem and it was driving me nuts. I found that it would operate in the same range that you had specified, but beyond that it gave me an extremely unstable output. I checked everything, just as you have, and it is definitely the chip.

I also purchased mine from China, and as you are having the exact same problem and purchased yours from there as well, I am pretty confident that those chips are counterfeits. The chip seems to work fine, just not in the voltage range shown in the datasheet (which makes me question whether it lives up to any of the other specs as well).