XR2206 with crystal

The XR2206 may be that good, but the resistor/cap rarely are!

And it's 20ppm/deg C.

It's my knee-jerk reaction to laugh at anyone proposing to use RC oscillators in anything better than 20% (200000ppm) tolerance situations. And even at 50% I'd be dubious.

Tim.

You could use the XR2206 as the VCO in a phase-locked loop, with a crystal oscillator as the reference. The sine wave is formed by shaping a triangle wave, so you can't just stick a square wave in there and get a sine out at the same frequency. Or go nuts and use a DDS.

Best regards, Spehro Pefhany

XR2206 typically operates upto 1 MHz. And within that range it exhibits a 20 ppm stability.

So, unless you intent to run the part on a 32.768 Khz crystal in an oven .... :-)

But seriously: no, I don't think you can.

But what's missing from the original poster's toolbox is that one can simply start with a crystal oscillator, and there you have it.

Like so many posts, people are asking questions for things they don't yet have the background for.

It sure sounds like the poster thinks crystal oscillators only come as square waves, and so he has to build up a workaround like the 2206 in order to get his sinewave.

Far, far simpler to just build a crystal oscillator that puts out a sinewave. Up until recent decades, that's what most oscillators did.

Michael

The pins for rounding of the triangle wave to a sine wave are very sensitive for capacitance.

If you look in an oscilloscope at the zenith of the sine wave generated by 2206 you will see a spike, lower or higher depending on the wires connected to pin 13 and 14. Longer wires/tracks gives a bigger spike.

Conclusion, if you want clean sinus waves, don't use the 2206.

If you want a function generator with triangle, square and sine wave, use 2206.

I have one connected as a triangle wave generator, because that is the only wave form I need.

For sine wave I use a tone generator with low distorsion.

Yes, it's more "nuts" from the pov of the technology that's brought to bear on the problem, sort of like using a micro to blink an LED. It may well be the best solution if a simple xtal oscillator isn't a practical solution for some reason.

And they need an output filter, which might be very simple for a fixed frequency. And they are generally SMT, which might be an issue for a hobbyist.

Best regards, Spehro Pefhany

Hello Spehro,

That's not nuts. Nowadays DDS chips have become a staple and are cheap, sometimes less expensive than these function generator chips.

The only problem might be that DDS chips usually need to be told in some digital format what frequency to put out, versus the simple potmeters of the olden days.

If the OP wants to check out DDS chips I'd start with Analog Devices.

Regards, Joerg

Hello Spehro,

True. Like the wood splitter I just used. A friend from church had built it. Big WW2 era motor and a hydraulic cylinder that could press north of

BTW, I once did blink an LED with a micro. Actually it did two short bursts each interval which had to be really precise. The micro was only a Dollar and the crystal and other stuff around 25 cents.

The filter is pretty simple unless you want to use them all the way to the Nyquist limit. Yes, SMT can be a bear.

Usually when people use a 2206 they want something flexible in frequency, that's why I thought of a DDS.

Regards, Joerg

Hello Roger,

I am not that fond of them either. Personally I have always gone the discrete way when designing oscillators.

The only time I used (or rather had to use) a 2206 was at university and there I have seen this effect on the scope. Didn't like it...

Even then I didn't.

I'd probably do it with a current source.

Regards, Joerg