twin-tee oscillator clipping

[elMickotanko]

Hi,

I've built this oscillator from **broken link removed**.

The oscillator stage works great, it looks like an almost perfect sine wave except for the peak of the negative part of the wave is slightly flattened.

The emitter follower section however extremely clips the negative half of the wave. Is this because the bias point is too high? I take it this stage is getting a bias from a DC voltage from collector of 1st transistor, how do I work out what it is and what I should change it too? Or could I put a capacitor in between and another resistor from supply to base?

Any help in understanding this circuit would be greatly appreciated.
Thanks,
Mick.

 

[Mikebits]

Most likely bias point is set to low on the follower, but I would caution value changes on the first stage as it will probably affect your osc. Perhaps you can isolate the follower from the ocs with a cap and setup your own bias for it.

 

[audioguru]

The oscillator transistor is biased wrong. Change its collector resistor to 1k and it will work better.

 

[Mikebits]

Changing the 4.7K may change the loop gain of the osc.

 

[elMickotanko]

Hi,

Thanks a lot for the replies.

The oscillator transistor is biased wrong. Change its collector resistor to 1k and it will work better.

I changed it to 1k and it's bang on now. Perfect sine wave. Can you explain why it is 1k and how you calculate it.

Cheers,
Mick.

 

[audioguru]

Hi,
Can you explain why it is 1k and how you calculate it.

The 2N3904 has a current gain of about 230 so with the total base resistor at only 54k and its collector resistor at 4.7k then it is almost saturated with a collector voltage of only about 1.2V. No wonder the bottom of the waveform is clipped.

If a 1k collector resistor is used then the collector voltage is about 2.3V then the transistor is not saturated (but it is nearly saturated).

 

[elMickotanko]

Thanks! Sorry, Ive been away. The 1k wasn't ideal as it started clipping again, more so when the battery ran down a bit.

Any lower than 1k killed the oscillations so I set up a seperate bias for the second stage and its great now. I added a switch now to select between guitar level and keyboard level output and used the ESP "better volume control" cos I had no log pots lying around. Finally stuck it in an iPod shuffle box.

Thanks for the help.