Colpitts, sinusoidal waveform,maybe crystal?

Thread Starter

cesare.magno

Joined May 19, 2021
18
hi everyone i have to design a colpitts oscillator (f = 4MHZ), the circuit is this. The output waveform is good but there are no 170n inductors that can be mounted on a board to test the circuit. I decreased the capacitance value to raise the inductance value but the output is no longer sinusoidal. The output signal from the hits I use to generate a square wave. I thought of using a pierce circuit (colpittz with crystal), maybe the waveform is better, I know that the crystal goes in place of the inductance but I tried to mount it and the circuit doesn't work, I have to adopt other arcongimenti to put the crystal? Which ? thank you all!!
(the first circuit is with the inductance at 170n the other has been scaled for an inductance value that can be mounted on a breadboard but waveform is not sinusoidal)
 

Attachments

Yffig

Joined Aug 25, 2019
8
170nH for a 4 Mhz oscillator is a rather small value. If you want to stay with this value, it very easy to build one that will fit a board (air inductor)... it's around 9 turns of wire on a 5 mm diameter , with a length of approx. 10mm allowing you to adjust the oscillating frequency (and of course you may vary the length, diameter or # of turns as needed by your form factor constraints).
 

Thread Starter

cesare.magno

Joined May 19, 2021
18
thank you !! but if I wanted to put a crystal at 4 Mhz, I just need to replace the inductance with the crystal? Should I change the order of magnitude of the capacities? if I used capacitances of 10p the value of the inductance to use would be 316uH. The crystal should be able to adjust the equivalent inductance right?
 
Last edited:

Yffig

Joined Aug 25, 2019
8
thank you !! but if I wanted to put a crystal at 4 Mhz, I just need to replace the inductance with the crystal? Should I change the order of magnitude of the capacities? if I used capacitances of 10p the value of the inductance to use would be 316uH. The crystal should be able to adjust the equivalent inductance right?
You'll find attached a LT-Spice model for a 4.0 MHz crystal I made and tested with 3 options: only the Xtal, with 15pF in // and 10 pf in series that show how these capacitors will affect the resonnant frequencies. Note that LT-Spice allows to put those values as parasitic elements of a Capacitor. Then have a look at:
https://www.electronics-tutorials.ws/oscillator/crystal.html
where you will find a Xtal Colpitts osc for 1 to 5 MHz.
Have fun !
Xtal model for LT-Spice.jpg
 

Yffig

Joined Aug 25, 2019
8
The data for modelling the 4.0 MHz Xtal went from a datasheet that , unfortunately I 'm just not able to retrieve (I've just made a search on my current PC that is not the one I used about 10 years ago when I made the Xtal model) .
There came from a vendor in France (iBizness near Lyon that went out of business).
At that time, I tested the Xtal on the bench with a VNA (minVNA Pro 100KHz...200 Mhz) the simulation and the results were very close (that is the serie resistor in really around 100 Ohms at series resonance).
As you may have noticed, these values are very accurate and I cannot have invented them: they were actual data from the manufacturer: the inductor value is very high (hundreds of mH), the series capacitor is very low (a few femtoFarad !), the series resistor gives a Xtal Quality Factor (serial resonance) of about 95 000 (this is a Xtal)!
You can definitely trust the 4 values given for the single Xtal...if any doubt about the quality of your sample, just increase the value of the serie resistor.
Have fun !
 
Top