Crystal oscillator using discrete CMOS inverter ?

Thread Starter

joeypc

Joined Jan 31, 2011
3
Hello Everybody,

I built a discrete CMOS inverter as shown below:

upload_2015-5-2_23-28-31.png

This CMOS inverter worked as an logic inverter very well. But when I tried to apply to crystal oscillator as below. It did not work at all! Could anyone here explain why ? Thanks all.

upload_2015-5-2_23-28-52.png
 

Papabravo

Joined Feb 24, 2006
21,157
My guess would be that the gate capacitances of the two MOSFETS are preventing the circuit from oscillating. For some additional insight you might try replacing crystal X1, with its equivalent circuit. Something like this:

1655307773540.png
 
Last edited:

Ian0

Joined Aug 7, 2020
9,667
Also, if the output biases to half-supply, each Vgs will be 2.5V, and an IRF9540/IRF540 will have very little gfs at that voltage.
 

LvW

Joined Jun 13, 2013
1,752
The feedback network for an inverting amplifier must be able to provide a phase shift of -180deg at the desired oscillation frequency. This can be realized using a 3rd-order lowpass in the classical ladder topology R-C-L-C (both C grounded). The inductance L can be realized with crystal - however the resistor R is missiung in your feedback network.
 

sparky 1

Joined Nov 3, 2018
756
I like to use circuit below even though there can be fewer parts, in developing and testing xtal osc more info usually helpful.
The load capacitance related the impedance and drive level. Starting with an abbreviated circuit is difficult at a concept level.
Notice on a more complete schematic the resistors RF and RL on the first stage. setting into linear region, the other feedback resistor. Notice 2nd stage VL
When all the numbers work Rout will be exact. Some of the mosfets used for crystal drivers are high gain low input capacitance. Off hand some start with BF.
Finding motional parameters on a series and on a parallel mode xtal oscillator and using online calculator is usually sufficient to design good Oscillator.
I have socketed Colpitts and Pierce circuits for crystals that are very adjustable in order to narrow the scope to get the finite parameters.
When all the pieces of the puzzle are known the simulator using equivalent circuit is handy.

https://www.giangrandi.org/electronics/crystalfilters/xtaltest.html
1655310177829.png
 
Last edited:

Audioguru again

Joined Oct 21, 2019
6,671
The datasheets for those IRF power Mosfets show a Vgs of 4V for some of them to barely conduct a very low current. They need a Vgs of 10V to fully conduct.

If you make a crystal oscillator with low capacitance little transistors then an amplifier made with IRL540 and IRL9540 Mosfets will produce a powerful output. The IRL Mosfets fully conduct with a 5V Vgs.
 

Ian0

Joined Aug 7, 2020
9,667
The feedback network for an inverting amplifier must be able to provide a phase shift of -180deg at the desired oscillation frequency. This can be realized using a 3rd-order lowpass in the classical ladder topology R-C-L-C (both C grounded). The inductance L can be realized with crystal - however the resistor R is missiung in your feedback network.
You don't often see the "R" in circuit diagrams for the Pierce oscillator using a 74HC04, but it is there as the output impedance of the gate. A physical resistor is only required for low frequency oscillators.
The requirement is to have an abrupt change of phase at the oscillation frequency. The RC gives the first not-quite-90° then the crystal or L-C network gives a change from 90° to almost 270° at the oscillation frequency
 

sparky 1

Joined Nov 3, 2018
756
Ian0, The generic schematic I think was a low voltage type I had modified so probably came from Dave Johnson ALDINC thanks
another source FET Principles And Circuits — Part 3 | Nuts & Volts Magazine (nutsvolts.com)

Just having fun with mosfets circuits 100 tested projects. traditional schematic p,140 there abouts. The crystal oscillator with fanstock clips. not too fancy. The audio and test equipment looks fun too. To me fun is high priority. 3N187 is an N-Channel Dual-Gate Depletion Type Transistor With Monolithic Gate Protection Diodes, used in RF,IF Amplifier and Mixer Applications up to 300 MHz Old stuff, analog type oscillators almost extinct.
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwiF5LzJh7D4AhVbJkQIHQeNB3I4ChAWegQIBBAB&url=https://worldradiohistory.com/BOOKSHELF-ARH/Author-Groups/Rufus-Turner/MOSFET-Circuits-Guidebook-Rufus-Turner.pdf&usg=AOvVaw121hFDDVF7DbwuUoyKuWJc
 
Last edited:

ag-123

Joined Apr 28, 2017
276
try one of those 74HC chips like 74HC04, like audioguru mentioned, Vgs could be a problem ('old' mosfets for some of them they want like 7-10V to turn on). 74HC can switch on as low as like 2v.
And like others mentioned, power mostfets has additional capacitances, which could change the params. In addition, if the crystal is in the Mhz range, I'd guess try to figure out if at those frequency, it may be 'RC filtered' out that there is not enough gains to oscillate. I think the phase shift isn't 180 deg, it could be like 90-135 deg but enough to cause a positive feedback loop due to the rather high gains in the transistor.
 
Last edited:

Thread Starter

joeypc

Joined Jan 31, 2011
3
The input capacitance of the Power Mosfets you are using is much Too High for the crystal oscillator circuit.
I think the 12MHz frequency is even too high for a Cmos IC oscillator. Old LS TTL ICs were used many years ago.
I cropped and enlarged your schematic to make it easier to see the text:
Dear Audioguru,

Thank you very much for your kind reply. Well, Could I ask you more question ? I would like to know what is the role of 0.25uf capacitor in the TTL version ? Thank you !
Clipboard01.jpg
 

Audioguru again

Joined Oct 21, 2019
6,671
Old-fashioned TTL logic? I used it 45 years ago.
The 0.25uF capacitor prevents the two 1k DC biasing resistors from causing negative feedback which would prevent oscillation.
 
Top