Hi, I would like to design an oscillator without using any ICs like LM741 and such. Is it possible? I am thinking if a simple LC circuit would do although I'm not so sure if it will work properly. I hope you could help me. Thanks!

Gab

 

Hello,

Most sinewave oscilators using fets or bjts are used for high frequency.
The following page will give you the fundamentals of them:
http://mysite.du.edu/~etuttle/electron/elect16.htm

Bertus

 

Look up "Wien bridge oscillator".

 

Thank you for replying on my thread, Alec and Bertus. I just have one last question. How can I design the bandwidth of the oscillator? I'm thinking about changing the values of the capacitors and resistors but I would like to do it mathematically like solving for the values of the components. I hope you could help me. Thank you very much.

 

Hi, I would like to design an oscillator without using any ICs like LM741 and such. Is it possible? I am thinking if a simple LC circuit would do although I'm not so sure if it will work properly. I hope you could help me. Thanks!

Gab

The Franklin oscillator; https://www.google.co.uk/search?q=f...pf4BA#imgrc=x_Umb05kJfiLBM:&spf=1501087136022

Is a possible option - no taps or feedback windings and it can be rigged for low damping of the LC network.

 

The Franklin oscillator; https://www.google.co.uk/search?q=franklin+oscillator&tbm=isch&imgil=x_Umb05kJfiLBM%3A%3Bh2m-4aMEhlId3M%3Bhttp%253A%252F%252Ftheradioboard.com%252Frb%252Fviewtopic.php%253Ft%25253D6486&source=iu&pf=m&fir=x_Umb05kJfiLBM%3A%2Ch2m-4aMEhlId3M%2C_&usg=__6PXNKmWszNsOX_eG2N9w2bUYMQA=&biw=1280&bih=908&ved=0ahUKEwiHwdWVsafVAhUjKcAKHQnvBU8QyjcIQA&ei=cMV4WYfJNKPSgAaJ3pf4BA#imgrc=x_Umb05kJfiLBM:&spf=1501087136022

Is a possible option - no taps or feedback windings and it can be rigged for low damping of the LC network.

Thank you for replying, ian. Do you have any idea on how could I set the bandwidth to a value of 50kHz? What I know is that the formula of Bandwidth is B=Fo/Q where Fo is the center frequency and Q is the quality factor.

Fo=1/2pi(LC) (the LC depends on what type of ocillator) and Q=2piFL/R where F is the frequency L=Inductance and R=Resistance.

I hope you could help me. Thanks!

 

Thank you for replying, ian. Do you have any idea on how could I set the bandwidth to a value of 50kHz? What I know is that the formula of Bandwidth is B=Fo/Q where Fo is the center frequency and Q is the quality factor.

Fo=1/2pi(LC) (the LC depends on what type of ocillator) and Q=2piFL/R where F is the frequency L=Inductance and R=Resistance.

I hope you could help me. Thanks!

Unless you want to get into bandpass filters - build the highest Q tank circuit you can and spread it by adding a "Q-spoiler" resistor.

Not sure how this applies to an oscillator though - they usually just burst into song at whateve frequency they have the most gain.

If you want a range of frequencies - tune it.

If you want loads of frequencies all at once - bandpass a noise generator.

 

A sinewave oscillator does not have a bandwidth. It produces only one frequency at any time, and if it is a clean sinewave then it has no harmonics.

ak

 

A sinewave oscillator does not have a bandwidth. It produces only one frequency at any time, and if it is a clean sinewave then it has no harmonics.

ak

You can design an oscillator that's easily "pulled" - but most people don't.

 

Ohh i see. So what I should do is to add a bandpass filter in order to measure the bandwidth of that signal?

You guys really helped me. More power to all of you.

 

You can design an oscillator that's easily "pulled" - but most people don't.

Most, if not all, oscillators are, by their very nature inherently pullable of pushable, and a great deal of time has gone into closing down that can of worms.

 

@Gab Cruz Apparently we do not understand what you mean by "bandwidth".

If you want to measure the noise or other modulation of the output of an oscillator in the frequency domain you can use a spectrum analyzer.

If you are not referring to noise or the oscillation frequency, to what are you referring?

Just guessing, but does the image below answer your question?

 

Maybe Gab Cruz means tuning range.
What frequencies do you need from the oscillator? And what is the output level, and what do you want to drive with it?

 

@Gab Cruz Apparently we do not understand what you mean by "bandwidth".

If you want to measure the noise or other modulation of the output of an oscillator in the frequency domain you can use a spectrum analyzer.

If you are not referring to noise or the oscillation frequency, to what are you referring?


View attachment 131968

My best guess would be distortion - The TS may be struggling with various frequency domain artefacts like a faint whiff of harmonics.

 

I like the mental image that generates: "...a faint whiff of harmonics."

 

I like the mental image that generates: "...a faint whiff of harmonics."

A notch of parasitics just after the peak isn't going to generate clearly defined harmonics like a square wave.