Designing 1MHz Sine Wave Oscillator for AM

[bejoynp]

Dear All

I want to design a 1MHz sine wave oscillator to be used as carrier wave for Amplitude Modulation. Which will be the best approach in terms of frequency and amplitude stability and why?

Can the technique be implement using high slew rate op amp and also can we use crystal oscillator?

Thanks and regards

 

[Averagesupernova]

The best frequency stability would be a crystal oscillator. For amplitude stability you could square the signal up and then filter the heck out of it.

 

[vk6kro]

There are a lot of crystal oscillator circuits or you could get a 1 MHz oscillator module.

These give a square wave output, but this doesn't matter if you are driving other stages which include some filtering. They should only cost a few dollars.

You can also make a variable frequency oscillator (VFO). These are not as stable as a crystal, but you can change their frequency.

It is usually bad practice to modulate the oscillator itself. It is better to use the oscillator to drive another amplifier (a buffer) which then drives the amplifier that is being modulated.

 

[skeptic2]

I have seen commercially available crystal oscillator modules, from a reputable manufacturer costing about $125 20 years ago, fluctuate in frequency and amplitude enough to cause problems for our high precision frequency reference.

You also may consider phase locking a colpitts oscillator to the crystal reference.

 

[bejoynp]

Dear all

Thanks for your suggestion. If I don't want to use a crystal, am I still left with any options or is it mandatory to use crystal for 1MHz sine wave without having to filter square wave

 

[Mike_In_Plano]

My personal fav is a Colpitts with your good 'ol 2n2222. Model it up in Spice and don't forget to put a little series resistance between the capacitor tap and emitter.
For improved stability, you can use a silver mica for the lower value cap.

1 MHz is a low frequency, so it's easy to play with using run of the meal transistors. However, the gain of most op amps will be nothing, or next to nothing, at these frequencies, so stick with a transistor oscillator.

If you have the makings for transformers, you can make a fairly straight forward modulator out of a couple of transformers and four signal diodes. Just build a balanced mixer and feed your modulation signal into the middle leg with a DC offset.

Finally, you can just put an RF amplifier after the oscillator and modulate the power supply to the amplifier. It's aweful and dirty, but it's the way AM stations weree built forever ;)

Best Luck

\Mike

 

[vk6kro]

There are many variable frequency oscillator circuits and you could make quite a stable one at 1 MHz with careful mechanical construction.

Two common ones are the Hartley and Colpitts oscillators. It will involve making quite a large inductor though.
See this link and the list of links at the bottom of it:
http://en.wikipedia.org/wiki/Hartley_oscillator

Broadcast radios for the AM medium wave band are often available free and the oscillator coil for one of these would be suitable for making an oscillator covering 1 MHz.

Crystals for 1 MHz are usually expensive compared with higher frequency crystals, but you can buy them.
The 1 MHz oscillator module does give a square wave out, but it is generally a cheap way of getting a stable 1 MHz output. Most discarded computer boards have crystal oscillator modules on them but these are usually made for higher frequencies.
The filter does not have to be very complex. It could be just a series tuned LC circuit with the output taken across one of the components.

 

[Averagesupernova]

Keep in mind that a higher frequency clock source such as from a motherboard can be divided down to 1 Mhz.