Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Ultrasonic oscillator

Status
Not open for further replies.

Granny

New Member
I want to build a sinusoidal oscillator circuit around an op amp or transistor. About 100kHz at a volt or so.
My first thought was that it should include a crystal for frequency stability, but the 2 problems I had are that I can't locate a crystal of the right frequency ( there doesn't seem to be anything between about 32kHz and 1 MHz) and that even if I could I don't know how to find its inductance so I can sort out the tank circuit.
I then breadboard some RC oscillator circuits I found online, but either they don't oscillate at all or the output waveform is a long way from sinusoidal! I look at the waveform using a Pico/laptop combination.
Help with any of these problems would be appreciated.
 
Look at a "Phase shift oscillator" - they produce a pretty good sine wave.

I'd suggest trying it at eg. 10KHz first then change the timing components for the frequency you actually want.
If you use an opamp, it should be a type rated for much higher frequency than you are working with.

Also make sure the power decoupling is good, eg. something like 100uF and a 0.1uF in parallel.
 
You can find 100 Khz xtals on ebay, they were quite common in communications receivers back in
the day.


There are programmable oscillators also -


And design methods - https://www.ti.com/lit/an/szza043/szza043.pdf


Regards, Dana.
 
Look at a "Phase shift oscillator" - they produce a pretty good sine wave.

I'd suggest trying it at eg. 10KHz first then change the timing components for the frequency you actually want.
If you use an opamp, it should be a type rated for much higher frequency than you are working with.

Also make sure the power decoupling is good, eg. something like 100uF and a 0.1uF in parallel.
Thanks! I thought I had tried that, but when I set it up just now it worked well at 10kHz. I used 1.2k resistors and 10nF capacitors with a 100k variable in the feedback loop of an op amp. Perhaps I got the feedback wrong before.
At just under 100kHz the waveform was not as good, but it will be amplified and used to generate noise using a small speaker that will be operating a bit beyond its published frequency range so the noise should be a reasonably pure tone.
 
A 3-stage phase shift oscillator needs a loop gain of 27. At 100 kHz that's a gain-bandwidth product of almost 3 MHz. That's a lot for a standard transistor or opamp. You would need something with extra bandwidth.

Digi-Key has 35 crystals in stock in the 50 kHz to 200 kHz range.


ak
 
Last edited:
A 3-stage phase shift oscillator needs a loop gain of 27. At 100 kHz that's a gain-bandwidth product of almost 3 MHz. That's a lot for a standard transistor or opamp. You would need something with extra bandwidth.

Digi-Key has 35 crystals in stock in the 50 kHz to 200 kHz range.


ak
Thanks. I hadn't realised about that gain- bandwidth product. Now I have a working oscillator I would prefer to stick with it rather than move to a chrystal oscillator.
The op amp I am using is a quad TLO74CN, and according to a data sheet I have just looked at online it's unity gain bandwidth is 4MHz. I assume that means its gain-bandwidth product is the same? Is that so - it's not something I have needed to know before? If so it should be ok.
 
I assume that means its gain-bandwidth product is the same?
yes, two ways of saying the same thing... Ft is the same thing, but for transistors., and i see some transistor data sheets also call it GBW or unity gain frequency.
 
Did you hear already about a GIC-based oscillator? (GIC: Generalized Impedance Converter).
It is one of the best two-opamp oscillator circuits - in particular, for high frequencies - because the non-idealities (limited bandwidth) of both opamps cancel each other (up to a certain degree).
See for example here:

 
Last edited:
Thanks. I hadn't realised about that gain- bandwidth product. Now I have a working oscillator I would prefer to stick with it rather than move to a chrystal oscillator.
The op amp I am using is a quad TLO74CN, and according to a data sheet I have just looked at online it's unity gain bandwidth is 4MHz. I assume that means its gain-bandwidth product is the same? Is that so - it's not something I have needed to know before? If so it should be ok.
Have you checked the PDF link I posted?, it takes you through the entire design process of the 100KHz oscillator? - you don't need to know, or calculate, anything.
 
Have you checked the PDF link I posted?, it takes you through the entire design process of the 100KHz oscillator? - you don't need to know, or calculate, anything.
I hadn't read it, but I have now.
If I can get the 28V 24mA bulb I shall give it a go.
 
I hadn't read it, but I have now.
If I can get the 28V 24mA bulb I shall give it a go.
This is the well-known WIEN oscillator. It also can be stabilized (instead of a bulb) using two diodes or any other amplitude-dependent resistive element (thermistor or FET)
 
I hadn't read it, but I have now.
If I can get the 28V 24mA bulb I shall give it a go.

I don't think the bulb is that critical, I seem to remember I used something different, that I got from RS Components. The point of it is to easily and cheaply stabilise the output and give low distortion.

This is the well-known WIEN oscillator. It also can be stabilized (instead of a bulb) using two diodes or any other amplitude-dependent resistive element (thermistor or FET)

There's a specific thermistor for Wien oscillators, the RA53 - but expensive if you can find one.

Here's another useful page:

 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top