Hi People!
Urgent help is needed. I have to make a Voltage Controlled Oscillator with Frequency varying Around 500 Khz or higher (more is better) and sinusoid output.
The problem is: I can only use transistors, capacitors, diodes, inductors and resistors (no varactors, Integrated Circuits etc. ) I want to make something simple (5-6 transistors at most), but all circuits I find on the web, either use varactors or IC's, or they are too complex for my level of knowledge (I just can't analyze them).
Any advise would be very appreciated. Maybe anyone has done something similar?
Thank you.

 

The imposition of the restrictions as to what you can and cannot use to implement your oscillator are a bit severe. Is this a class assignment or are the contraints self-imposed?

hgmjr

 

This is an assignment. I tried really hard to find useful circuits on the web, but it seems really hard. I am not sure that I can do that myself. In the attachment is an example of what I have found on the web. They use op-amps as a differential amplifiers, but i think they can be modeled with simple differential amplifier. This circuit suits all my requirements, except that it doesn't work (In NI Multisim 10 at least). But this is an example of what I am looking for.

 

Is this perhaps an assignment to reinforce a topic you have recently covered in your class? If so, what circuit configuration do you imagine your instructor is expecting you to use?

hgmjr

 

This is an Analog Electronics course (3rd year) final project. We mostly studied operation of BJT and MOS transistors, spending much time on differential amplifiers etc. As well as concepts of feedback and stability. Assitant told me that solution is based on using positive feedback, but also mentioned that we will have to make some research on he web.

 

A couple of possibilities come to mine. You can start with either a wien-bridge oscillator or a twin-t oscillator. You would then need to come up with a way to vary one or more of the frequency determining elements using voltage control.

You did not mention how much frequency range adjustability you are needing to provide.

hgmjr

 

Look for an oscillator circuit you are familiar with
A Wein bridge should be able to reach 500Khz for instance.
Decide what the frequency determining components are
For instance they are two resistors and two capacitors in a Wein bridge
Think how you can make one of these vary with voltage.
Since varactors are out that leaves resistors in my example
FET transistors make good voltage dependant resistors, that do not need a ground reference and have isolation between the control voltage and the oscillator.

 

This is an assignment. I tried really hard to find useful circuits on the web, but it seems really hard. I am not sure that I can do that myself. In the attachment is an example of what I have found on the web. They use op-amps as a differential amplifiers, but i think they can be modeled with simple differential amplifier. This circuit suits all my requirements, except that it doesn't work (In NI Multisim 10 at least). But this is an example of what I am looking for.

Don't rule out that it doesn't work! Eventhough it is probably the way you created the circuit in Multisim that is causing the problem, simulations have been known to be a bit faulty at times.

I had an analog design project last term and experienced this first hand. A really good way is to sit down with some ideas, build it, and check it with an oscilloscope. I have seen a lot of people walk into the lab, with their fancy multisim schematics along with pretty waveforms, to find out soon enough that their circuit is hopeless.

Steve

 

Thank you studiot.. I think Wein circuit can be of much help here.. I did some research on its variations, looks promising..

 

There is another member here in the forum that is looking to do basically the same thing you are doing. Here is a link to a recent reply that I posted. You may be interested in this approach also.

hgmjr

 

Here is a link to my reply to the other poster dealing with a similar design. In it I have furrnished a link to an example of a transistor-only wien-bridge circuit that can serve as a starting point for your design.

hgmjr

 

By making components in one arm of the Wein bridge much larger than the other it is possible to go for a single resistor control. See my sketch 1. you will have to sort out gain stabilisation and other stuff in your amp.

A second approach does not yield a sine output. Consider any oscillator that works by cyclically charging and discharging a capacitor through as resistor. If you connect the charge/discharge path to a variable voltage this will control the rate of charging and thus the frequency of the oscillator. My sketch 2 shows an astable multivibrator working on this principle.

 

I have a really nice LC VCO design (no varactors, 7 transistors), but I'm at a loss as to how to describe it so our OP could design one, short of providing a schematic. This sorta defeats the point of homework.

 

I have a really nice LC VCO design (no varactors, 7 transistors), but I'm at a loss as to how to describe it so our OP could design one, short of providing a schematic. This sorta defeats the point of homework.

I'll just bet you ran out of room writing the design equations in the margins. Perhaps another 300 years or so will pass until the design is rediscovered.

 

I'll just bet you ran out of room writing the design equations in the margins. Perhaps another 300 years or so will pass until the design is rediscovered.

That's sorta how I got it. I "rediscovered" it. About 40 years ago, I was designing PLLs for video timebase correctors, and while researching VCOs, I ran across a Bell and Howell patent that was very clever (as patents are wont to be). I'm thinking it used an LM1496, although I'm not sure they were available then. At any rate, that's what I used. Discrete transistors can replace the IC, circumventing our OP's requirement that he not use ICs. I just don't want to give it away, since this is a homework assignment.

 

The circuit I described in the audio VCO thread can be used at 500kHz if you use high-speed switching transistors in the multivibrator. I was able to sweep it in simulation from about 100kHz to 1.3MHz. The range can be changed by changing the caps.
Here's what I posted in that thread:

You can make one with a two-transistor astable. Use a current mirror to convert the control voltage to a proportional current which controls the frequency. I simulated one, and it had a fairly linear transfer function.
It uses 5 transistors.

 

Hi again. I decied to use wein bridge, but i wasn't able to increase frequency to 500 KHz. Not even to 250 KHz Anyone can suggest such design? Should I modify the circuit somehow? I used the basic wein bridge: 1 op-amp, 2 capacitors and 4 resistors.

 

The design of a wien-bridge that operates at 500KHz is going to involve the use of transistors. Most Operational Amplifiers are not likely to have sufficient open loop gain to implement 500KHz.

hgmjr

 

Here is a two transistor wien-bridge that may serve as a guide to a design of your own. Scroll down to the second schematic for the wien-bridge design.

Also, I think the use of a photo-cell and a white-LED as the variable resistance element will surely provide the adjustment while at the same time providing a measure of isolation between the control signal and the oscillator.

hgmjr

 

hgmjr: Is it a night-bulb circled on the schematic of wein bridge you gave me? Can i use something else there? I don' think using LED or bulbs will be accepted as a solution, though it seems very reasonable to use them.