500 - 3000 MHz Signal Generator 4

[mrkenneth]

What is the simplest circuit for generating a sine wave or a square wave from 500 up to 3000 MHz? The signal does not need to be modulated; just a simple sine wave or square wave.

I would prefer the power output to be greater than -10 dBm at 50 ohms, but I can add an amplifier if needed.

The output frequency to be controlled by a voltage (using a potentiometer or a DAC) or digitally via I2C, SPI, etc.

I have looked at VCOs, but I have not come across any that can cover such a large frequency bandwidth.

Thanks in advance!

---------------------------------------------------------
Operation Radiation:

http://www.mrkenneth.com

 

[BrianG]

Agreed, VCOs do not cover the bandwidth directly - but here's a standard way of getting almost any frequency span:

Use a fixed frequency reference oscillator to feed one input of a mixer. A VCO feeds the other input of the mixer and the output is the difference of the two frequencies.

If you use high enough frequencies for both the reference oscillator and the VCO, the mixer difference output gives the required variable frequency span for a (relatively) small change in the VCO frequency.

 

[biff44]

Mixing two VCO's together is a great idea. You do end up with a lot of harmonics and spurious products, however. Also, If you need low phase noise, that is probably not the way to do it.

 

[zeitghost]

I've got a sweep oscillator that uses that very principle.

It's about 50 years old.

Uses the mains as modulation by varying the magnetization of a core in one oscillator and thus varying the inductance.

All thermionics of course...

 

[nbucska]

The simplest is : Mix a fix osc. and a VCO.

Depending your application, you may be able to
use the result without filtering.

How fast do you have to change the Fr.? Is step
change acceptable? Range switching (manual/auto) ?

Would mechanical tuning be OK? perhaps with motor ?
If square wave acceptable, you can tolerate harmonics?
What about other noise e.g. mixing artifacts?








<nbucska@pc33peripherals.com> omit 33 Use subj: ENG-TIPS
Plesae read FAQ240-1032

 

[nbucska]

Spirit:
I used for up to 200 MHz a regular ( TV IF type) pentode
connected as reflex klystron.( plate=reflector , 1/4
wave transmission line pair connected to gr3 and gr2, gr1 was AM)



<nbucska@pc33peripherals.com> omit 33 Use subj: ENG-TIPS
Plesae read FAQ240-1032

 

[mrkenneth]

Thanks to all for the quick replies!

I still do not quite understand the concept behind how a small change in the VCO frequency can produce a large change in the output frequency of the mixer. If the mixer takes the difference of the two inputs, would not a change in one of the inputs result in the same amount of change in the output of the mixer?

Example: Input 1 is at 1 GHz and input 2 is at 1.5 GHz. I assume the output is 0.5 GHz. When input 2 is increased to 1.6 GHz, then the output would be 0.6 GHz. So a 0.1 GHz change at the input still produces a 0.1 GHz change at the output.

I am surely misunderstanding something. Is the example above correct? I have tried searching on Google for similar applications/circuits (using a mixer, a VCO, and a fixed oscillator) to no avail.

The frequency change does not have to be quick. A second for 100 MHz is no problem. I would prefer to tune the frequency using a DAC in the final application.

nbuska, what do you mean by mechanical tuning? I would prefer everything to be controlled by a microcontroller.

For my application, I would like to generate a sine or square wave of a single frequency, so harmonics should be as low as possible. (Sorry, I do not have a quantatitive requirement, but perhaps -10 dB maximum for the harmonics?)Phase noise or any other artifacts should not be a problem.

By the way, nbucska, was your last post intended for another thread? I did not understand it at all...

Thanks in advance for the help!

---------------------------------------------------------
Operation Radiation:

http://www.mrkenneth.com

 

[biff44]

It is all about percentage bandwidth. It is pretty hard to make a 0.5 to 3 GHz VCO (140%). It is easier to make a 10.5-13 GHz VCO (21%) and mix it down with a fixed 10 GHz signal.

 

[nbucska]

"Zeitghost == spirit of time" . The post was for anyone who may be interested in it.

Mrkenneth:
You could switch the range
Let say F1= 10 GHz, F2=11 to12 gHz -- the difference
Fout=1 to 2 gHz. i.e. a 10% input change causes
100% output change. ( The change in frequency
is the same but the ratio of the change to the
nominal frequency changes with the nominal freqency.



<nbucska@pc33peripherals.com> omit 33 Use subj: ENG-TIPS
Plesae read FAQ240-1032

 

[mrkenneth]

Thank you for the explanation.

Where can I find such a VCO that generates such high frequencies? I have not seen any over 3 or 4 GHz...

Are there other methods? What do commercial signal generators use?

Thanks in advance!

---------------------------------------------------------
Operation Radiation:

http://www.mrkenneth.com

 

[nbucska]

Evidently you don't know how to specify the generator.

Perhaps you could tell us what do you want to use it
and under what condition ( e.g. hobby, amateur building
or repair, manufacturing etc.)


<nbucska@pc33peripherals.com> omit 33 Use subj: ENG-TIPS
Plesae read FAQ240-1032

 

[mrkenneth]

I am just building this for fun, and will probably only produce one or two of these circuits.

I can only make this on double-sided FR-4 printed circuit board, so I am not sure whether I can use frequencies as high as 10 GHz without much loss...

Would using multiple VCOs to cover the entire bandwidth be easier? If so, how would I connect the output of all the VCOs together?

Thanks in advance!

---------------------------------------------------------
Operation Radiation:

http://www.mrkenneth.com