Designing a 2.455 GHz Oscillator

[hashmos]

Hi All,

I would really like your help in my design project, which is about desgining, as mentioned in the title, an oscillator with 2.455 GHz frequencey.

I want it to be efficent to the greatest extent, meaning the lowest power consumption possible and greatest stablitiy.

I'm confused on which oscillator to use, is it the colpitts, or a crystal with a frequencey multiplier.

I also need resources to help, anything from ebooks to tutorials on simulations.

I would be greatful if you help me with your expertise.

thanks

 

[negitron]

Not to sound too negative, but gigahertz-range oscillator design is VERY different from the sorts of low-frequency applications in which you find oscillators built from discrete components. 2.455 GHz is almost exactly the frequency used by consumer microwave ovens, for example, and they use a magnetron tube to generate the RF. Cordless devices using the 2.5 GHz band nearly always use a monolithic (built on a single chip) purpose-designed oscillator such as the MAX2753. Since these are intended for battery-powered applications, they already have high efficiency and low power consumption.

 

[hashmos]

thanks negitron, I examined the mangetron but it is not practical to use it in my application, nor is it acceptable to use the ready chip since there will be no design invloved in the process. In other words its to easy. Still looking for advice.

 

[negitron]

What exactly are you attempting to do? Is this a class project of some sort?

 

[hashmos]

This is part of my design project for graduation, obviosuley I am majoring in electrcial engineering, and our advisors stressed the fact that buying chips and just petting them together is not considered designing, they want everything to be desinged and put togther by us. So...

 

[negitron]

In that case, you'll need to show what effort you've put into this point before I'll offer any further guidance. Hope you understand.

 

[hashmos]

This is the thing, dear negitron, I still have nothing solid yet. I'am familiar with the low frequencey oscillators and the LC tank feedback system, but I didnt find a type that can assembled from discrete components. I read about the colpitts but the image is still vauge. This is so far the effort I have put into this project.

 

[waht]

When designing a free running microwave oscillator, hopefully you are familiar with S-parameters, transmission lines, microstrips resonators, smith chart, Rollet's stability factor, and stability circles?

Frequency multiplier chain is so much easier to design. Start with a low frequency colpitts oscillator you are familiar with, and apply successive saturated amplifier stages to double or triple the frequency, and filter out.

 

[MATLABdude]

This is part of my design project for graduation, obviosuley I am majoring in electrcial engineering, and our advisors stressed the fact that buying chips and just petting them together is not considered designing, they want everything to be desinged and put togther by us. So...

Maybe I'm missing something here, but isn't this just a small part of the something that you're designing? The design and engineering aspects would come in
a) designing the rest of the SYSTEM, and getting everything to work together, and
b) designing the firmware to operate all these chips (if applicable)

When I was a TA for the senior EE design project at my school, people were expected to buy chips and modules that carried out various functions, but they were also supposed to integrate it into the rest of their systems. (We had a similar restriction against cookie-cutter engineering). So the students would be able to use the chip in, say, a microclimate measurement device that had wireless transmission / reception capabilities, but, if they were designing a wireless presentation mouse, they wouldn't just be able to head down to FutureShop and buy one. (As ridiculous as this sounds, supposedly someone tried to do just that, though it may also have been buying one of those DIY electronics kits, and just building that and reverse engineering for documentation).

You really don't have the experience or knowledge necessary to design / make something like that yourself, but I also doubt that is the objective of the course. You should talk to the professor / lab instructor for the course and figure out whether it's permissible to use that product in whatever it is that you're building.

 

[hashmos]

Matlab Dude, thanks for replying, you are absolutley correct. I lack the knowledge and experience, and that is why I came here for help. If you can provide me with resources and links I would be grateful. Coming to your point about using ready chips, you make a good point, and if there is any ready made circuits, I would also be grateful if you provide some links and suggestions.

Thanks all

 

[Pumblechook]

You need to specify what power and how stable.

Very stable oscillators - which remain within a matter of Hz for hours on end need to be phased locked to a stable low frequency Xtal osc, possibly in an oven OR you need a multiplier chain starting with an overtone Xtal osc in the VHF region.

Simple free running cavity osc will drift maybe 10s or 100s of KHz in a short space of time.

 

[Pumblechook]

You might find circuits for oscillators used in downconverters and transverters for the 13cm amateur band such as designs by Charles Suckling, G3WDG.

 

[vk6kro]

You have probably covered this topic in your course work and you should follow the techniques you were given.

2.4 GHz is way above the frequencies where you could use coils and capacitors as tuned circuits. A quarter wave transmission line is about 3 cm long at this frequency and this may take the form of a pair of tracks on a printed circuit board.

There are programs for designing such circuits, but using them may not be in the spirit of your assignment.

Free running oscillators at this frequency are very unstable and it would be normal to divide down from them to some stable oscillator's frequency and then do a phase comparison. This produces a DC voltage used to pull the oscillator onto frequency.

 

[hashmos]

Thanks guys for your support, I really appreciate it. If there is any links or resources that can help me, please post them.

 

[Pumblechook]

http://www.g3wdg.free-online.co.uk/oscillat.htm

http://www.g3wdg.free-online.co.uk/fig1.jpg

http://millimeterwave.free.fr/G4DDK004.pdf

 

[hashmos]

Thank you Pumblechook, any more links guys ??

 

[jason12345]

Thank you Pumblechook, any more links guys ??

Try this link:

Design a Direct 6-GHz Local Oscillator with a New, Wideband, Integer-N, PLL Synthesizer
http://www.analog.com/library/analogDialogue/archives/35-06/ADF4106/index.html

You need to do some research into the various methods of designing oscillators and Google scholar is your friend:

http://scholar.google.co.uk/

Your supervisors want to know that you have looked at the current literature on the subject and have made an assessement of what the best design procedure to follow is: Improve on existing designs, or just copy them? Either way *DON'T* reinvent the wheel or try to be too creative because you haven't the time. The project is there to see if you can think and organise your time to achieve your goal.

So read a general book on designing Ghz oscillators, then look at some cutting edge papers and their references, then make a decision on the design route you will take, bearing in mind you have a few weeks/months to design it, rather than years of investigation. Your initial research into the project will be marked.

Regars,

Larry