# Design of oscillator

I'm looking for a good tutorial about oscillator design (electronically) for transmitter. The problems with current tutorials are almost all provide only schematics, some explain how it works. But noone show how we determine the transistor (select beta, or frequency range), determine resistor values (bias voltage and output modulation voltage), or capacitor values, etc. It is so hard to experiment and learn by myself about transmitter.

Not to mention about next stage amplifier, how sound added to the signal, and how filter works (I heard for x'tall filter, we should select x'talls with the closest frequency as possible. But why, I don't know. No need to answer this.). Basically, not practical thus cannot learn in depth.

• Hi! Welcome here! Can you link to one of the better ones, so that we might help find something that is similar, but better. Generally, you can put arbitrarily much math and physics to these kinds of problems, and I'm trying to gauge where your interests lie and how deep you want to dive. Sep 5, 2020 at 9:47
• Also, please state in which frequencies you're interested in, and to which accuracy / stability (alternatively, state a purpose). You do something completely differently when you need a 1 ppb-accurate 12 GHz oscillator than what you'd do for a +- 4% 32kHz one. Sep 5, 2020 at 9:54
• You'll have a hard time building an oscillator, and the rest of a transmitter, if you don't first understand transistors. Sep 13, 2020 at 12:55

I would like to applaud your interest not only for homemade transmitters but also a theory! I started to build my own receivers and transmitters not a long time ago and completely understand how frustrating it can be at the beginning.

However this is a broad and complicated topic. If we consider only oscillators for transmitters, there are LC oscillators - Clapp, Colpits, Hartley, and several other, crystal oscillators, various types of VFOs, also PLLs, DDSs, you name it. People write not a single tutorial but books on this subject.

OK I can give you a simple VXO schematic that I used in a CW transmitter not a long time ago:

This is a so-called Super VXO. Q2 is a buffer, Q1 and all the left part is an oscillator. How does it work? OK basically it's a Clapp crystal oscillator, but a crystal is de-Q'ed with a second crystal and also L1 and VC1. What brings a question like "what is Q", "what I need a buffer for", "what are R5 and C3 for", also about oscillator stability, amplification (it's output is only 4 dBm, now what is dBm and how much of it do I need...), filtering, keying, and Barkhausen criterion. If you are interested in this particular schematic here is a tutorial. It's in Russian, but Google Translate should help.

However the goal was to show how a simple schematic creates more and more questions, which create even more questions, and this is why it's impossible just to write a simple tutorial on oscillators.

Here are several books that helped me a lot:

1. ARRL Handbook. Consider it a starting point with a list of your future projects and references to further reading.
2. Hands-On Radio Experiments, vol 1, 2 and 3. Ditto. Vol 1 is especially worth reading.
3. Practical Electronics for Inventors, 4th Edition. Good source of information on various topics, not excluding oscillators and filters. Especially filters.
4. Amateur Radio Transceiver Performance Testing. This one explains in great detail what are IMD, MDS, etc and how to measure them. You will need this.
5. Solid State Design for the Radio Amateur. It's not new, circa 1977. But it explains many things well, mixers especially. The good thing about SSD is that it's available for free.
6. Experimental Methods in RF Design. The newer book by the authors of SSD. It's great but I wouldn't recommend to start from it. Some context is required for better understanding.

My advice would be to read at least (1), (5) and (6), but better - all of them. Start with simple projects. Build an attenuator or a 50 Ohm dummy load. Measure them. You will need both to test your transmitter. Then build a simple LC oscillator. Measure it. Is it stable? Why not? Build a crystal oscillator. Is it better? Is it possible to make the frequency adjustable? How clean is an output signal? Is it possible to clean it up? Build a filter. Is it any good? Did it help?

In other words proceed one little step after another. And soon you will be on the air with your first QRP rig. 73s de R2AUK

• High frequency oscilllator design by Randall Rhea is also a good book for reference Sep 20, 2020 at 17:46
• Did you understand my suggestions? Sep 21, 2020 at 2:56

I find these Electronics Tutorials on Oscillators to be very instructive. LTSpice, the free electronic simulator from Analog Devices, includes example circuits for Clapp, Colpitts and Hartley oscillators. Perturbing the values of the example circuits' elements can help you develop a fuller understanding of how each affects the oscillators' performance parameters.

The good news is you must know that engineers must learn to read and write specifications before they can learn how to design. So don't even try. You must start with good specs !!! (lots of measurable values)

What does it do? [ppm, ppm/'C, Rs, parallel, series or harmonic (overtone) etc ? tolerance?/stability?/aging? sine or square out?

Once you know what is required, your options are "make or buy" on every little circuit. Discrete, IC, module or the final product? What makes radios bad or better?

Fortunately, XO's are cheap now so you don't have to learn a book's worth of info. Just buy a ready-made Crystal Oscillator for not much more than the crystal. But what specs?

X = Crystal or Xtal (x looks like a criss-cross)
XO = Crystal Oscillator
VCXO = Voltage Controlled XO
TCXO = Temp. Compensated XO ( often -40 ~ +70'C or 0 to 70'C) graded by tolerance in ppm, parts per million (error) at room, 25'C and stability = over Vcc & temp. range.
VCTCXO Volt. controlled TCXO (THese can be nulled)
OCXO ( Ovenized SC cut crystal ~1000x more stable and 200x more expensive for 1e-11 type accuracy. The others above are all AT cut X's. Of course, DIY AT-cut OCXO's are cheap if you know-how but that takes a 2 stage thermal design and a 2 stage heater at the 3rd order minimum.
DTXO Digitally tuned XO with PLL and you serially program the string into its non-volatile memory, so one xtal can do many frequencies.

I've probably left some acronyms out, but you'll figure it out.

Digikey has all of these.

## What is a crystal

What are the common attributes must be understood?
https://www.changpuak.ch/electronics/Quartz_Crystal_Filter_Designer_1.php

Q for an X is ~>10k whereas,
Q for an LC resonator based on impedance ratio with R might be 100 to be reasonably stable.

I can remember being an undergrad and reading the ARRL Handbook with glazed eyes as if, I was reading a book on vector economics and then 30 years later asking a grad student from Philly when I was in San Diego, "what is a crystal ?" and he admitted no idea. Yet upon graduation in EE, I had the necessary theory and self-taught skills to understand how to read datasheets and schematics and designed several 2 stage Xtal VLF filters for a Doppler Rx easily in my 1st job.

## What are the Filter Specs?

Again same thing , don't think about design until you find your specs.

What is the best filter spec and is it possible to learn how in a short time? For each stage, RF, IF , Antenna, PS, how many stages, what group delay, BW, Q stop band?