Quartz crystal as basic frequency reference

[sahko123]

I want to make a very simple frequency reference with a quartz crystal and somewhat obviously don't have the budget to buy pre-built standard references. All i want to know is how to make an oscillator using a crystal and how accurate should i expect it to be.

[ArthurDent]

From what you are saying about your knowledge you will be far better off to buy a OCXO than try to build your own oscillator. Making a quality frequency standard from scratch involves a lot of technology an a fair bit of magic. Not knowing what you are using the standard for prevents any good recommendations. Even if you build one you have to have a higher accuracy standard to adjust yours to the correct frequency.

Here is a really simple inexpensive one.
https://arachnoid.com/frequency_standard/

[sahko123]

sorry guess i should of specified the use case but the stuff im gonna be referencing it against is only gonna be used down just above the audio rang like 30khz and under and itll be used as a transfer reference to go from one scope to another. And in the end im just looking to make a small oscillator circuit thats stable enough over like a minute

[Conrad Hoffman]

Look for a DIP crystal oscillator. Should only cost a few dollars. Typically you just supply power and it will have TTL or similar output.

[james_s]

I bought a couple of surplus Trimble double oven precision 10MHz oscillators for $15 each, they are many orders of magnitude more stable than anything you can easily build yourself. If you want just modest stability pull a TTL oscillator out of some scrapped device, or buy one of the $12 DDS boards from eBay.

[Etesla]

If you're set on using a crystal, you can get most crystals going with a CD4060, which outputs pulses at 1/8, 1/16, 1/32... of that frequency. There are plenty of other chips that do that same job, but I haven't looked into it.

You could also just buy an oscillator with your desired output type off digikey for a buck or two (https://www.digikey.com/products/en/crystals-oscillators-resonators/oscillators/172)

[tanveerriaz]

old mobile phone hase tcxo 13 and 12.8 Mhz use divider cir. for down frequency.

[Doctorandus_P]

With "normal" crystals you can expect an accuracy of about 200ppm. A part of the drift will be because of temperature, part of the drift will be because of imperfections in the crystal itself.

The pierce oscillator is the simplest to build, and indeed with a CD4060 which also has a divider you get a bunch of output frequencies.

Wikipedia has a nice article about crystals, including how they look inside, and an overview of the different pre-built oscillators such as TXCO
https://en.wikipedia.org/wiki/Crystal_oscillator

[drussell]

For a simple, portable frequency reference at relatively low frequencies, consider simply amplifying/dividing/multiplying and zero-beating a (preferably with WWV or CHU or whatever, but really any) radio station carrier, since those should be very closely controlled these days.

That's how the olde timey people calibrated their 100 kHz "reference" "calibrators" for use in the shack.

Still works. 

[james_s]

Ideally you'd use that technique to calibrate a nice stable oscillator though rather than having to do it every time you want to calibrate something.

10MHz is the standard used for most everything that can use an external reference so that's probably what I'd aim for unless I had other specific needs.

[drussell]

Ideally you'd use that technique to calibrate a nice stable oscillator though rather than having to do it every time you want to calibrate something.

10MHz is the standard used for most everything that can use an external reference so that's probably what I'd aim for unless I had other specific needs.

True.

In this case, though, I thought to start the OP was just wanting to compare a reasonably stable signal across multiple oscilloscopes.