Accurate Frequency generation
Accurate Frequency generation
We use a crystal in one of our pc boards that the user can replace with different values to get different carrier frequencies. The crystals on both ends of the communication devices have to match. I would like to build in a programmable frequency generator, but it needs to be accurate like a crystal. Any ideas? We currently use 6.3, 6.8, 7.37, 7.8, 8.7, 9.2 and 9.5mhz crystals.
Take your set of frequencies. Express them as fractions (i.e. 6.3/1, 6.8/1 etc). now find the least common denominator of that set of fractions. This is your fundimental frequency. This frequency can be divided into your desired frequencies using clock divider chips and counters. The numerator in your new fractions (with the common denominator) is the factor you need to divide the clock by (hopefully you can work it out to have whole integers). All you need is one stable crystal at or a multiple of your fundimental freq.
Sorry, I'm too lazy to do the math for you. Hopefully the fundimental frequency you get is a commonly available crystal otherwise, you'll have to pay to get custom crystals tuned to what you need.
I do believe there are other ways that do not use crystals but a PLL circuit instead. Look into how its done for FM radio. Search on "pll synthesized chip" and you will find some VCO clock chips of higher frequency than you want that can probably be divided down to what you need. Tuning with a VCO is less precise so you'll need fine tune adjustment or a way to lock in on the carrier.
Sorry, I'm too lazy to do the math for you. Hopefully the fundimental frequency you get is a commonly available crystal otherwise, you'll have to pay to get custom crystals tuned to what you need.
I do believe there are other ways that do not use crystals but a PLL circuit instead. Look into how its done for FM radio. Search on "pll synthesized chip" and you will find some VCO clock chips of higher frequency than you want that can probably be divided down to what you need. Tuning with a VCO is less precise so you'll need fine tune adjustment or a way to lock in on the carrier.
-
- Posts: 2277
- Joined: Wed Nov 24, 2004 1:01 am
- Location: ASHTABULA,OHIO
- Contact:
I have to agree with Jwax.Swithed crystals through a diode switching network are simple, cheap and easy. If your system is even semi broad band, you don't even need individual trimmers, thereby taking up even less board real estate. PLLs can get involved unless your lucky enough that all the math falls in the right place for your project. BTW, what kind of freq. accuracy are you looking for?
Re: Accurate Frequency generation
Beng, are those frequencies exact or approsimate? Could you please post the frequencies with all of the decimal points?ljbeng wrote: We currently use 6.3, 6.8, 7.37, 7.8, 8.7, 9.2 and 9.5mhz crystals.
Just curious, are these satellite audio subcarrier frequencies?
Bob
- Dave Dixon
- Posts: 436
- Joined: Thu May 01, 2003 1:01 am
- Location: Wichita, KS
- Contact:
Who is online
Users browsing this forum: No registered users and 145 guests