Adafruit is open, safely ... and shipping all orders at this time! Read more!

SI5351 and FM (Arduino or other platform)
Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.

SI5351 and FM (Arduino or other platform)

by LU8DAM on Thu Oct 26, 2017 8:58 am

Greetings to all the forum. I would like to know if anyone has experienced frequency modulation (FM) with SI5351. I need to read about it. I have an FM VHF transmitter to which I need to replace a 48 Mhz (48 x 3 = 144 Mhz) crystal with an SI5351. Without modulation the SI5351 works perfectly with the transmitter.
Best regards.

Posts: 1
Joined: Thu Oct 26, 2017 8:49 am

Re: SI5351 and FM (Arduino or other platform)

by tjrob on Thu Jan 03, 2019 12:10 am

I just built a small clock generator box using a Raspberry Pi Zero W and the Adafruit Si5351 board. Works great! Details below. The output is as stable as a crystal oscillator, with only slow drift of a few parts per million per hour.

If you are thinking of modulating the Si5351 at audio frequencies, I doubt that will work, as the only control is via I2C and its bandwidth isn't high enough. Note the Adafruit board uses the Si5351A, but another version of the chip, Si5351B, is specifically designed to be frequency modulated and has an on-chip VXO.

My details:
I built this box specifically to test a frequency counter I am building with a LC4256 CPLD and a Raspberry Pi 3B+. It uses a 10 MHz crystal oscillator as a timebase. The CPLD has a 20-bit counter that is latched and then read at 1 kHz by the RPi; each sample is converted to frequency and written to a file. The CPLD analysis says it should work up to 215 MHz, and tests show it does work at 150 MHz (the highest frequency I can get the Si5351 to generate).

I have run overnight with the Si5351 set to 150 MHz (see note), and the output file has a repeating pattern: a series of entries 149997000 then one 149998000. That pattern is just showing that the crystals differ by about 20 parts per million (easily within their accuracy specs). I saw no frequency variations other than a very slow change in the number of 149997000 entries -- I attribute that to slightly different drifts of the two crystals.

Note: see my warning about electromagnetic interference here:

Posts: 3
Joined: Wed Jan 02, 2019 10:13 pm

Please be positive and constructive with your questions and comments.