#### Si5351 Frequency Calculation

KI7MWA

I've been playing with a 5351 breakout board & have a question about the way the frequency is calculated, and its limitations.

Setting the PLL frequency is based on a crystal frequency (nominally 25 or 27 MHz) and a (three part) multiplier.  In the simplest case, the multiplier can be in the range (15...90), which, for a 25MHz crystal, would imply a (nominal) range of 375MHz to 2250MHz.

In various discussions, it is noted that the PLL frequency is (should be?) limited to a range of 600-900 MHz, implying a multiplier range of 24-36, given a 25MHz crystal.

I make the assumption that multiplier values outside this 24-36 range are there to enable the use of crystals (or other frequency sources) other than the nominal 25/27MHz.

Or is there something I'm missing?

Shirley Dulcey KE1L

Although the specified range of operation of the Si5351A calls for a PLL frequency between 600 and 900 MHz, the chip will work with PLL frequencies outside that range. People have gotten output frequencies near 300 MHz out of the chip (the specified maximum is 200 MHz), which requires a PLL frequency well above 900 MHz. (The nanoVNA uses that to achieve more frequency range.) Frequencies below 600 MHz are useful to generate quadrature clocks at lower frequencies (the QCX uses that technique on 80 meters) or to extend the minimum frequency lower into the audio range.

When you operate the Si5351A outside its specifications, the results are not guaranteed. Some chips might go to higher frequencies than others, and the supply voltage may matter. (Just like your CPU, using more voltage will probably allow the Si5351A to reach higher frequencies.) Power consumption may rise above the specified limit so heat could become a problem, especially if you are using it at ambient temperature well above normal room temperature. The chip may not meet its jitter specification, which will mean more phase noise on your local oscillator. The waveform of the output has less and less resemblance to the desired square wave output as frequency increases.

Warnings aside, the Si5351A seems to work well enough at frequencies outside its specs to be useful for a number of applications.

On Sun, Oct 11, 2020 at 8:42 AM KI7MWA via groups.io <entilleser=yahoo.com@groups.io> wrote:
I've been playing with a 5351 breakout board & have a question about the way the frequency is calculated, and its limitations.

Setting the PLL frequency is based on a crystal frequency (nominally 25 or 27 MHz) and a (three part) multiplier.  In the simplest case, the multiplier can be in the range (15...90), which, for a 25MHz crystal, would imply a (nominal) range of 375MHz to 2250MHz.

In various discussions, it is noted that the PLL frequency is (should be?) limited to a range of 600-900 MHz, implying a multiplier range of 24-36, given a 25MHz crystal.

I make the assumption that multiplier values outside this 24-36 range are there to enable the use of crystals (or other frequency sources) other than the nominal 25/27MHz.

Or is there something I'm missing?

Ron Carr

If you look at the B and C variants those numbers seem to make sense.  ( and maybe even the A variant but I didn't wade too far into AN619 application note ).   It says an external clock source needs to be in the range of 10 to 40 mhz.

Allan Nelsson

There is a Si5351A tutorial here:

https://www.rfzero.net/tutorials/si5351a/

Not that I know much about the device myself I only use it :-) There is also an online register calculator:

73, Allan.

 1 - 4 of 4