LMK05028: Hitless Switching

Hello Kudo-san,

KAEDE KUDO said:

1．What is the maximum value of tHITLESS?

We don't have a specified maximum for tHITLESS.  In the ITU-T G.8262 compliance test results for the LMK05318 (snaa316) figure 37 shows the phase transient captured during a switch between a random phase of 25 MHz inputs.

It shows within +/- 0.15 degrees of a 10 MHz signal.  Based on this I calculate (1/10 MHz) * (0.15 / 360) = 41.67 ps.  Or with-in +/- 41.67 ps.

KAEDE KUDO said:

2．About the following D / S description.
　  Does 0ppm mean that there is almost no difference in frequency between the two inputs, that is, the same source?

That's correct.  Only the phase is different.  As described in the application note, when one of the dividers is toggled on/off, the phase of the divider is unknown with respect to the other clock.  Actually, when using a 2500 MHz VCO on LMK04806 to generate 25 MHz outputs.  Because the divide is /100, there are actually 1 of 100 different possible phases distributed at 400 ps steps.

73,
Timothy