Message	Date	From	Subject
765	2007-05-12 14:53:03	Ed Almos	Switching Time of Phased Lock Loop
If I have a transceiver that uses a PLL for frequency control then it will need to change frequency very quickly. As an example, If I work 'split' on full break in then the PLL needs to change from the receive frequency to the transmit frequency within the time period of a morse dot. A DDS such as the AD9851 will do this very quickly but a PLL will be slower because of the loop filter. How can I measure the amount of time a PLL will take to change from one frequency to another and what sort of switching speeds should I expect to see? Edward Almos HA6SST
766	2007-05-13 15:33:18	Allison Parent	Re: Switching Time of Phased Lock Loop

767	2007-05-13 18:46:33	Andy	Re: Switching Time of Phased Lock Loop
> How can I measure the amount of time a PLL will take to change from > one frequency to another Put scope on VCO control voltage, trigger scope on the digital inputs that change the PLL's frequency. Do it several times, and use various frequency steps, in both directions. The loop is nonlinear, and the response is not necessarily 100% repeatable. You could also look at the phase error (phase detector's output), to see when the phase error (which is related to frequency error) has settled to something small. > what sort of switching speeds should I > expect to see? That could be anything! It depends on the loop's characteristics, which in turn are usually related to the reference frequency (input to phase detector). The latter might be anywhere from Hz to MHz, so the switching speed could be anything over several orders of magnitude. > If I work > 'split' on full break in then the PLL needs to change from the receive > frequency to the transmit frequency within the time period of a morse dot. Instead of one PLL, you could have two and switch between them. Andy

> How can I measure the amount of time a PLL will take to change from
> one frequency to another

Put scope on VCO control voltage, trigger scope on the digital inputs that
change the PLL's frequency.

Do it several times, and use various frequency steps, in both directions.
The loop is nonlinear, and the response is not necessarily 100% repeatable.

You could also look at the phase error (phase detector's output), to see
when the phase error (which is related to frequency error) has settled to
something small.

> what sort of switching speeds should I
> expect to see?

That could be anything! It depends on the loop's characteristics, which in
turn are usually related to the reference frequency (input to phase
detector). The latter might be anywhere from Hz to MHz, so the switching
speed could be anything over several orders of magnitude.

> If I work
> 'split' on full break in then the PLL needs to change from the receive
> frequency to the transmit frequency within the time period of a morse dot.

Instead of one PLL, you could have two and switch between them.

Andy