Hi The PLL cocks in MCU’s are abut the worst jitter monsters out there. It’s clear they only make them as good as they need to be for the internal MCU logic. The PLL’s in the FPGA’s I’ve tested are much better jitter wise. Part of the issue is that the FPGA VCO’s will “pull” a few percent. Some of the MCU VCO’s are good over a decade.
Bob > On Jan 20, 2016, at 4:32 PM, Bob Stewart <[email protected]> wrote: > > Hi Poul, > > Your comment about on-chip PLLs got me to wondering about the > dsPIC33FJ128MC804 I'm using in my GPSDO. So, I put it on the 5370A to see > what's up. Input to the gate port is the drive signal for an LED that I > flash at 1PPS (controlled by an on-chip timer). I delay that with a 10ft > piece of RG-58 to drive the start gate. The stop gate is driven by the OCXO > output from the GPSDO. Internally, the chip's PLL is running at 160MHz > (10MHz OCXO / 2 * 32), which is divided down to 80MHz for FOSC and then by 2 > again for FCY, if you know these dsPICs. > > The ADEV at 1s is about 1.7E-10, declining at about a decade per decade. I > take this to represent the jitter of the PIC's onboard RC oscillator as > driven by the OCXO. > > Bob > > > > -------------------------------------------- > On Wed, 1/20/16, Poul-Henning Kamp <[email protected]> wrote: > > Subject: Re: [time-nuts] Generating a solid PPS from 10Mhz source > To: "Discussion of precise time and frequency measurement" > <[email protected]>, "Attila Kinali" <[email protected]> > Date: Wednesday, January 20, 2016, 8:13 AM > > -------- > In message <[email protected]>, > Attila Kinali w > rites: > >> The test results showed a quite more >> detailed structure with few delays over > 100ps and most being between >> 20ps and > 80ps. Interestingly, some were close to 0ps, for which >> we have no explanation good explanation. > > Any on-chip PLL's with > "spread-spectrum" to fudge EMI tests ? > > -- > Poul-Henning Kamp | UNIX > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
