The integer greatest common divisor 10MHz/25.576MHz is 16KHz so a simple PLL should go through that frequency.
On Fri, Jul 6, 2012 at 6:38 PM, Bill Dailey <[email protected]> wrote: > Tips on this process if anyone wants to do this. From the cards I have > there are 2 different types of oscillators/resonators... one is SMD with 4 > pads and the other was a through hole crystal. The right sided hole when > reading the label is the one you want.. With the SMD (most of the > oscillators that I have looked at of this caliber). The upper right pad is > the RF out. So... simply remove the oscillator, find a convenient ground > nearby (I used a nearby pad that I verified to be ground) pre-tinned the > shield and center conductor and the hole/pads, applied a little liquid > flux.. touched both with the iron and done. A little hot glue to hold > everything in place.. Bulkhead connector through a hole in the box (for ext > usb) or in a pci cover (for pci cards). > > I searched and searched for a card that would take 10MHz and they aren't to > be found. I didn't want to invest in a word clock that may have oddball > timing issues that I don't understand so I went this route. It seems to > work very well and is a simple modification. The only problem is the > 24.576 MHz.... I have tried to find a board that was plug and play but the > frequency is not a nice one to play well with 10MHz. There are numerous > ways to get 24.576 disciplined but since I am not an electronics guru it > appears the best for me is the valon synthesizer. If someone wants to > build a better mousetrap.. I have a nice VCXO that is begging to be put in > a PLL and would probably give better phase noise performance. I have > considered injection locking it with the synthesizer but can't get my head > around how to get that done. To me that would provide the best of both > worlds although it may introduce some injection-locking phase noise nearby > the carrier (from what I have read). > > I am getting incredible performance and have phase stable audio in spectrum > lab without any calibration which is what I was trying to achieve. Also, > it seems to make the audio better when listening to music although I can't > verify this claim and don't want to start another audiophile thread. > > -- > Doc > > Bill Dailey > KXØO > _______________________________________________ > 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.
