Hi A little math:
You can get an OCXO that will do 1x10^-10 per day and has a sub 5x10^-10 temp performance. It's likely that the temperature impact will be minimal in a normal environment. If you are after 1x10^-8, you need to correct roughly every 100 days. That's not anything you need a cpu to handle. Even if you are after 1x10^-9, the aging is likely predictable. A bump every 10 days and a measurement once a month likely does the trick. Still not much of a need for a CPU. For a much easier / cheaper / less work solution - grab a cheap rubidium. Manually set it to ~ 1x10^-11. It will hold your 9 digits for at least a year. With some luck I'll hold 10 digits. Bob On Dec 8, 2012, at 1:59 PM, Chris Albertson <[email protected]> wrote: > On Sat, Dec 8, 2012 at 9:30 AM, <[email protected]> wrote: > >> Hello all - >> >> "People talk about good deals on Thunderbolts but I have >> yet to see one. It seems peak Thunderbolt passed before I was >> seriously looking." >> >> >> *The box you put it in - suspending it above the bench on inspirational >> thought probably >> will not work. A nice box is easily $ 50.00 so who cares about adding >> another flip flop or >> counter at 25 cents each. > > > > The goal, well my goal is to build a GPSDO to this set of requirements > > 1) well under 1/2 the cost of the t-bolt. > 2) can be made with common parts and skills most people have > 3) is completely modifiable (open source software) > > So #1 above means you use a low cost box, perhaps the case and old CDROM > was once housed in or maybe you gut and old PC chassis > > #2 pretty much means you have to keep theparts cound way down. Yes a chip > might cost only $1 but as soon as you get more then about three of those 25 > cent cips you need a PCB. PCB making is not a common everyday skill so #2 > means a way-low parts count > > #3 means the uP is gong to have to come with a very easy to use and free > open source toolchain. Adruino is like that, there are others, but a bare > uP chip is not going to work > > > You are right that one way to engineer a product is to first set > performance numbers. But another more common way is to first identify a > target customer and then set a target price. In this case the target > customer is a hobbyist with only basic skills who wants a low cost easy to > build GPSDO that has can understand and modify himself. > > Next with those very hard to meet requirements (low cost, low parts count, > simple to build) we can ask if the exected perfomance is "good enough" > What is good enough. I'd say if you could use the GPSDO a a local > frequency standard for a counter that has 8 digits you are doing "well > enough". If you can get to 10 digital it is pretty good > > -- > > Chris Albertson > Redondo Beach, California > _______________________________________________ > 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.
