[time-nuts] WWVB d-psk-r down conversion method
kb8tq at n1k.org
Tue Sep 23 20:23:35 EDT 2014
If the idea is to decode in software in order to “de-psk” with a switch, then things are indeed easier.
My *guess* is that the loops in most WWVB systems are tolerant of noise bursts already. If not, they would have a hard time with the normal RX environment. There are a number of ways to build a PLL and make that happen, even with long time constants involved. They probably used one or more of the common approaches.
So here’s a plan:
10 MHz / 168 = 59.532xxxx KHz
168 is 8 x 7 x 3. That’s two 4 bit programable divider chips and a divide by 8. With 10 MHz in, it does not have to be very fast. There are (cheap) CPU chips that will do it on a PWM output. If it’s DIP packages, one 74AC161’s should do the job for the programable sections. You would need a NAND gate for each of the programable sections as well. The divide by two’s can be done several ways. I’d put one at the tail end and the other two at the front end.
The 467 Hz note is slow enough to be pretty easy to process with a simple CPU.
On Sep 23, 2014, at 11:54 AM, paul swed <paulswedb at gmail.com> wrote:
> Wow. What 8 hours can do for responses.
> Great comments and appreciated. A piece of the puzzle appears to have
> gotten lost from the start of the process.
> This down converter system. Only detects phase change. It is not intended
> to be a phase locked oscillator system that replicates what the old radios
> Its output flips a signal path to invert or not invert the incoming wwvb
> signal that feeds the old radios.
> There is no need for an instantaneous response in the flip circuit. .3-.5
> seconds seems fine for the old radios. Want to see if I can get down to
> .1-.2 sec.
> Pretty sure from what I have seen so far there is no need to lock the LO
> nor must it be some precise offset of 100 Hz or anything else.
> In watching this on a scope with a stable local 100 Hz reference not locked
> to anything. The phase changes were obvious and useful.
> One of the comments in the thread was that you could use a good local
> reference and adjust for the LO drift. I do believe thats the case. I am
> pretty good at dividing in decades. ;-) But getting ahead of the other
> criteria. Simple as possible.
> OK time to fix the HP3335a that failed. Lost my LO for the moment.
> On Tue, Sep 23, 2014 at 7:37 AM, Bob Camp <kb8tq at n1k.org> wrote:
>> Actually stability isn’t the main problem with the progamables. The issue
>> is usually phase noise ad spurs. Often they use odd multi modulo divides
>> rather than a PLL. That gives them a low cost chip, but the output spectrum
>> is pretty poor. Figuring out if you have a “bad one” or not is tough from
>> many data sheets. It’s rare that cheap XO’s come with phase noise and spur
>> On Sep 23, 2014, at 12:04 AM, Hal Murray <hmurray at megapathdsl.net> wrote:
>>> paulswedb at gmail.com said:
>>>> Did try lots of frequencies and divider math to come up with a simple LO
>>>> scheme for 61 or 59 KHz. Messy.
>>> There are companies that will make a crystal or oscillator at any
>>> you want at a not silly price. Delivery is not overnight.
>>> Beware: There are several companies selling instant delivery of
>>> running at any frequency you want. They are using a programmable PLL so
>>> won't get the stability you expect from a crystal.
>>> These are my opinions. I hate spam.
>>> time-nuts mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to
>>> and follow the instructions there.
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to
>> and follow the instructions there.
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts