[time-nuts] 50:50 duty from a PRS10

Magnus Danielson magnus at rubidium.dyndns.org
Fri Jul 22 09:01:43 UTC 2011


On 07/22/2011 05:38 AM, Dave M wrote:
>
>> Hello,
>>
>> I have a PRS10 rubidium which gives a 1pps output. The output is a
>> 10 us positive pulse.
>>
>> I need to convert that to a 50:50 duty cycle pulse. Still 1pps.
>>
>> I'm hoping for a simple circuit rather than having to use a pic.
>>
>> I don't mind a bit of propagation delay, but I need to preserve the
>> low jitter on the rising edge, and hopefully also the falling edge.
>>
>> Any ideas out there
>>
>> Regards
>>
>> Steve
>
>
> The simplest way that I can think of is to feed both inputs of an XOR
> gate with your 1pps signal to double it to 2pps, then feed the XOR
> output into a flipflop to get back to 1pps with 50% duty cycle. You
> might have to delay the 1pps into one of the XOR inputs with a small RC
> to reliably trigger the flipflop. Use 74HC or AC gates and the jitter
> should be tolerable.

The rise and fall transitions which trigger the XOR gate will cause the 
XOR output to occur at those times and not with a 500 ms spread as you 
wish them to be.

You really need some form of memory that recalls that we have a "high" 
period for 500 ms rather than the 10 us. The method (and inherent 
precision) will vary, but there is no alternative to this memory. 
Dividers from the 10 MHz ensures it is a synchronous timed transition. 
If the fall transition isn't critical, the 555 solution or similar will do.

Cheers,
Magnus



More information about the time-nuts mailing list