[time-nuts] Can (will) a M12+T generate a negative sawtooth message ?

[EWKehren at aol.com]

The question is what is the definition on time. Yes it will be always 140  
nsec late to what the M12 calls zero. Good for a GPSDO. How ever if you want 
it  to relate to NIST time more hardware is needed unless you can 
compensate that  140 nsec in antenna delay.
A simple solution would be a preset counter that at 10 MHz takes out  
9999999 counts and the last 100 nsec are controlled by the DS chip. Using 100  
MHz and you can use 10 nsec and 0.1 nsec. resolution. Counter is simple to  
implement synchronizer on input and output and a PIC in between, or all 
discrete  logic or a 32 cell Altera G/A.In all cases the counter is hard wired and 
the GPS  receiver controls the DS chip.
Bert Kehren
 
 
In a message dated 5/18/2014 7:59:06 P.M. Eastern Daylight Time,  
tvb at leapsecond.com writes:

Hal,

Yes, there are negative delays. The goal is that the  physical 1PPS output 
is, on average, exactly on-time. If designed right, that  means as many 
negative offset pulses occur as positive offset pulses. The  spread gives you 
the RMS value. 

This is exactly what you want for a  GPS timing receiver.

/tvb (i5s)

> On May 18, 2014, at 1:42  PM, Hal Murray <hmurray at megapathdsl.net> wrote:
> 
>  
> The ref output is the minimal delay through the chip covering the  input 
and 
> output pad buffers.  It will vary slightly with  temperature and voltage.
> 
> There are no negative delays in that  sort of chip.  It's just a bunch of 
> gates/buffers with a  carefully calibrated delay.  (For a negative delay, 
you 
> would  need something like a PLL.)
> 
> If the delay from the M12+T might  be negative, set the antenna cable 
delay to 
> be a bit short and add on  a constant in  software.
_______________________________________________
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.