[LinuxPPS] Why is my ATOM clock a falseticker?

[Paul Simons]

I choose the clear edge based on this quote from the wiki:

# "PPS pulses are usually short, and the leading edge is the on-time mark,
# so by looking at the time of two adjacent edges with cat /sys/class/pps/pps1/{assert,clear}, 
# you can see which one leads the other."

And, based on this diagnostic,:

PPS event on source 1 at 1233330678.493769677
capture clear seq #76087 for source 1
[IRQev] PPS clear at 615823591 on source #1
PPS event on source 1 at 1233330678.993769372
capture assert seq #76087 for source 1
[IRQev] PPS assert at 615823716 on source #1

the clear edge leads.  I can state that the scope trace was in the positive direction, and it looks to me like the clear edge leads.  I do not know if that makes sense.  Judging from your response, my thought that a falseticker is bad is correct.  I added the atom clock to an NTP configuration that was already working and that included five peers.  I saw that the atom clock was becoming a falseticker, so I stripped out the peers, deleted the drift file, and let ntpd run overnight.  I then reported the results to this list.

On Friday 30 January 2009 02:46:39 pm Paul Simons wrote:
Oops, here is the relevant section:

server 127.127.22.1 minpoll 4 maxpoll 4 # ATOM (PPS)
fudge 127.127.22.1 flag2 1              # Use the "clear" edge of the
signal

I think this may be an issue.  In your original note you wrote "...declared it 
a decent RS232 signal (0/+5v, 200ms)" but wouldn't that mean that you need to 
being using assert edge and not the clear edge?  I know that my OnCore GPS TTL 
PPS signal that I have hooked up to my serial port goes high, which is +5V, 
when it pulses and I use the assert edge.   But I am using a different driver 
so it could be different.

Hal 


----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ml.enneenne.com/pipermail/linuxpps/attachments/20090131/73b5fc95/attachment.htm