The expansion bus communication in the SyncInjectors and PDUs is
redundantly powered through their main inputs and the expansion. So yes,
you'd have to cut both the main and the exp. power to cycle them.
I have seen many issues with the SyncPipes over the years. Sometimes
they just go stupid and need a power-cycle. But those were usually the
older revisions. The newer ones with the GTop receivers have been much
more reliable. Don't GPS receivers have sort of a "memory" to reacquire
lock quickly after a power-cycle?
For my issue though, I don't think it's the pipes or injectors. If it's
not a long cable bounce issue, then it has to be software on the radios.
At all of the sites where this is happening, the pipes are tracking 8+
sats, most 11+. At a site where this happens once or twice a week, the
SyncInjector shows only a couple 2D and 3D Fix events and no 1PPS events
in the last four months. And no ignored, early or late pulses.
On 11/10/2015 11:54 AM, Bill Prince wrote:
Since we're talking about sync pipes and sync injectors....
We have a site that is on the 11th story of a 13 story building. The
building is shaped like a big flat V (as seen from above), and 190°
(or so) of the sky is blocked because there is close to 30' of
concrete going above the site (and we are in the shallow part of the V).
The number of satellites used drops to 4 regularly, and 3
occasionally. When the used satellites drops to 3 (or fewer), we will
occasionally lose sync for a second or two. I guess because of the
geometry of the orbiting satellites, this will happen about once or
twice a day; and "about" the same time. As time goes (and satellites
orbit), the actual time of day slowly precesses.
However, we discovered recently when we had to replace the charge
controller, that this once/twice a day loss of sync stopped completely
for a little over 3 weeks after power cycling he entire site.
It's started up again, which leads me to believe that there is
"something" in the sync pipe and/or the sync injector that is "cured"
after power cycling.
We're going to reboot the sync pipe tonight to see what (if any)
effect this has on the issue. But I'm wondering if this doesn't affect
the issue, is there a way to reboot the sync injector? There is an
"Expansion Power" control in Binary I/O, but I don't know if the sync
injector will be power cycled by this because it has a separate power
input. Does anyone (besides Forrest) know if the sync injector logic
is powered by the expansion bus, or is it powered by the separate
power input?
bp
<part15sbs{at}gmail{dot}com>
