<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
I'd like to thank everyone who showed up yesterday for the SnoDEM
repair + upgrade work. It was great to see such a big turnout!<br>
<br>
The ground crew did a great job, and kept both the climbers busy
throughout the day with no stalls of work up on the tower! The
pre-work before deploy day (cutting CAT5 to length, pre-terminating
ends, color-coding the cables, partially assembling the shields,
pre-assembing the Lookout dish) saved so much time on the day. So a
BIG thanks to the few who did this extra task to help make the day a
success!<br>
<br>
Here's what we managed to do:<br>
<br>
<ol>
<li>Upgraded all sectors to MIMO units, which roughly doubles
their data rate.</li>
<li>Upgraded all sectors with shields, to minimize crosstalk and
maximize throughput.</li>
<li>Installed the shielded + MIMO Lookout Mtn dish for that
upcoming site link.</li>
<li>Upgraded the Haystack dish with a shield to minimize
crosstalk.</li>
<li>Replaced ALL the HamWAN cabling on the tower, and secured it
with proper tower cushions / hangers. No more zip ties that
fail!<br>
</li>
</ol>
<br>
We didn't manage to hit 100% of our goals for the day however.
Here's what remains to be done:<br>
<br>
<ol>
<li>Gold dish still has a broken modem, and needs a shield.</li>
<li>ETiger dish still needs a shield.</li>
<li>"Bruce" dish (no current link, but maybe later this summer)
still needs a shield AND MIMO modem upgrade.</li>
<li>Cable ladder is missing several cable cushion tie points that
need to be installed.<br>
</li>
</ol>
<br>
My estimate for the speed of cable installation with proper cushion
hardware was way too low. This had a few root causes:<br>
<br>
<ol>
<li>I missed the presence of cross-arms all around the tower that
prevented direct access to the cable ladder by newly hoisted
cables.</li>
<li>The general strategy probably had some flaws:</li>
<ol>
<li>It called for the hauling of CAT5 in 4-cable bundles to each
of 2 heights on the tower. I figured this would save time by
having fewer hauling trips up the tower. It turns out the
ground crew was really fast at hauling, and that I was really
slow at anchoring and routing the cable, and even slower when
I realized I had pulled way too much excess cable up and had
to push everything backwards to achieve proper lengths. It
probably would have been faster to haul up 1 cable at a time,
set it to the correct length for its destination modem, and
install the cushions on it.</li>
<li>I split the 2 climbers into 2 teams so each could work on
independent stuff concurrently at different heights on the
tower. While this made some sense, I suspect a greater speed
boost would have been achieved had the climbers been on the
same level most of the time. I spent a bunch of time in tying
things off just to move around, whereas a 2nd set of hands
could have easily held something instead. There were plenty
of concurrency opportunities at each level, where each climber
could have taken on the installation of each modem, or
splitting work up between cable routing and dish replacement,
etc.<br>
</li>
</ol>
</ol>
<br>
On the plus side, the time estimates for the ease of sector + dish
replacement / installation were pretty spot on.<br>
<br>
If you have any further insights into what went well or poorly,
let's discuss it on the deploy@ mailing list. I only included psdr@
as CC to let people know the status of SnoDEM, so please remove it
from your replies.<br>
<br>
I'll be reaching out to our climbers to see when we might schedule
the follow-up visit to finish the work @ SnoDEM.<br>
<br>
And again, big thanks to everyone who helped! Couldn't be done
without you.<br>
<br>
--Bart<br>
<br>
</body>
</html>