[prev in list] [next in list] [prev in thread] [next in thread] 

List:       soekris-tech
Subject:    [Soekris]  extreme cold deployment
From:       Landon Curt Noll <soekris-mail () asthe ! com>
Date:       2011-09-11 18:23:32
Message-ID: 77C74FB3-A8B2-4B58-8A74-CC6BAD7172EC () asthe ! com
[Download RAW message or body]

Hello Mike,

I do not know if someone answered already … If so here is another one.

Our experience was with equipment deployed at and near the South Pole.
That environment differs from your situation in some ways, but it is similar in others.

The biggest problem you will face is humidity and the second biggest is thermal cycling.

The interior of Antarctica is very very dry.  The typical outdoor humidity was about 5% to 10% max.
Indoor humidity, and where people congregate, is higher.  Our most significant problems
occurred when cold equipment was brought in contact with warmer moist air.  Frost
would form and then as the equipment warmed up (or worse power on in that environment)
the frost would melt and pow!  (Think spraying mist all over powered equipment … not good! :-))

We developed a protocol where cold equipment, powered off, was put into a sealed plastic bag
containing lots a desiccant.  (BTW: When we ran low, dry rice worked reasonably well)  .  The bag
was shaped to minimize the amount of air in trapped in the bag.  The equipment was brought in
and kept in the sealed bad until it was warm to the touch.  Only then would we open the bag.
Yes, frost formed on the outside of the bag, but inside with desiccant (lots of desiccant) it did not.
The frost would melt as the bag warmed. We were careful to dry the outside of the bag before opening.

Before returning to the outside, we would recharge the desiccant.  We would squeeze
the air out of the bags before going outside.  The equipment was placed in a cooling
box that was below freezing but not as cold as outside and where people did not spend
a lot of time breathing moist air.  The equipment would sit and acclimate before going
all the way outside.

The above takes time … a maddening amount of time to some … however
everything takes much longer in Antarctica.  Getting replacement equipment takes months so
spending an hour in these warming / cooling stages is much better than having a failure
due to contact with water.  :-)

Thermal cycling will take a toll on your equipment.  So minimize the the number of time it will
warm and cool.  Leave the unit powered on as much of the time as possible when it is outside and cold.
Don't bring the unit from the below-zero cooling chamber to the outside until you are ready
to it plug it on and power it on.

Use an SSD drive that is compatible with the Soekris instead of a hard drive.  Most SSDs are
better than hard drives in dealing with the cold.

Eliminate unnecessary disk I/O.  Mount your filesystems with the "noatime" option.

Consider placing desiccant inside the box.  Use desiccant that is in those little bags.
Don't put so much as to impede the airflow / vent flow.  There is spare room in our net5501
box so you should be able to attach some bags of desiccant into a corner that is not touching
the electronics nor covering up air holes.

Thermal cycling will be harsh on your plugs and connectors, particularly those next to
warm devices such as drives.  Consider wrapping electrical tape around the device and
connector to pull the connector next to the SSD drive as much as possible.

Consider buying more than one unit in case of a failure.

Good luck!

chongo (Landon Curt Noll) /\oo/\

_______________________________________________
Soekris-tech mailing list
Soekris-tech@lists.soekris.com
http://lists.soekris.com/mailman/listinfo/soekris-tech

[prev in list] [next in list] [prev in thread] [next in thread]