We hit –69 here in Utah.  (and +116 too, 185 degree spread)  
http://www.deseretnews.com/article/116318/PETER-SINKS-UTAHS-COLDEST-SPOT.html?pg=all

From: Bill Prince 
Sent: Tuesday, May 19, 2015 7:55 AM
To: [email protected] 
Subject: Re: [AFMUG] Enclosure low temperature

-40 is almost 100 degrees lower than we ever see inside any of our cabinets. 
Just checking around our network, about the lowest I see anytime in the last 
year has been ~~ 50 degrees.


bp
<part15sbs{at}gmail{dot}com>

On 5/19/2015 6:50 AM, Chuck McCown wrote:

  I would think the worst case scenario is being cold soaked.  
  Power is off for a week.  Power comes back on.  -40 outside, the chip will be 
–40.

  Pour some liquid butane on it to warm it up...

  From: Bill Prince 
  Sent: Tuesday, May 19, 2015 7:47 AM
  To: [email protected] 
  Subject: Re: [AFMUG] Enclosure low temperature

  How about actual data? For the last week, the ambient outdoor temp at this 
site has been between 50 (night time), and 60 (daytime). The SiteMonitor temp 
inside this 18x16x10 enclosure has been like below. So for at least this 
situation, it's been running between 10 and 20 degrees above ambient. The 
daytime temp is probably higher because of sun exposure.



bp
<part15sbs{at}gmail{dot}com>

On 5/19/2015 6:34 AM, Chuck McCown wrote:

    Back to the original question, need more data.  It is a question of 
physics.  Jaime touched on it.

    Live load.  
    Ambient temp.
    Thermal conductivity of the enclosure.
    Surface area of the enclosure.
    Wind speed at coldest temp.
    Thermal convection currents inside the enclosure.
    Thermal conduction currents inside the enclosure.

    Not trivial.  

    Perhaps use some transistor grease and place a rectangular type metal plate 
resistor on the memory.  Add a tiny bit of foil and insulation.  It would take 
so little energy to get the memory up into the safe region that it could 
probably happen in a few seconds.  As has been mentioned, the pre-heat on boot 
would certainly do that.  Once booted, measure the temperature and heat or not 
heat.  Be a fun experiment.  The memory thermal mass is so small the resistor 
probably would have more thermal mass.  You could get a 50 degree rise in maybe 
5 seconds with 1/4 watt if insulated properly.  




    From: Forrest Christian (List Account) 
    Sent: Monday, May 18, 2015 8:09 PM
    To: af 
    Subject: [AFMUG] Enclosure low temperature

    What is the lowest temperature that each of you would normally expect to 
see in your enclosures?

    The reason I'm asking is that I'm in the process of developing up a few new 
products.


    To date, all of the packetflux products are designed with components rated 
-40 to +85C (I.E way cold to way hot).    I'd like to retain this rating, but 
I'm running into a minor snag:


    For storage, I'm planning on integrating a SD card (probably microSD) in a 
socket.   I only need a GB or so, and SD card memory is inexpensive at that 
range.


    Unfortunately, all the reasonably priced SD cards are only rated down to 
about -25C or -13F. which are about ~$3 in qty.   "Industrial temperature 
range" ones which are good down to -40C/F are available but they add at least 
$30 to the cost for non-name brand, and even more for known brands.  When 
you're talking about a $100 end-user price, a $30 1GB SD card seems excessive - 
and probably isn't even possible if I want to meet the $100 price with some 
margin.


    So, I'm currently playing the 'what options do I have' game.   I hate to 
ship a product only rated down to -25C/-13F, but I know for at least a large 
chunk of my customer base they never see below this temperature, especially 
when you add a watt of power dissipation in the case with the device.  Which 
leads me back to my original question:  What's the lowest temperature most 
people would expect to see inside their enclosures.


    -- 

          Forrest Christian CEO, PacketFlux Technologies, Inc.

          Tel: 406-449-3345 | Address: 3577 Countryside Road, Helena, MT 59602
          [email protected] | http://www.packetflux.com

             






Reply via email to