http://talk.jeelabs.net/tags/cold Discussion about Jee Labs projects en-US Sat, 11 Sep 2010 02:46:03 +0000 http://bbpress.org/?v=1.0.2 q http://talk.jeelabs.net/search.php http://talk.jeelabs.net/topic/95#post-405 Tue, 29 Dec 2009 20:36:53 +0000 amvv 405@http://talk.jeelabs.net/ <p>Hello!</p> <p>I have been trying to use a JeeNode for a homemade weather station! It is now in beta testing. It has been outside for more than 2 weeks now!</p> <p>Indeed we don&#39;t even get -10C that often here, with that already being extreme.</p> <p>I had problems with the power saving stuff, as it had been discussed until a few months ago (didn&#39;t try the recent developments), which after pulling lots of hair, I attributed to the cold outside - The jeenode would hang after being put outside...that didn&#39;t happen if there was no power saving active, I think it had something to do indeed with the oscilators and watchdog timers...and it was even at not that low temperatures - about 10C.</p> <p>I will try to report more on that one of these days. Now I only have the Arduino power saving and watchdog timers, and I am not using the radio timer as suggested.</p> <p>But anyway, It has worked fine running from 3 AA batteries down to -5C at least! This is probably not very useful, since -5C is very far from -50C!</p> <p>I will report on the station once it is completed! </p> http://talk.jeelabs.net/topic/95#post-404 Tue, 29 Dec 2009 19:17:51 +0000 olliver 404@http://talk.jeelabs.net/ <p>An interesting topic!</p> <p>What does AVR specs say about temperature? I&#39;ve the feeling that the operating temperature of standard chips is usually above zero Celsius and some special design can run below it (like those used in cars etc.). </p> <p>I think the standard solution is to have some kind of heated (and isolated) cabin or case for the electronic parts. Sensors can (must) reside outside of this artificial atmosphere. Heating of course takes more energy and should also be controllable to prevent overheating (thermostat). So I guess, battery powered solution is out of question. Batteries are drained in cold temperature very fast anyway. </p> http://talk.jeelabs.net/topic/95#post-395 Mon, 28 Dec 2009 20:14:55 +0000 Jean-Claude 395@http://talk.jeelabs.net/ <p>Just saw that the RFM12B operating temp range is &quot;only&quot; down to -40C (storage down to -55C). </p> http://talk.jeelabs.net/topic/95#post-394 Mon, 28 Dec 2009 18:53:12 +0000 Jean-Claude 394@http://talk.jeelabs.net/ <p>I have no experience whatsoever with such low temperatures, -10C is rare here, and -20C exceptional.</p> <p>If you&#39;re concerned about this you could change the fuses so the ATmega runs at off its 8 MHz internal oscillator. This would get rid of the resonator issues, and brings the ATmega fully in spec - something which might be important at extreme temperatures, see http://news.jeelabs.org/2009/02/27/out-of-spec/</p> <p>Haven&#39;t used JeeNodes outside yet (I want to get all the low-power battery / solar issues resolved first), but I&#39;d be very interested to hear how it holds out in &quot;real&quot; weather conditions.</p> <p>Since you have AC power - you could also put the whole thing in a big mitten along with a resistor, heating it up ever so slightly... </p> http://talk.jeelabs.net/topic/95#post-393 Mon, 28 Dec 2009 16:56:19 +0000 dkb 393@http://talk.jeelabs.net/ <p>I would like to use a JeeNode outside for basic weather measurements but I am not sure how well it will do in cold temperatures. It frequently drops to -40C in the winter here and has been as low as -51C within the past 15 years. The resonator for instance is only rated to -20C and we&#39;ve been below that already. I will however have AC power available so a battery will not be an issue. Will there be issues in this environment? </p> http://talk.jeelabs.net/topic/92#post-379 Sat, 19 Dec 2009 20:18:54 +0000 Jean-Claude 379@http://talk.jeelabs.net/ <p>FEC as in forward error correction? Not sure that will help much - radio waves at these power levels are finicky. There will always be periods when nothing gets through at all. I sometimes see my weather station on the roof drop out for up to an hour - presumably it&#39;s at the limit of the range, but still...</p> <p>I would expect to get more effect by making the code a bit smarter - instead of staying in high-power mode until the final ack is received, the node should really go to sleep in between as well. Then that last remaining large current drain period would be reduced from seconds to milliseconds.</p> <p>This definitely needs to be addressed. I&#39;m seeing similar behavior for some new nodes here as well - there&#39;s a lot of reinforced concrete in the house, range is extremely variable. </p> http://talk.jeelabs.net/topic/92#post-378 Sat, 19 Dec 2009 18:16:53 +0000 charlie 378@http://talk.jeelabs.net/ <p>Hello,</p> <p>Concerning the power consumption I was thinking retransmission is not always the best option. In my case the central node was switched of after &quot;an evening playing with jeenodes&quot; while the temperature probe was still outside measuring the freezing cold -9°C.</p> <p>Due to the fact that the central node stopped ack-ing, the temperature probe kept on re-transmitting, which ultimately had the batteries running on empty by the morning...</p> <p>Ok, my batteries are not of that quality and were probably not fully charged, and I have no clue how NiMH batteries behave at -9°C, but nevertheless it captured my mind today...</p> <p>Anyhow,<br /> jcw, have you ever considered (adding) FEC to make the transmission more reliable iso re-transmitting (or a combination).</p> <p>Regards,</p> <p>Charlie </p>