Apparently the DALLAS sensor will only last 20 mins on a coin cell. What ia the lifetime of the thermistor alternative? What is the temperature range of that device? How accurate does the voltage need to be if powering externally?
Obviously a 3.3v regulator is one alternative. How about using a pair of AA rechargeables?
Paul

It all depends on how frequent the broadcasts are. For example every 5 mins the average consumption is 12uA, this equates to many many months on a coin cell and an awfully long (in years) time on AA's. The self discharge rate of the batteries will be more a factor.

2v-3.6v, the rest is taken care of in internal code.

We have never tried the upper and lower limits. As the code is provided for free the amount of testing is, as you would expect, somewhat limited. For the £6 to £10 that LLAP devices cost, what we provide in the documentation is at the moment the limit of our information provided. Hopefully someone on the forum can help a little more here.

Sorry, what documentation? I was able to find the doc online for the DALLAS sensor but not the thermistor model nor the action of the firmware for the XRF (that determines duty cycle).

What is the temperature range of that device? If used outdoors one would want to be able to handle temperatures significantly < 0 deg C

Paul

Really?

openmicros.org/index.php/component/search/?searchword=thermistor&ordering=&searchphrase=all&Itemid=484

or

openmicros.org/index.php/articles/87-llap-devices-commands-and-instructions/122-xrf-firmware-temp

It'll go well below zero, here's the current thermistor datasheet

www.vishay.com/docs/29049/ntcle100.pdf

Farnell say -40 to +125, the XRF at -40 might be what starts to drift first

On further investigation it turns out it needs a full 3.3 volts. A pair of AA cells, although fully charged, are insufficient.
Which means that supplying the DALLAS-based kit with a coin cell slot and no instructions is misleading to say the least.
I understand that coin cell operation of the thermistor-based kit is more practical. I don't have one to try out yet.

(BTW. When the voltage is marginal we observed that the temperature reads as 127 degrees. I guess that might be used as a battery-low indication.)

Spencer has removed the coin cell from the shop advert, and has mentioned it needs to be permanently powered.

In the documentation for the dallas it already states "The Dallas firmware is capable of operating as a permanently powered "polled" device with up to 2 Dallas DS18B20 sensors connected. "

As the slip in your order says, all documentation is on line

I have an insane idea. I'd like to replace the 10K thermistor with a 1K version, then put the (very carefully sealed) whole unit INSIDE my -70°C freezer and see if I can monitor temperature. I'd try it on hourly reporting, with the hope that enough chemistry is going on in the battery in that time to load the capacitor for the XRF...
Has anyone been crazy enough to try this?

One issue will be the temp readouts, the internal calculations will be based are the 10k value, not 1k
I know the BVAL of the thermistor can be changed via a LLAP command but I don't think this will give the correction needed to the conversion calculations.
CisecoDev will know better.

Another option would be use the basic Analog firmware, take the raw ADC value and do the mapping to temp yourself.

As for battery I have no idea, give it a try and see what happens.

Matt

I'm going to give it a try! The freezers have a temperature readout on the door, so I'll use that to calibrate. I can simply adjust the raw data to match that number. Next I'm going to see if I can get a low temperature battery...

The sensor as shipped has a 10k series resistor with the thermistor. If you change the thermistor to 1k then you should really change the resistor to 1k as well (otherwise you will lose resolution). The LLAP command axxSRES1000- will set the value of the series resistor to 1k. axxRNOM1000- will set the nominal value of the thermistor to 1k.
Also axxBVAL3900- will set the B Value of the Thermistor for the calculations (just insert the value for the thermistor you want to use).

There is only power across the series resistor and the thermistor whilst a reading is taken so changing to 1k should not significantly affect the battery life.

pbalfe,

My CCB Temperature Sensor went inside my freezer and gave very accurate results at -24c (that was the maximum my freezer can go down to). I kept my 2xAA batteries in the freezer.

-70c is very low and I think the batteries may not work well down to that temperature, so I would extend the wires and place the battery pack outside the freezer!

p.s. Is it a commercial freezer?

Bitter experience tells us that running a wire through the door seal on this type of freezer is a recipe for disaster. Icing up happens really fast if you do that! It is a standard commercial freezer, science labs are full of them (ours is a New Brunswick). Generally they run at -70 to -80°C. I'm going to put the sensor in the freezer in a cell cooler box (1°C / minute chill rate) and monitor it cooling down, I'll post here the temperature at which it stops working.

PB

I'm really looking forward to hearing any results, there's a mining company that operates within the artic circle using our radios and I couldn't give them anything better than the TI datasheet figures of I think (from memory) -40 deg.

If you do any damage through any tests, I'll personally cover a free replacement, feel free to abuse that radio

My expectation is that the xtal will drift first (thus take the frequency out of band), might be worth putting an LED on pin 6 just to see the heartbeat is still running, the reason I mention is we might be able to tweak for this

Miles

Drop one in a vat of liquid nitrogen and see what happens

OK, well I wasn't quite brave enough for liquid nitrogen. But the XRF transmitter chugs along nicely at -80°C!

Apparently Lithium batteries are most capable of continuing to deliver a voltage at this temperature, so I'm using 2 AA batteries to run the XRF (currently after 2 hours at -80°C they're reporting a voltage of 2.7V through the XRF). I've got the XRF waking every 60 seconds to send a reading, which gives the capacitor some time to charge (at -80°C the capacitor has <10% of room temperature performance).

All is not so well with the thermistor itself, once the temperature hit -50°C it started giving some very odd data, first a few "+170" readings, then a few more at around -50, now it's settled firmly at sending -273°C. The true temperature (according to the built in probe on the freezer door) is -78°C.

Tomorrow I'll be able to tell you if it's still running. But meanwhile, any great ideas for replacing the standard CISECO thermistor circuit?