Hi,
I’m looking for a node to monitor temperatures in ultra low temperature freezers (- 80°C) using TTN.
This is exactly what I need but the lowest temperature is too high (Range: -40 to + 80°C) : https://www.dragino.com/products/lora-lorawan-end-node/item/169-lsn50v2-s31.html
The price of this model is around 50 €, I’d like to find something around this price (or lower of course…).
Can someone help me to find the right node ?
This doesn’t really sound like a LoRaWAN application, likely you want things like local datalogging which record if there’s ever been a temperature excursion (particularly in a medical or research integrity setting), and reporting through a building network or more likely a traditional mobile network.
LoRaWAN makes sense when you have a whole lot of relatively low value assets distributed through an area and a “nice to know but not essential” data requirement, it’s not a good choice for a situation where you have a small, sparse collection of very high value assets with critical needs.
In terms of the sensing technology itself, try a web search on things like “cryogenic temperature sensor” and see what Omega etc have to offer.
You and everyone else that wants to provide in-transit & storage quality control for mRNA BNT162b2.
Common silicon like TMP36GZ will work down to -50℃ and could form the basis of a node at around €15 in bulk. After that, it’s more specialised sensors with more sophisticated input circuitry.
Given the critical nature of such an application, is TTN really the best platform for data delivery?
Test/Dev on TTN then move to TTI or other pro/supported service for production 
There are many hospitals, research institutes and other related sites enabled with LoRaWAN networks for just such sensitive uses and protect/monitor/audit of temperature sensitive assets…including medical supplies. Outside of fixed sites there are similarly many logistics implementations now that monitor/alarm in transit either with LoRa/LoRaWAN alone or as back up/compliment to conventional cellular coverage. Often a vehicle or vessel in transit will go out of cell coverage but often remain in LoRaWAN coverage depending on where in the world. Also we are now seeing LoRaWAN with satellite coverage as emerging application making it even more viable beyond classic mobile comms, along with sensors talking to mobile GW’s mounted on vehicles and vessels with back haul then implemented over satellite if not conventional cellular comms.
Far, far more typically the reverse. Keep in mind the context - these are precious supplies, flown by air, escorted in transit. When they end up in freezers, they’re in extremely expensive ones no doubt already equipped with monitoring, because such equipment is only used in situations where an auditable record of storage is needed.
A DIY effort is unlikely to be allowed anywhere near it
The most important thing to understand about LoRaWAN, is what sorts of needs it is a fit for, and what sort of needs it is not.
Hi Cedric,
I don’t think you will find a commercial LoRaWan node for temperatures in the -80C range. It is outside of the normal testing range for most commercial semiconductor sensors and electronic assemblies.
For research purposes, or to create a proof of concept, Adafruit have I2C and SPI thermocouple modules that can accept a Type T thermocouple, which should put you in the correct temperature range. There are cheaper options available from the usual global online market places as well.
Put that together with a dev/test module that has reasonable low power performance such as one of the Heltec dev boards, with some distance from the thermocouple and some good RF decoupling, you have a quick to build, cheap and cheerful, test, dev, PoC, or non critical system. Though I suspect it might be more than 50 euros.
A quick Google search provides a number of companies that make LoRaWan IoT nodes that have Type K thermocouples. For a potential commercial application, perhaps approach one of them to see if they can recalibrate for a Type T thermocouple and retest their product at say -100C.
PT100 or PT1000 in my opinion is a more suitable choice for this temperature range. It is far more linear than thermocouples and produces a reasonable change in measurand per degree C.
The project is more likely a proof of concept.
I think lorawan is a good choice because the freezers are located in different buildings in different locations. The freezers store DNA plant samples.
We already use professional services for monitoring our freezers (not using lorawan) but it’s very expensive and the solution is proprietary.
And other teams I work for can’t afford the money so they have no monitoring system at all !
So something is better than nothing 
Of course if the POC is satisfying I will look for ways to improve the reliability and create the sensor. I’m already in touch with an electronic engineer school which could create low-cost sensors from scratch.
Some people in Africa also asked me for a low-cost monitoring system for fridges (4°C), that’s also the reason why I’m trying to realize this POC.
Like french unit research my african friends can’t find the money for deploying professional monitoring systems. So even if it’s less reliable, low-cost systems are better than no systems.
@descartes : The TTN network is not enough reliable for production ?
Yes and no. It’s provided free of charge on a best endeavours basis. If there is any issue with any part of your solution that needs some help with the LoRaWAN elements, you do not have access to commercial support.
Now that you’ve provided enough detail to indicate the application & context, it does not appear to be quite as critical as monitoring Covid vaccines, but who knows, you can make more vaccine, the plant DNA may be irreplaceable. If you are a commercial operation that could suffer financial loss if you lose samples because you weren’t alerted in time, I doubt the directors of the business could justify the use of a free service as you already have budget for it. So a TTI instance may be appropriate.
You may want to consider that you may not need to monitor the actual temperature - most of these freezers have some sort of local alarm which you may well be able to tap in to.
Your colleagues may want to consider their backup strategy - they have the resources to spend on creating samples and procuring & running the freezers, but not the resources to make sure the work is kept safe!
Ok, I understand.
I will have a look.
That’s smart !
Yes, this is a very good advice.
I dont disagree - if only for the fact that the density of cellular coverage is generally much higher than LoRaWAN GW’s in most places. Where there are co-located cellular & LoRaWAN GW’s the LoRaWAN reception area will often be much wider than that for cellular hence many telco/cellco’s have deployed as a complimentary technology 
In this specific case mentioned by Nick yes, but cannot be assumed for OP and often so. My comment relates to general cases, however, just as applicaple where the content is so precious - as an addition (re)assurance!
In such cases there is often both a datalogger (e.g. that can be checked on delivery to ensure no excursions/refreezing during transit) and a mechanism for best efforts live reporting and particularly alarming(*) - hence involvement of LoRaWAN. Indeed the fact these goods are so valuable - both in terms of actual COG’s and in terms of industry or societal value - often justifies the use of such multiple layered protections and monitoring.
(*) Best catch an excursion condition at the time and take immediate corrective action before material spoils, than review a datalog once delivered or days later when the damage already done and its scrap!
True other than for development/POC activities, though I know of many university and hospital engineering & support tech labs working on such projects, and they are extremely professional vs DIY/Hobbyists…and I am actually advocating professional vs hobbyist for precious cargo monitoring. There are obviously many players in the LoRaWAN ecosystem - many of whom are active on the Forum, and/or within the LoRa-Alliance, friends like Laird Tech, who have cold chain protection as a core use ( https://lora-alliance.org/sites/default/files/2019-09/get_smart_with_your_cold_chain_monitoring.pdf ) case for LoRa come to mind, as do smaller start-ups such as Ben, Steve, Ali & the team over at Koolzone here in the UK
Absolutely agree - and that requires an open mind and not prejudging what can or cannot be done with the technology - after >9 years with LoRa then LoRaWAN I am pleased to say I still get a nice suprise every month and that is what helps keep me going with the tech despite odd sticky patches…I like a challenge 
Recognizing basic technical realities is key - for example LoRaWAN not really supporting closed loop reliability with any degree of air efficiency, due to its “uplink mostly” design.
I see all ready a lot off response, but if you build something self with the Dragino as example then look to this manufacturer: https://www.te.com/usa-en/products/sensors/temperature-sensors.html
the range there off some temp sensors are below -50 till -250 degree
Thanks, that sounds good : Platinum RTD Sensor - PT1000 - 3 Wire 1 meter long : ID 3984 : Adafruit Industries, Unique & fun DIY electronics and kits
Make sure the cable used is rated for those temperatures as well. You don’t want it to fall apart…
I came across this node by chance: https://www.milesight-iot.com/lorawan/sensor/em500-pt100/
The specifications cover the required temperature range, you may want to investigate further @CedricGoby
Thanks for mentioning PT100 sensor
We already had a case using PT100 to monitor vaccine temperature during the cold chain supply. @CedricGoby