Although the HW has longer range (about 2x the range), they have different temperature tolerances.
RFM69W -40 to 85 C (-40 to 180F)
RFM69HW -20 to 70 C (-4 to 158F)
There are quite a few days when the temperature goes below -4F if you live in the the northern states. As for the higher temperature, if you stay out of direct sun, you should be OK in most parts of the US, right?
For humidity, there's really no way around it. An outdoor wireless node is bound to get condensation on it in the mornings. Is that OK? Those with the mailbox notifier, do you run into issues? Can you saran wrap the moteino to keep condensation off the board?
I haven't looked at the other components on the chip. I think the ATMEGA chip is good down to -40C and up to 85C, but I can't find the exact datasheet for the Moteino processor. But there's diodes and capacitors on the chip too, and I don't know how those are rated.
I think taking some basic measures to weather seal your outdoor project will do the job.
BTW the "Moteino processor" is the Atmega328 ;)
After I found my ice cream melted while in the freezer, because someone left the door slightly opened, I wanted to monitor the temperature in the freezer. I found the Moteino worked well at 0 degrees Farenheit and the temp sensor (TMP36) worked fine as well.
Problems started below 0 degrees F.
1. Bad data was being sent, possibly the TMP36 started to fail?
2. Moteino stopped transmitting after 3-4 hrs in the freezer, but "came back to life" after being thawed out.
This is likely not how either was designed, but where's the fun without trying?
I attempted to solve the moisture problem by putting the entire setup in tupperware and adding a few silicon dessecants.
So, this project shares some similarities with what you mentioned.
I figured I'd share my experiences for you to build/learn off them.
My attempted setup
Moteino 433Mhz + Gateway Moteino
TMP36
9V battery
Tupperware with twist on lid
Remarkably, I was getting reception from inside the freezer, 2 floors away from the Gateway!
I've not given up, specially since its for the sake of ice cream after all. :)
I also tried transmitting from the refrigerator (not freezer) and it worked a few hours and then quit. I took it back out and it came back to life and then transmitted for several more days until the batteries finally died.
I traced the problem to the fact that I was using alkaline batteries which (like many batteries) perform poorly at low temperatures. The best option for batteries at cold temps are lithium "primary" (non-rechargeable) batteries like Energizer Ultimate Lithium. Another good option are NiMH rechargeable Eneloops. They also have good performance at cold temps, but have lower capacity than the Energizers and also have a higher self discharge rate. For extremely low power sensor/transmitters used at cold temperatures that you want to last for months or years, the Energizer Ultimate Lithiums are the way to go.
Energizer makes a 9V version Ultimate Lithium battery, but I used 3 AAA's for higher capacity and to get closer to the 3v3 operating voltage of the system
Interesting results, thank you for sharing your findings guys.
I can probably confirm that freezing temps take a hard toll on the batteries not the chips.
For my mailbox notifier - I never had a reason to believe a freezing temps affected the radio itself but rather the battery being drained by the cold to the point of failure. There was a clear relation between falling temps and lower battery voltage. The continued cold probably stressed the battery to the point of failure but it was some considerable time before that happened and needed battery replacement.
We've had a brutal winter here in Michigan with record lows and constant freezing temps.
I cannot bid for ultra reliability of these radios since I don't make them, I just redistribute them. I don't know how the chinese source their silicon, if it's authentic or slimy copies. But it's cheap compared to other solutions so that is why we're even having the conversation.
I don't know how significantly sensitive they are to different factors (environmental changes, ESD, vibration, etc).
Bottom line is that the price/performance ratio still appears to be in favor of them, at least that's my conclusion.
People want cheap and good, which is difficult in the DIY/hobby market. I hope by having more exposure more factual answers will emerge. Perhaps the market will shift in the meantime and we might see a different product before getting real answers. Technology does move fast...
In cases like this where you have a particular node that has RF challenges (inside the metal fridge, inside the mailbox, furthest point in a network) can you use an RFM69HW node (of the same frequency obviously) to interface with an RFM69W Gateway?
I guess you might also want to bump the Gateway to an HW, but can it then interface with unchallenged nodes?
LG
The difference is in power output, not in whether they can communicate or not.