There are lots of them. Google "Lora certified"
And where would one procure one such item?
This reminds me even more urgent need to have LoRaWAN stack implemented in your controller to mitigate risk of "modem" vendor decay (and suffer costs of PCB redesign due to modem vendor switch), though nobody would assume Microchip could be an example.
Yesterday I was lucky to buy 5 modules RN2483A at arrow.com. They had stock around 40pcs. What was shocking, the delivery to Europe (Czech rep.) was $0 extra via DHL. Looked today at Arrow module is again on backorder. So they exist after all!
I can confirm that both RN2483 & RN2903 are in production and shipping, but they go out in order and the demand is high, so the stock is always exhausted. It should be caught up after the next batch.
We received the first batch of 1200 pcs RN2483A and started production of a new batch of SODAQ ExpLoRer boards with it!
So that's where all those RN2483's went to.
I can confirm new revision RN2483/A with firmware 1.0.3 indeed handles sleep power consumption much better than 1.0.1. It's now below 10uA - I couldn't measure exact number, my amp meter displayed just 0.00mA.
Are you using an old RN2483 module updated to firmware 1.0.3 or a new RN2483A module?
Would love to know actual power specs and also TX/RX peaks
The RM186 we use now has no problem hitting 4 μA in sleep doze but up to 10 would be OK as well so could be an alternative.
That test was with RN2483A, e.g. new module.
Most distributors seem to have ample stock again.
The datasheet is here: http://www.microchip.com/wwwproducts/en/RN24831.6uA nominal (room temp and 3.3v) and 2.6uA max at high temp. (Other products will have the same temperature dependance but probably not documented)
Upgraded the firmware to 1.0.3 on the older RN2483 (without A) and sleep current went down to 2.8 uA at 3.3V.