Any success with LoRa low-power listening..?

[Jim]

Hi,
I have been struggling with this topic since a year now :O

Is there a chance  to use a rfm95 in receiver with a couple of aa batteries for a couple of month?

[Jim]

Hi,
are there are any news about it in the meanwhile? Still looking for a solution in 2020 and would be happy for a few words or hints.
I'm looking for a battery powered (max 4 AA batteries) receiver with a responds time of approx 2ms that is able to receive a data word of 8 or 9 bits. Operation time should be around 6month with those batteries.
Is there any chance to do it?

[Felix]

2ms? No. A roundtrip of packets will take longer, and thats if your node is actively listening.

[Jim]

incognito wrote something about 2ms. If that's nno possible in the meanwhile I would also be happy with 5ms. That would be great. I'm trying to figure out what is possible in terms of speed and lowest consumption

[Felix]

If you are looking for copy-paste code that works I'm sorry I'm afraid I don't have that to provide.
But please see this thread where a long discussion on increasing throughput with MoteinoM0 took place. Reading that might be eye opening for reaching your goals or providing some answers, there's some sample code there too.

IMO to get into the lower ms transmissions with RFM69/FSK, is possible. For that purpose, forget LoRa.
You will likely have to take a deep dive into the registers of the RFM69 and walk through the library a bit to understand how a packet is formed and transmitted. Then turn off all features like encryption which add extra processing time. Increase bitrate to the maximum 300kbps. See some examples in RFM69 library to set that up, or at the thread above I believe there is some posted as well.

Keep packets to an absolute minimum length. Use a logic analyzer (to analyze the SPI data) and/or CurrentRanger and measure your packets precisely. You will need a lot of patience, but it's very gratifying work when you chase a specific performance aspect out of the RFM69.

[arf]

Any new techniques, insights, guides or approaches with the original post's premise of low power listening with the LoRa RFM95?

Surprised there isn't more interest in this.

[craign27]

I've also being researching this very topic with not much success for my hopeRFM95.
This was the last stage of my node project and has had me stumped for some time, What i've managed so far with the radiohead library is by using "ischannelactive" that i can see a positive rising interrupt on DIO4, but it only remains in this state for a very short time so with some nodes transmitting "hello world" whilst resetting my test unit i occasionally get the interrupt captured on my scopemeter.
When the person wrote this document it was pre version 1.62 (17/8/2016, the later versions have stuff related to CAD) so when you try and add the example given it has errors compiling.
I'm a beginner with arduino and C++ so its tough going.
I'm guessing that a timer to do with the "ischannelactive" is modified it might work to be able to retain it in CAD mode and then use the DIO4 as an external interrupt to wake the arduino / micro controller.