RFM69HCW 433 1/4 unipole - reestablishing an out of range point to point link.

[ozpos]

Hi All.

Thank you for publishing your excellent library for RFM69HCW module.  I am sorry to admit that I am not using your hardware because I had to get something quick before my crazy rescue lurcher goes missing again and does not come back.  The last time he went chasing deer he ran into a post and was could hardly walk.  Also to be clear, I am not seeking any help with this hardware.

I am using the Feather M0 RFM69 with your software and downloaded this https://lowpowerlab.com/forum/rf-range-antennas-rfm69-library/node-payload-clash/msg16236/#msg16236
example code from 2016.  The code works just fine and I get about 300m range with un-tuned 23swg unipole 163mm long antenna.

Thanks to your library the battery life is fantastic and more than adequate for my needs. My intention is to either

• Just have the single Feather M0 RFM69HCW as the receiver and use a DIY tape measure yagi attached to the Tx Feather with a display showing the signal strength (so I know which direction to go in) if he fails to return or
• Use a Feather M0 RFM69HCW plus a FeatherWing GPS module for the receiver and at the Tx end use the same Feather with bluetooth Wing linked to a mobile phone with a map.

I realise that I will probably need to not use both the RFM69 module at the same time as the GPS or bluetool modules in order to get the best signal to noise ratio.

On to my questions:
Today I did a range test by leaving the single Feather tx at home and walking the dog while holding the single receiver Feather until the link was lost.  On my return the link was reestablished however it was probably 100m shorter than when the link was lost (i.e. ~200m from home).

Could there be a software reason for this ?
Should I modify the application code so that if the link goes down for more than say a minute I set the Tx to full power ?
Is there a similar adjustment I should make at the Rx end ?

Best regards,
oz

P.S. Sorry about the list numbering and paragraph layout, I have not used this html editor before.

Just had a thought about sensitivity.  When I set off the receiver pcb and lipo battery would have been at room temp and the walk took ~ 1hr at arround 4 degree C, could that have knocked off 100m ?

Some more info if anyone is interested.  I put the receiver in the fridge for an hour before doing the same range test again and observed ~300m outbound before the link was lost and on my return the module reestablished the link ~280m from the transmitter.  Overall very impressive.

Thank you again for this library.

[Felix]

You're welcome, I'm happy that my work is appreciated.

Thanks to your library the battery life is fantastic...

You should try it on a Genuine Moteino, and then let's talk about battery life

Today I did a range test by leaving the single Feather tx at home and walking the dog while holding the single receiver Feather until the link was lost.  On my return the link was reestablished however it was probably 100m shorter than when the link was lost (i.e. ~200m from home).
Could there be a software reason for this ?

I did a recent blog post on range testing with LowPowerLab Moteinos and RFGateway, but I used helicals instead, specifically for indoor to outdoor signal strength evaluation. Still the distances are very good, and the code examples to help evaluate range/RSSI should be a little better that what was posted before.

To your Q: It's always hard to estimate what's wrong when you have a significant drop like that. With RF there's many factors to consider. It's always best to try to replicate the same test/comparison conditions, as much as possible. In other words always control/change a single variable at a time. Otherwise a single thing that is not apparent, can throw off the whole comparison. I do not believe there is any software component that will cause this, if things are done right (but that's a comprehensive implication!). So to get a better idea how you got 300m then 200m when you walked the dog, then maybe we need to talk what was different in the second test .. the dog maybe haha? 

RFM69_ATC is essential to get good battery life, especially in frequent transmitters. The library already does a great job auto-tuning TX power (after several no-ACKs the transmitter should already reach full power). So IMO you should not have to mess with that much, unless you manually control TX power and ATC is disabled. A gateway/RX does not adjust TX power (it only tells the distant transmitter how much the RSSI was, so the transmitter can use that info to auto-tune TX power), so no, you don't have to deal with RX tuning. The RX side usually does not have a battery life concern so there is no need to dial down TX power on it, it will generally always use max TX power level. Now of course, there are ways to increase the link sensitivity but that is getting more complex, the library is tuned so it can give you great all around RF settings that work in the majority of cases. In the vast majority of my own personal and consulting projects I have not had a need to tweak RF settings from the defaults.

[ozpos]

Thank you very much for your comprehensive reply.  It is important for one to know where to place their confidence especially when working in unfamiliar territory.

Best regards oz.