How close is the device to the gateway - it looks like it was heard on two frequencies at the same time.
You need 10m &/or a brick wall between a device & a gateway other wise the shouting in the ear of the gateway can cause it to think it’s receiving on two or more channels at the same time.
What I would do is remove the antenna. Funny as it sounds I have done it before, Run the node but forget to put the antenna, still see packets when I am up to a meter close to the gateway. Quick and dirty way to lower your EIRP for testing purposes.
Or use this:
at+set_config=lora:tx_power:X
Set the TX power level. X - The level of TX power. I
Crank it down a bit and observe, hopefully this intermodulation product you seem to be seeing will attenuate sufficiently to be below the sensitivity level.
I have another question: my gateway and node both have antennas. Do I just need to remove the antenna of the node? Or remove them both? Or what kind of permutation?
This is in no means a good practice
Generally you should absolutely not do this, but for a quick test it will work. I am only suggesting it as I have forgotten to put mine on the RAK811 numerous times (unintentionally) and the result was a lot lower RSSI, nothing else.
Leave the gate alone it is expensive, you don’t want to mess with it (not that I haven’t done that too, again unintentionally).
But try reducing the power with the AT command first just to stay safe observe the results and if need be go nuts, ahah.
It sounds like you will operate your device very very close to a gateway. Just use serial comms and wires
Anyway, I don’t suggest you remove the antenna of the gateway. It is more expensive compared to your device and also, other devices might connect to it as well. For devices, I have lots of time running stuff without antenna connected too. Sometimes I am just lazy to put the antenna do to quick test
The only bad thing of disconnecting the antenna is the reflection of the RF signal going back to your RF chip. Unless you add a 50ohms resistor to match the transmission lines to match it.
You don’t need to touch the gateway, it can all be done over the network - so just put it at the other end of the room or in another room with a note on why it’s plugged in there.
You could even explain to your supervisor that it won’t work unless you do this.
I have a derived question: as shown in the third line of the gateway log below, is it normal to receive two or more RF packets in a single device transmission, or is it also caused by too close distance?
Normally, only one packet will be received??
The difference between this problem and the original problem is that the RF packets mentioned here will be two or more different packets instead of the duplicate packets mentioned at the beginning.
A device can’t transmit two packets at the same time as it only has one transmitter circuit. So if you are consistently getting two or more packets arriving at the exact same time, then yes, it almost always is device & gateway too close, particularly when you get CRC errors as well.
What is the RSSI and SNR of the uplink that survived? A device close by would be on something like -50, a device much further away would be -110 with all values in-between.
This is about the laws of physics - you don’t get to negotiate, neither does your supervisor. Just give it a try with the gateway further away - 10m &/or brick wall is the ideal as usually people just shove the gateway a couple of meters away when you tell them that but if we said 2m they’d try 50cm.
There’s a firmware & backend expert over on the TTN forum, a really great bloke, very helpful, ask for @descartes - he’ll confirm that he answers this question all the time even though he’s not so much in to gateways & RF. But he has a little secret. He sticks a glass of water over the spring antenna on the RAK833 on his test gateway on his desk and puts the antenna of his test device behind one of his iMacs and that does a great job of attenuating the signal enough to not cause any issues.
The lowest rssi value is about -20, and the value of lsnr is always around 14.
It’s really abnormal compared to the value you described. I see that when multiple packets are received at the same time, the RSSI of the extra packet always jumps to - 70, which may be a reason.
You said the packets are different this time around. How do you know?
If they are different can the answer simply not be that they are not from this node at all?
This is a RAK7258 right?
In fact, I also see from the gateway log that the ‘difference’ here probably refers to receiving CRC correct and wrong packets at the same time, which is not strictly equivalent to the packets sent by different devices.
INFO: Received pkt from mote: 26041304 (fcnt=8)
INFO: Received pkt from mote: A22082B5 (fcnt=36936)
JSON up: {"rxpk":[{"jver":1,"tmst":235580894,"chan":7,"rfch":1,"freq":487.700000,"mid": 8,"stat":1,,"modu":"LORA","datr":"SF7BW125","codr":"4/5","rssis":-17,"lsnr":13.5,"foff":-238,"rssi":-16,"size":18,"data":"QAQTBCaACAACzaMSupFAJTka"},
{"jver":1,"tmst":235580894,"chan":2,"rfch":0,"freq":486.700000,"mid":10,"stat":-1,,"modu":"LORA","datr":"SF7BW125","codr":"4/5","rssis":-84,"lsnr":-7.2,"foff":-226,"rssi":-76,"size":18,"data":"Y7WCIKLASJACmANO6WTYrKgb"}]}
