Systematic data pauses between gateway and application

Hi All,
I have a problem with the data of my sensor. I run a sensor with a simple “Hello World” uplink every 5min. In the gateway I see the uplink message every 5 minutes. In the Device, however, only 3 contiguous and then half an hour no data.
Here the Device:

Here the Gateway

In parallel to the Device Live Data window I have both a MQTT and a HTTP query running. These also show me only the data I see in the live data window.

Here is the configuration of the device

image

The airtime per transmission averages about 0.061s, so . Here I see no violation of the 1% airtime.
For test purposes I have removed the checkmark in the LoraWan settings to be on the safe side.

image

What I noticed is that the timestamp goes 2 hours after. As an example instead of 16:30 time the timestamp 14:30.

I suspect that there is a timer or counter active somewhere that limits the number of uplinks that go through. But I don’t know V3 that well so far. Have TTN last used 2 years ago. Would be happy if someone could help me :slight_smile:

Greetings
Hotte

Welcome,

First please stop sending plain text in the payload, waist of air time.

Search on the forum how to reduce your rssi, the node is shouting at the gateway.

1 Like

The console is not a source of truth. You’ve not told us what the integrations receive.

No need to load the community server with both MQTT and a web hook, a web hook being much more preferable.

The times are UTC.

Redacting the DevAddr which is shared by multiple devices is pointless. The only secrets to redact are the AppKey and any API keys you have.

As above, the device is too close to the gateway and will be overloading the gateway input.

What is the device, are you using ABP or OTAA, look at it from the perspective of the the volunteers answering for free, what info do you think would be useful. More is less.

As kindly pointed out by @Johan_Scheepers, whilst the payload/uplink integration is UTC, the console shows it in local time which is very much dependent on your computers settings, something for you to look at - and perhaps compare with the payloads timestamps.

Hello together,
sorry for the late feedback. First of all, thank you for the feedback on the topic. The RSSI is currently so high because I have the sensor sitting right next to the gateway for testing and code adjustments. The problem did not change when it was further away (RSSI at about -100).
The integrations both show exactly the same as the terminal in the console. So there is no deviation here. The timestamp is generated by the code. I assume that here simply another time zone was selected. But this has no influence on the transmission of the payload.

I use a Dragino LoRa shield v1.4 in conjunction with an Arduino Uno and a standard program from the LMIC library. The activation is done via ABP.

I was able to solve the problem. It was because the code of the sensor uses several frequencies to distribute the load. I had not entered these extra whereby always only 3 standard frequencies have come through.

My current problem is that sending data to the application freezes after an undefinable time. From this point on, the end device no longer receives data despite the display on the gateway. The problem can be solved manually by resetting the MAC state and resetting the sensor.
Therefore the question.
Is there a possibility to reset the MAC state automatically?

Up to date versions of LMIC do include this. As do older versions. But you have highlighted why ABP sucks.

This is most likely hardware related as I have nodes on hundreds of thousands of uplinks on LMIC on ABP

By freezing do you mean the debug serial log isn’t showing any more info? [Please think ahead to what information will help the volunteers with answering] Do you mean downlinks which is the end device receiving? If the firmware has frozen then nothing a gateway transmits is going to cause a device to restart.

As above, I’ve devices that are doing just fine. But getting LMIC in to an ATmega328 leaving room for local variables takes some configuration. What optimisations have you implemented?