I’ve been messing with my the things uno lately, i am building 2 IOT weather stations out of them. but i cant get them to run on a battery reliably with a “100ma” solar panel with a 18650 attached to it, i would expect that practically would be enough. After messing with my code some time and after looking at the things forum i learned that i can put the board and the radio in to sleep. this got my power draw from about 28 to 5mA. Also removed the resistor in from of the power led to save power. according to those forum posts here its possible to get this board running in the µA range, i can’t get my board to come even close. i use the arduino low power library and a BME280 sensor, more to come but first want this to run good. this sensor draws about 0.5 mA. so in total my setup currently draws 5.5mA, wich is way too much… Can anyone help me?
I’d take the project apart, and build it back up step by step making sure the power consumption is low in the minimal configuration and remains low as each piece of hardware and software is added back.
So for example for the board itself, you want to make sure that you’re getting a much, much lower sleep power configuration. Something is clearly still running that shouldn’t be. Probably do this first without any LoRaWAN code at all; unless you need to do something to get the module into its lowest power state.
The BME should sleep in the low microamp range - unless it’s on a board that has something wasteful like another regulator, different pull-ups, etc, or is not being left in the correct mode.
Make sure you don’t have any voltage across resistors in a sleep condition, as that’s going to consume power. If there’s a pulling resistor, the signal needs to be at the pulled level during the sleep condition.
I have a feeling that the radio isn’t going in to full sleep though and that that is a big part of my issue. because the current doesn’t change much if i only trigger the ttn.sleep(2000000); function only.
Is there a way i can confirm that the radio goes to sleep?
Do you see a drop in power consumption when you put the radio to sleep? You could write some simple code that starts, waits for 10 seconds and then only put the radio to sleep, not the controller as well. While measuring power consumption you should see a difference when that succeeds.
without ttn.sleep(): my board uses 7.85ma in arduino sleep mode, when i trigger the ttn.sleep function power usage of the board goes down to 5ma, so it seems like the radio does go in to some form of sleep
Oh thanks! i did not know that site existed. anyways, before i found that out i removed the regulator, now my power usage went down to about 0.32ma, wich is way better!!!
Is there anyone who used the The Things UNO much in the past? i want confirmation from people that used it in the past to see if my 0.32ma is a good stand by consumption or if i can do way better.
Most people used TTN Uno for a quick prototype not for power optimization as that was not the main target use case for that board. You having to remove the LDO already suggests that…
Okey! thanks for your reply, i can imagine it being an prototype board to attach shields to etc., i also have two ttgo oled esp32 boards, i like the uno boards because of their ease of programming to get into lorawan, and this is the only thing that i could get via my school, lora boards and especially gateways are really expensive right now with the helium trend going on in the netherlands, cant seem to find any student budget TTN capable gateways overseas as well, i now have two ic880a based gateways and i’m really getting the hang of it. I’m gonna close this topic and mark it as solved!
