It was the LDO-voltage regulator on the Lora e5 mini that was the problem, when I desoldered I get down to a sleep current of 1.9uA!
I think I’ve been a bit confused as the SoC in itself is called LoRa e5 module and when I’ve been reading module in some posts I was thinking of the LoRa e5 mini-board and thought it was possible to get down to 2uA on it without modifying it.
Hey Alexander,
I am currently working on the same problem.
Is it possible to have a look at your code for achieving good deep sleep?
Best regards,
Aleksander
Code sharing is not that trivial for number of reasons.
Best way to go is to start with all-out blinky app. Make sure it works with less than 2uA in sleep mode. Then add 1-by-1 all the features you need. Good exercise anyway. At least, this is what I’ve done when I needed to write firmware for another Lora E5 module device.
This time I’ve used AT-slave template for LoRaWAN part (instead of End Node), managed to make it low power without any problem. Already 1 month in the field and I do not observe any voltage drop on 2xAAA batteries. It has enabled AT-shell on LPUART for changing LoRaWAN keys etc.
You could also hire me if it is an option or write PM, maybe I can help you
Did you disabled more things than the GPIOS? I also disabled all unused GPIOS ( analog mode ) in cubemx but this doesn’t changed the power consumption. My E5 still consumes about 2.5uA. Of course this is great, but your 1.3uA looks way better
I think they’ve fixed it in the latest firmware package. So, blinky works just out of the box, even if you add UART (I use LPUART in my latest firmware).
Do you have anything connected to your module besides power line and antenna? 1.2uA is a very low current, can be leaking via some pin just because you still have LED or debug interface connected, even if you do not use them. Or if UART wires are still connected. Anything.
What it also can be: power regulator, decoupling capacitors leaking etc. And you are using LoRa-E5 module (not board), right? Like you’ve made your own PCB and soldered it all together and nothing is galvanically connected to the module besides power and may be antenna?
I am also not sure about 1.3uA much better than 2.5uA. It’s about 10.5 mAh per year of difference. Even CR2032 has like 235mAh. You are not going to run your device on CR2032, do you? You still need a lot of power to drive the radio. 1.3uA is achievable in sleep, but below certain value it doesn’t matter any more because most of the power will be spent when device is doing something useful.
I made my own board, but for testing I only soldered one single E5 chip to the board without any other component. How do you measure the current? I am using the nordic semiconductor power profiler kit ii.
Yeah you are right it doesn’t make much difference, but as a developer, we always want to achieve the best, or?
Yes, I also use Nordic PPK2. Can it be calibration problem? Try smallest example, like blinky - I do not think you will get anywhere lower than that. At least, you exclude the firmware. And measure the current with good multimeter also, just to compare with PPK2.
Now I get also 1.7 uA. I forgot to disable SWD. In between I have a DCDC buck converter, so I think without it I am also at 1.3 uA. Very nice. Thank you for your time