Author Topic: coin cell powered rfm69 module  (Read 1698 times)

0 Members and 1 Guest are viewing this topic.

Offline mmrveljTopic starter

  • Newbie
  • Posts: 6
  • Country: hr
coin cell powered rfm69 module
« on: March 28, 2019, 08:54:29 am »
I am trying to build low power (coin cell CR2032 powered) simple device that would be used to transmit occasional messages (pushbutton or beacon) by using RFM69 module.

For this I have selected to use ATmega328p MCU mainly because I have (some) experience with it.

I did manage to have a working setup, however only when I power the device externally. It is not working when powered with CR2032 (because it draws too much current). I am still trying to figure out why, I will start debugging tonight, however I would appreciate if someone with more experience could take a look at my schematic and comment if there are obvious design flaws.

I am measuring current consumption of around 40mA  :o , although I am putting ATmega and RFM69 to low power modes (using LowPower library for ATMega and radio.sleep for RFM69).
« Last Edit: March 28, 2019, 08:56:48 am by mmrvelj »
 

Offline mmrveljTopic starter

  • Newbie
  • Posts: 6
  • Country: hr
Re: coin cell powered rfm69 module
« Reply #1 on: March 28, 2019, 08:55:43 pm »
Update, with RFM69 disconnected I measure rest of the board sleeping at 18uA. So probably I am not putting RFM69 to sleep properly..
 

Offline viperidae

  • Frequent Contributor
  • **
  • Posts: 306
  • Country: nz
Re: coin cell powered rfm69 module
« Reply #2 on: March 28, 2019, 10:33:50 pm »
40mA is quite a bit for the RFM69. It's only supposed to draw that much power in transmit mode. The datasheet says 33mA at +10dBm.

You're also probably going to get up to 1V drop from the internal resistance of the CR2032 if you pull 50mA while transmitting, assuming 20R.
 

Offline Rolo

  • Regular Contributor
  • *
  • Posts: 206
  • Country: nl
Re: coin cell powered rfm69 module
« Reply #3 on: March 29, 2019, 06:54:50 am »
Also check at what brown-out voltage your ATmega is set, as said the internal resistance of these coin cells give a drop in Vcc, if it's goes under your BOD voltage your mcu will stop working or give unexpected behavior.
I'm using mysensors node's with the RFM69W and have sleep currents of 6 uA, with no other sensors attached, just an atmega and the radio module. So a good sleep is possible.
 

Offline mmrveljTopic starter

  • Newbie
  • Posts: 6
  • Country: hr
Re: coin cell powered rfm69 module
« Reply #4 on: March 29, 2019, 12:47:08 pm »
Thanks, I am guessing that I might have messed up something in software part (at least I am hoping). ATMega BOD is set to 1.8V, I hope this should be good enough?

I know there are similar designs that do work with CR2032, so in theory, it should be possible :)
 

Offline Rolo

  • Regular Contributor
  • *
  • Posts: 206
  • Country: nl
Re: coin cell powered rfm69 module
« Reply #5 on: March 29, 2019, 03:49:40 pm »
A CR coin cell is considered empty at 2.0V so in theory the BOD at 1.8V is fine, but, at the end of it's life the cell has higher internal resitance making the Vcc drop even higher, for short burst below the 1.8V. This makes it also hard to measure when the cell is empty.
This could be an interesting read: https://www.eevblog.com/forum/projects/cr2032-battery-voltage-with-ble/msg2277234/#msg2277234

 

Offline mmrveljTopic starter

  • Newbie
  • Posts: 6
  • Country: hr
Re: coin cell powered rfm69 module
« Reply #6 on: April 13, 2019, 05:17:59 pm »
It turned out it was software issue. Not sure what was wrong, but now I have sleep current ~ 6 uA.  It might be that RFM69 needs to be initialized before you can put it to sleep, and my testing sketch was instantly sending it to sleep mode.

Now I need to see if CR2032 can consistently provide sufficient current for the transmission. I am able to send a single messages once per minute for the last 30 minutes. Now testing to see if it is stable.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf