Author Topic: RFM69HWC sleep mode current > 100nA  (Read 2224 times)

gussabina

  • NewMember
  • *
  • Posts: 42
  • Country: us
RFM69HWC sleep mode current > 100nA
« on: August 20, 2016, 01:12:48 AM »
Hello:

Did someone validate the 100nA in sleep mode for the RFM69HWC modules?
I'm using an mcu which consumes 1.8uA in Sleep mode, so I would expect ~2uA with the radio, but I'm getting almost around 4uA.
I measured both power separately and the difference is taken from the radio module. Is there anything else that need to setup other than setMode(RF69_MODE_SLEEP)?

Thanks
Gus

emjay

  • Full Member
  • ***
  • Posts: 119
  • Country: nl
Re: RFM69HWC sleep mode current > 100nA
« Reply #1 on: August 21, 2016, 12:55:44 AM »
@gussabina,

Are any of the DI/O pins on the RF Module 'floating'? 

perky

  • Hero Member
  • *****
  • Posts: 873
  • Country: gb
Re: RFM69HWC sleep mode current > 100nA
« Reply #2 on: August 21, 2016, 06:53:47 AM »
I think all the DIO pins are actually outputs when configured in packet mode (selectable by a mux in the sx1231h), and the only bi-drectional DIO pin in continuous mode is DIO2. So if it's configured in packet mode there should be no external pull-ups on them to stop unwanted current.

Mark.

gussabina

  • NewMember
  • *
  • Posts: 42
  • Country: us
Re: RFM69HWC sleep mode current > 100nA
« Reply #3 on: August 21, 2016, 08:42:15 PM »
Hello:

The only pins connected to the mcu are the SPI lines (MISO, MOSI, SCK and SS) and DIO0 (interrupt line). All the rest of the radio module are floating. Should I tie them to vcc or gnd?

What should be the I/O internal configuration for those lines (input/output) to keep leakage at minimum?

Thanks
Gus

perky

  • Hero Member
  • *****
  • Posts: 873
  • Country: gb
Re: RFM69HWC sleep mode current > 100nA
« Reply #4 on: August 21, 2016, 10:10:06 PM »
The DIO pins are all outputs from the radio when in packet mode so they don't float, you shouldn't pull them up or down otherwise it might drive the signals into a state that draws current through the pull-ups or pull-downs when asleep. The SCK and MOSI should be driven low by the MCU, the SS# pin driven high, and a pull-up enabled on the MISO on the MCU to prevent it floating (the radio will tri-state the MISO when SS# is high).

It might be worth checking that ClkOut is turned off, bits 2-0 of RegDioMapping2 register (0x26) should be 111. This is not the hardware default value on power up (it's actually 101 which is FXOSC/32) and enabling that buffer even when it's asleep and FXOSC is not running might be the cause of the extra few uA you're seeing.

Mark.

gussabina

  • NewMember
  • *
  • Posts: 42
  • Country: us
Re: RFM69HWC sleep mode current > 100nA
« Reply #5 on: August 22, 2016, 01:33:31 PM »
Hi:
I was in doubt about the clk running while the radio is in sleep, but I though the chip should handle that automatically..
Anyway, good tips... I will check that and share results.

Thanks
Gus

perky

  • Hero Member
  • *****
  • Posts: 873
  • Country: gb
Re: RFM69HWC sleep mode current > 100nA
« Reply #6 on: August 22, 2016, 08:31:41 PM »
Yes, putting the radio to sleep (as opposed to standby) will stop the TXOSC, however the buffer itself might still be enabled. I'm unsure what effect that would have on the current but it is noted in the datasheets that enabling clkOut could cause extra current consumption. Worth checking.
Mark.