I measured the current drawn from the RTC backup battery and was horrified to find it was 175 µA. Yes, one hundred and seventy five micro Amps!

I measured the current directly using an analogue multimeter on the 50 µA range. The needle hit the end stop. The next range up was 5 mA and the needle barely moved. So I fitted a 1K Ohm resistor between the battery and the VB pin and measured the voltage drop using a DVM on a 300mV range.

The mbed spec says 27 µA. The LPC1768 datasheet on the NXP web site says TBD. I was hoping for less than 10 µA.

At 175 µA, my 48mAh 1225 size coin cell will only last 11.4 days. :-(

Has anybody else measured this? Do the mbed team have an explanation?

Paul

Paul Griffith wrote:

I measured the current drawn from the RTC backup battery and was horrified to find it was 175 µA. Yes, one hundred and seventy five micro Amps!

I measured the current directly using an analogue multimeter on the 50 µA range. The needle hit the end stop. The next range up was 5 mA and the needle barely moved. So I fitted a 1K Ohm resistor between the battery and the VB pin and measured the voltage drop using a DVM on a 300mV range.

The mbed spec says 27 µA. The LPC1768 datasheet on the NXP web site says TBD. I was hoping for less than 10 µA.

At 175 µA, my 48mAh 1225 size coin cell will only last 11.4 days. :-(

Has anybody else measured this? Do the mbed team have an explanation?

Paul

paul, I measure 136uA (!) when the device is plugged into usb,
and -500 ~ 0 uA when the device is unplugged.

Not sure why the reading is negative (maybe the mbed capacitors are charging the battery??) but drawing 136uA while plugged in USB is definitely not good. Will have to look into it to figure out exactly what's happening. 

There was a bug on 2368, I wonder if there's something similar in 1768:

Vbat.1
Increased power consumption on Vbat when Vbat is powered before the 3.3 V supply used by rest of the device.

Introduction
The device has a Vbat pin which provides power only to the RTC and Battery RAM. Vbat can be connected to a battery or the same 3.3 V supply used by rest of the device (VDD(3V3) pin, VDD(DCDC)(3V3) pin).

Problem
If Vbat is powered before the 3.3 V supply, Vbat is unable to source the start-up current required for the Battery RAM. Therefore, power consumption on the Vbat pin will be high and will remain high until 3.3 V supply is powered up. Once 3.3 V supply is powered up, power consumption on the Vbat pin will reduce to normal and subsequent power cycle on the 3.3 V supply will not cause an increased power consumption on the Vbat pin.

Workaround
Provide 3.3 V supply used by rest of the device first and then provide Vbat voltage.

Igor,

I cannot see how this workaround can work. If we want to maintain the RTC when power is off, then Vbat has to be present when power is off. It will therefore be present when the main 3.3V supply comes on. If we were to remove Vbat just before turning main 3.3V on and then restore it, we would lose the RTC values.

Paul

Hi Paul, Igor,

Chris is writing up a report on this after having done some investigation. If you can hold out 24hrs, hopefully you'll be pleased with the results we've got :)

Simon

Hi All,

I don't want to keep the suspense going for too long...

I've got some more investigation to do tomorrow with higher sensitity measurements, i'll update my notebook page as I go. In the mean time, here is my notepage so far, comments please!

http://mbed.org/users/chris/notebook/rtc-battery-backup-current/

Cheers,
Chris