How to reduce MCF52259 VSTBY  current draw

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

How to reduce MCF52259 VSTBY  current draw

995 Views
CarlFST60L
Senior Contributor II

Hi,

 

I am just after some clarification of current draw from VSTBY on the MCF52259. Is there a way to reduce the standby current draw from 20uA down to around 1uA (Which is what it is when VCC is on)? Is it possible to 'turn off' some features to get the current draw down? i.e. stop VSTBY from supplying SRAM in standby  and just runt the RTC?

 

It seems strange to draw nearly 20uA to run a RTC, even 1uA is a big much really... I did some initial testing a couple of years ago and noted it was drawing 1.5uA at 3.3V in standby, however, I now realize that if you power it up and then down (initialize RTC etc), it jumps to nearly 20uA which is just to much for our application.

 

The smallest viable lion battery (CR1220) doesn't even make 100 days standby before it dies!  CR2450's etc are just way to big to run something as light weight as a RTC, in cost and space...

Labels (1)
0 Kudos
4 Replies

580 Views
TomE
Specialist II

20uA is what the Data Sheet says. I wouldn't expect there would be any way to lower it. hat's what you have to design for if you've selected this chip.

 

If you need a lower power draw then use an external SPI/I2C-connected RTC.

 

Search for "VSTBY" in this forum and you'll find a detailed post from March 2009 listing the power draw - up to 31.7uA in one test.

 

https://community.freescale.com/thread/51895

 

The usual reason why these chips draw more than the watch on your wrist is that they're made with different silicon processes, optimised for higher speed and power than your watch needs. The RTC and SRAM need to be read QUICKLY when the CPU is running. Sometimes there are unexpected "leakage paths" inside the chip as well that draw current. I've read Chip Errata detailing these problems. Sometimes they release new silicon to fix the problem, and other times it is "will never be fixed".

 

I'd also suggest measuring the current and then start grounding any floating pins on the CPU. It might be that a floating pin on a powered input gate is drawing some extra urrent.

 

Tom

 

0 Kudos

580 Views
MrBean
Contributor I

External RTC, if you want low power, as Tom says.

 

250nA, I2C $0.95 @1k

PCF8563

 

130nA, SPI $0.70 @1k

PCF2123

0 Kudos

580 Views
CarlFST60L
Senior Contributor II

Thanks for the links, I ended up removing the parts from my board and using our network for time sync.

 

Would be nice for the RTC in the processor to be 'useful' ; )

0 Kudos

580 Views
mjbcswitzerland
Specialist V

Note that the RTC design in the Kinetis (a potential migration path for Coldfire users) is a big improvement at 550nA typical current draw (1.71V..3.6V supply).

 

It is however only a second counter (so the SW needs to do the convertion to years, days, hours, minutes etc.) and it doesn't have any non-volatile RAM backup, but should save the use of external parts in much more designs.

 

Regards

 

Mark

 

0 Kudos