Battery backup for RTC in AT91SAM9261

[rickman]

The AT91SAM9261 uses 1.2 volts on the backup voltage for the RTC. It
only needs 2.5 mA, so it will run nicely from nearly any battery. But
to use a primary cell requires some sort of regulator to reduce the
voltage. Anyone have any suggestions on how to do this without
increasing the battery drain enormously? All the off the shelf LDOs
have quiessent currents of 10x and higher.

I just bugs me tha they designed an RTC that you can't just connect to
a battery. Couldn't they have made it work with a 3 volt lithium
battery like all the other RTC circuits???

 

[Martin Riddle]

The AT91SAM9261 uses 1.2 volts on the backup voltage for the RTC. It
only needs 2.5 mA, so it will run nicely from nearly any battery. But
to use a primary cell requires some sort of regulator to reduce the
voltage. Anyone have any suggestions on how to do this without
increasing the battery drain enormously? All the off the shelf LDOs
have quiessent currents of 10x and higher.

I just bugs me tha they designed an RTC that you can't just connect to
a battery. Couldn't they have made it work with a 3 volt lithium
battery like all the other RTC circuits???

Looking at the Eval board schematic they have a 3 terminal device R1100DC121C.
Ultra low power LDO http://www.ricoh.com/LSI/product_power/vr_linear/r1100d/r1100d-e.pdf

National's part might be a sub http://www.national.com/pf/LM/LMS5258.html#datasheet

Cheers

 

[rickman]

Looking at the Eval board schematic they have a 3 terminal device R1100DC121C.
Ultra low power LDO http://www.ricoh.com/LSI/product_power/vr_linear/r1100d/r1100d-e.pdf

National's part might be a sub http://www.national.com/pf/LM/LMS5258.html#datasheet

Thanks for the info. The Ricoh part looks interesting, but I will have
to see how hard it is to get and the price.

The National part has a way higher power consumption...
Quiescent Current (mA) = .14 mA

Having to add the regulator makes the RTC less useful. The total power
is now 4 uA vs. 0.5 uA for a real RTC. For about the same price as the
LDO, I can add a RTC chip at a much lower power consumption with a
smaller battery and be done with it.

I guess none of the ARM MCUs are really perfect. The entire TI family
is not 5 volt tolerant. I guess you can get away with that when you
are going for ultra low power, like the MSP430 or selling FPGAs, but I
am surprised that an embedded processor family would not be 5 volt
tolerant.

 

[Gerhard]

The AT91SAM9261 uses 1.2 volts on the backup voltage for the RTC. It
only needs 2.5 mA, so it will run nicely from nearly any battery. But
to use a primary cell requires some sort of regulator to reduce the
voltage. Anyone have any suggestions on how to do this without
increasing the battery drain enormously? All the off the shelf LDOs
have quiessent currents of 10x and higher.

I just bugs me tha they designed an RTC that you can't just connect to
a battery. Couldn't they have made it work with a 3 volt lithium
battery like all the other RTC circuits???

Have a look at the Seiko S-817-A25 regulator
Specs: 2.5 volt, 50 mA max, 1.2uA quiesent, 0.3V dropout
available in a number of SOT packages and TO92
Can be orderd from Mouser in small quantities.

Gerhard van den Berg
CSIR

 

[Gerhard]

Have a look at the Seiko S-817-A25 regulator
Specs: 2.5 volt, 50 mA max, 1.2uA quiesent, 0.3V dropout
available in a number of SOT packages and TO92
Can be orderd from Mouser in small quantities.

The Seiko S-812C25 is actually a bit better.
Specs: 2.5 volt, 50 mA max, 0.9uA quiesent, 0.3V dropout
This is for the TO92 package, the dropout for the SOT packages
is 0.5Volt. Also available from Mouser.

Gerhard van den Berg
CSIR

 

[rickman]

The Seiko S-812C25 is actually a bit better.
Specs: 2.5 volt, 50 mA max, 0.9uA quiesent, 0.3V dropout
This is for the TO92 package, the dropout for the SOT packages
is 0.5Volt. Also available from Mouser.

Thanks for the info. The S-812 will not do the job. The required
output voltage is 1.5 volts and this line only goes down to 2.0 volts.
The S-817 will work down to 1.1 volt output. This is no great part.
The dropout can be as high as .99 volts, at least at 10 mA, I guess it
would be a lot lower at uamp currents. The regulator current can be as
high as 2.5 uA which would double the current from the battery.

This is a lot better than no backup circuit, but it still irks me that
they designed a circuit for a non-existent backup battery and left it
to the designer to figure out how to provide the power.

 

[Rich Grise]

Thanks for the info. The S-812 will not do the job. The required
output voltage is 1.5 volts and this line only goes down to 2.0 volts.
The S-817 will work down to 1.1 volt output. This is no great part.
The dropout can be as high as .99 volts, at least at 10 mA, I guess it
would be a lot lower at uamp currents. The regulator current can be as
high as 2.5 uA which would double the current from the battery.

This is a lot better than no backup circuit, but it still irks me that
they designed a circuit for a non-existent backup battery and left it
to the designer to figure out how to provide the power.

What about a 1.2V ni-cad, that you'd leave on trickle charge as long as
the equipment is powered up?

Or is the whole thing battery operated? If that's the case, then never
mind.

Good Luck!
Rich

 

[rickman]

Have a look at the Seiko S-817-A25 regulator
Specs: 2.5 volt, 50 mA max, 1.2uA quiesent, 0.3V dropout
available in a number of SOT packages and TO92
Can be orderd from Mouser in small quantities.

This turns out to be an adequate regulator. With a quiescent current
of up to 2.5 uA, it can double the power consumption of this circuit.
It will require a larger battery than I would have liked to have used,
a 2032 will give me almost 5 years of backup compared to a 1632 for the
RTC alone.

The S-817A12ANB regulator comes in a package as small as the SC82 which
is about 2 x 2 mm. So at least size is not an issue. Cost at Mouser
is only $0.30 (@3000) so cost is not an issue either. But it still
irks me if for nothing else because I have to add another part to the
board... >:-(

 

[Gerhard]

Thanks for the info. The S-812 will not do the job. The required
output voltage is 1.5 volts and this line only goes down to 2.0 volts.
The S-817 will work down to 1.1 volt output. This is no great part.
The dropout can be as high as .99 volts, at least at 10 mA, I guess it
would be a lot lower at uamp currents. The regulator current can be as
high as 2.5 uA which would double the current from the battery.

This is a lot better than no backup circuit, but it still irks me that
they designed a circuit for a non-existent backup battery and left it
to the designer to figure out how to provide the power.

Sorry for the incorrect info. I spotted my mistake just after the second
post
but when I tried to reply our news server was unreachable (I could see the
cashed items). The server only came back 24 hours later.
I am glad you spotted the S-817A12ANB. I checked to see if the
S-817B12AUA (B series without short circuit protection) had a lower
quisent current, but it was the same :-(.
Yes I agree, an onboard 1.2v backup regulator would have been great.

Gerhard van den Berg
CSIR