Powering an AVR with Li-Ion battery

Go To Last Post
19 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello to everybody,

I'd like to power my AVR project using this battery http://www.sparkfun.com/commerce... and this Li-Ion charger (http://www.sparkfun.com/commerce... ).

All the system should work at 3.3V, because that's the power supply voltage for the Atmel Dataflash, also there are so many ICs that uses the voltage, and the battery level is 3.7V (nominal). My question is related with how to "regulate" and supply the MCU with a stable 3.3V from a power source with 3.7V nominal. It will be great if I can still have 3.3V when the battery is near 3.4V.

Does anybody use a LiIon battery to power an AVR or any other MCU at 3.36V?

Thanks very much!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Does it really require a regulator?

The question here is what's the max voltage of the other components.

If you need it, just pick a low-drop one from the manufacturer you like the most.

Jörg Wunsch

Please don't send me PMs, use email if you want to approach me personally.

Last Edited: Tue. Jan 29, 2008 - 12:01 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The problem is that some ICs need between 3.0 and 3.6V and I dont' want to be "on the edge", I would prefer a stable 3.2 or 3.3V suppy capable of source nearly 1000mA.

Why do you doubt about using a regulator?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A voltage regulator is not needed. You may use a diode (or two in parallel) in series with the power supply. The voltage drop may be adjusted by selecting regular diode (1N4007) or a schotky diode (1N5819). Get the diodes datasheet to see the current versus voltage drop. This will also avoid damage if the battery polarity is reversed by user misoperation. I use an Atmel Dataflash at 3.4V. The main supply is 5V. With 2 1N4007 and a 1N5819 the voltage is drop to 3.4V. I use a resistor (2K2) in parallel with the dataflash to keep a small current flowing through the diodes and the their voltage drop. Without the resistor, when the dataflash is in power down, the voltage would be near 5V.

Eng.Marco Aurélio Carvalho

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
I would prefer a stable 3.2 or 3.3V suppy capable of source nearly 1000mA.
A buck/boost regulator should do all that you need. Maxim (and others) makes some.

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

I'd like to power my AVR project using this battery ...

Quote:

... capable of source nearly 1000mA.

So a 1Ah battery will last one hour?

Lee

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I use a low dropout linear regulator succesfully for an application. And in my case is even worse, since I have a schottky diode in series with the backup LiPol battery (to switch from mains or battery if the first fails). Anyway, my consumption in the worst case is about 30mA peak at bursts every many minutes (up to days), much lower usually (<1 mA). This way, I can run from a 550 mA batt about one week.

Guillem.
"Common sense is the least common of the senses" Anonymous.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

> Why do you doubt about using a regulator?

Because the AVR could tolerate the full voltage range of a LiIon
cell. You didn't write what other stuff exactly you've got
connected to the circuit. If other devices are picky about an
uppper limit of 3.6 V or so, yes, pick a low dropout regulator
from the chip manufacturer of your choice. The issue with these
is that all of them need additional current to operate, so they
cause more load to your battery.

If that is any concern to you, an option is to select a low-drop
regulator that has a shutdown input. Then, power the AVR directly
from the LiIon cell, and let it control the operation of the
regulator.

Jörg Wunsch

Please don't send me PMs, use email if you want to approach me personally.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well, I have a screen and an I2C device which works at 3.3V with an upper limit of 3.6V and a lower limit of 3.0V.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I would say you need very low drop voltage regulator.
LiPo cells voltage vary 4.2V per cell till only 3V when discharged, in the middle it likes to stay around 3.8V or so depending on load.
Similar situation with LiIon:
http://www.uniross.com/images_in...
This is huge variation.

Also remember that this cell voltage vary much with load.
Also remember that voltage regulator will filter out external interference from f.ex blinking diodes connected to the same power source.
Don't forget that without decent voltage regulator you cannot use A/D converters in uC.

Using diodes is risky since their forward voltage drop varies with temperature and voltage.
Also remember voltage regulator protects from short circuit, a diode will just explode at 1A?5A? you never now.

Also remember in order to prevent LiIon or LiPo from total discharge (that damages battery in some 80% cases irreversibly) you need disable pin in voltage regulator, and most likely reverse battery connection protection.

All in all Micrel 29xxx (MIC29151-3.3BT?) or something similar to LM2940 but for 3.3V is well worth the price (heard about negative experiences from some ppl using LM2941 when tuned under 5V, try to avoid potential trap).

Last Edited: Tue. Feb 5, 2008 - 04:31 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi,
I also suggest the LDO regulator as your best choice.
However, what concerms me is that your choice of battery does not appear to include a PCM protection module, I hope you are planning to add one. See the following website for example PCM's
http://www.batteryspace.com/inde...
Also see
http://www.sparkfun.com/commerce...
This battery inlcudes a PCM it is a bit bigger but also offers larger capacity.

Regards
Dez Ellis

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well, if you integrate the entire circuit with some kind of protection
circuitry (which could be a charging device that reliably stops when
reaching the terminal voltage), I think you don't need a separate
protection device. This is only really mandatory if the battery unit
can be taken away, and in particular if it can be charged outside the
final device.

Jörg Wunsch

Please don't send me PMs, use email if you want to approach me personally.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you want charge and monitor li-ion. you should measure how much power you put in your battery and how much you take it. I use Nokia 3310 li-ion battery and max1555 charger IC. USB-UART chip CP2102 has 3V3/150mA LDO so i use it. Current measurement is done by program. I charge battery via USB.

Hope, there is a better/easier way to monitor battery charging.

Here is some pics for my battery powered MEMS pedometer/datalogger project.

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

My charger it's also a MAX1555. It works great! It stops charging automatically, it's a nice IC. I don't have a baterry monitor, but I will include one, and my idea is to shutdown the regulator when the Battery voltage it's critical (near 3.2V i think).

I'm looking at some Linear Technology ICs which are step-up/step-down switching regulators prepared for 1cell Li-Ion batt. May be that will help a bit to get a stable 3.3V power supply.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you shut down around 3 V anyway, I don't think the additional overhead
of the buck/boost converter is worth the while. After all, it's going to
waste energy most of the time.

Jörg Wunsch

Please don't send me PMs, use email if you want to approach me personally.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

SMPS tend to add some noise, that is not a good friend for accelerometers and theyr ADC readings... And probably is not worth, as Jörg stated. If your load to the batt is low, then the critical battery voltage tends to increase a little, since batt voltage decreases with higher loads. Since I cut out power at about 3V3 with my curren circuitry, and I measured about the real consumption, it seemed that I can extract a little more than 80% from the battery (550mA@, up to 30mA occasionally, usually peaks about 16mA every minute or so, but can't remember the mean current consumption). So I concluded that for the extra overhead that would give me the DC/DC converter, the noise plus cost introduced don't compensate.

Since my battery is fixed, I also don't have 'protection circuitry', except a simple 1206 fuse.

Guillem.
"Common sense is the least common of the senses" Anonymous.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Guillem Planisi wrote:
SMPS tend to add some noise, that is not a good friend for accelerometers and their ADC readings...

Every time I think SMPS, I think about SMPS followed by linear lowdrop. So far nothing filtered better than that. Of course you get 80% efficiency from SMPS, suppose it outputs 5.5V, then if you got linear lowdrop to 5V, you lose 0.5/5.5 % of power (plus quiescent current). However you are sure you got your desired voltage, and you can use internal RC oscillator (which would vary a lot with voltage) instead of external quartz, that in turn limits RF noise emmited by the circuit locally (near uC), but you are left wit SMPS somewhere close the the battery itself.

I dont know the application, but I think accelerometers deserve good external voltage ref like REF02 or newer ADR02.
This is because any linear voltage reg drifts a lot with temperature, while accelerometer-assisted robots tend to fly high in cold wind, while being heated in crimped space by powerful motors and ESC.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

dezellis wrote:

However, what concerms me is that your choice of battery does not appear to include a PCM protection module, I hope you are planning to add one. {...}
This battery {that} inlcudes a PCM it is a bit bigger but also offers larger capacity.

I think this is purely about specs and the statistical quality of chemistry. They suppose that with PCM you can unload LiIon or LiPO to 3V (2.8V?) per cell every day, and that it will not wear out so quickly etc etc so in the end they claim you will recover every time 80% of charged energy, while for cheaper product you might recover 80% but it is not recommended without playing with dangerous safety margins. There are no reason why Good Quality cells should need protection electronics in order to have purely high efficiency - I know that because in RC Aeromodelling where weight matters, all high-end LIPOs are unprotected (if you short - it smokes and often bursts, so what :twisted: ).

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Take a look here for a schematic from Peret and a description of my solution:

https://www.avrfreaks.net/index.p...

Ingo