# Can I run an Arduino off 8 x AA batteries?

Can I run an Arduino off 8 x AA batteries? The documentation says it can run from 5-12v which is in the limits of the the specs. I wanted to ensure I had more than enough power for anything else I wanted to drive through it. Would it be better to run the Arduino off a 5v regulator from the 8 battery supply then draw anything else off the power supply directly?

-
What regulator IC does your Arduino use? Can you post a link? – Russell McMahon Dec 29 '11 at 20:46
The UNO uses the MC33269D-5.0 which has a maximum $V_{in}$ of 20V. – Majenko Dec 29 '11 at 21:23
@Mech Software - Can you please confirm: / The requlator used / the Arduino you are using and preferably links to both, but MC33269D-5.0 data sheet here . | Note that if the regulator is 5 volts out then it needs 1.25V headroom at 500 mA and 1.35V at 800 ma. So Vin min = 5V + 1.35V = 6.35V.6 cells are getting marginal. 8 are OK. – Russell McMahon Dec 29 '11 at 22:59

The first thing the power jack goes to inside the Arduino is a 5V regulator.

This can (for the UNO at any rate) supply a maximum of 800mA.

Any more than that and you will have to use the external power.

How much current the AA batteries can supply depends on the chemistry.

But yes, you can run (and many people do) off 8 AA batteries.

-
Thank you much. – Mech Software Dec 29 '11 at 20:02

What regulator IC does your Arduino use?

Yes, you can PROBABLY run you Arduino from 8 x AA batteries.
There is a small risk that brand new cells will cause problems - see below.

If you care about battery life you can get a useful gain in lifetime using a buck regulator to step the voltage down. If your Arduino regulator will really accept 5V then you may get 50%+ xtra battery life by using a buck converter to convert battery voltage to 5V

IF the Vin min spec is really 5V then you would be better off using 6 x AA.
This allows operation per battery down to 5V/6 = 0.8333 V. That is essentially flat for all cells. An alklaine cell has a very tiny % of total energy left.

If you only use a linear regulator then using 6 cells instead of 8 has very little downside and you need 6/8 as much battery capacity.

If the spec is 5 - 12 V in then you will exceed that voltage using new AA Alkaline batteries !!!
That is PROBABLY OK but you want to check the Arduino's 5V regulator spec.
Note that some regulators have a low Vio max (12V or less in some cases) and exceeding it even slightly MAY cause damage.

An Alkaline battery will typically provide 1.55V when new.
8 x 1.55 = 12.4 Volt.
The 0.4 Volt seems like a trivial amount to worry about, and it probably is unimportant - but in electronics you should always be wary of exceeding data sheet rating sby even the very smallest amount.
It MAY be OK to do so.
You MAY choose to do so.
But you MUST check and know what you are doing and why and what the implications may be.
Failure to do this will mean you end up with "designs" like those that most people produce :-).

Worst case an Alkaline cell can produce 1.6V or slightly more.
1.6 x 8 = 12.48
Not a lot different.

A Nickel Zinc cell produces 1.6% V or more when new.
1.65 x 8 = 13.2 V
Definitely getting into the "better check the requlator" territory.

8 x 0.8V = 6.4V so lots of headroom.

NimH or NiCd should not be discharged under 1.0V and have a max voltage just after charging of about 1.3V + - so inside Alkaline range at both ends.

It has been suggested by @Majenko that the regulator is a MC33269D-5.0

If so, the regulator outputs 5V and needs a MINIMUM of 6.35V to be guaranteed to operate at 800 mA under all temerature conditions. Slightly ove 6V in may be OK in most cases.

If this is the regulator used it is 5 volts out an needs 1.25V headroom worst case at 500 mA and 1.35V at 800 ma.

So Vin min = 5V + 1.35V = 6.35V.
6 cells would need to make 6.35/6 = 1.06V each to operate this worst case.
In practice it will usually be able to be somewhat lover than this.

So, 6 cells are getting marginal BUT you get reasonable but not complete utilisation of cell energy.

8 cells will be OK.
Vin max is 20V so no problem there.

If this IS the regulator IC then the spec should be about 6.6V - 20V.

For higher Vin's energy losses get severe. A buck regulator or similar would make much sense.

Energy loss as heat = (Vsupply - Vout) x Iout.
At 800 mA and say 10V in and 5V out
energy loss = (10-5)*0.8 = 4 Watts.
A substantial heatsink would be needed on the Arduino regulator.

-
We haven't considered heat. I've hooked up an Arduino to 12V with a Zigbee and ethernet module on it. I can tell you, the regulator is getting very hot (too hot to touch). 'Very hot' is not something I actually measured, but it may be that 12V can damage the regulator if it's operated at high loads. – Hans Dec 29 '11 at 20:54
@Hans - The regulator on most Arduinos has thermal overload protection, so you won't damage the regulator or circuit. It could cause the device to shut down, which might be problematic, but it won't burn anything up. – Kevin Vermeer Dec 29 '11 at 21:19
It's not that we haven't considered heat, it's that the user has asked a question that has less data than could be provided. So far requests for extra data have not produced answers. Could be a time zone issue and having to work to make ends meet. At 12V in at max regulator current Pdiss =~ (12-5)*0.8 = 5.6 Watt. A goodly amount. Whether the Arduino has a heat sink to handle anywhere like this I know not. – Russell McMahon Dec 29 '11 at 23:04
I've been told by manufacturers in the past that NiCd can possibly reach 2.0V while charging and NiMH can reach 1.8V while charging. Have not personally seen them go that high (more like 1.45V), but those are what I'd consider the safety limit numbers - have you seen the same thing, Russell? Not so much a consideration if the product does not charge the battery in place, e.g. you have to remove AA cells to a separate charger. – Matt B. Dec 30 '11 at 23:37
A NimH that reaches 1.8V under charge is being grossly abused. At 1C charge they terminate under all typical termination methods (negative delta V, delta Temp, absolute temp) at almost exactly 1.45 V almost always. I have often watched terminal voltage at end of charge on many many charges and never seen any healthy call get as high at 1.5V. (I've been involved with 1 - 2 million NimHs in products the last few years. That's not my main focus but I've played with quote a few along the way. NiCd I can't say as much for BUT see no reason for them to be vastly different. – Russell McMahon Dec 31 '11 at 0:27