voltage regulator for battery powered clone?

Can you please suggest a voltage regulator that is efficient for battery power of an arduino clone of my own design that I make?

I am planning to make an arduino clone for a toy project of mine that is based on the BBB and the single side board but I read that the 7805 is not efficient for battery operation...

Thanks Harry

what voltage are you trying to input? and how much current do you need?

I've used the TS2940. A quite affordable low drop regulator with 0.6V drop at 1A. For small currents the voltage drop is below 100mV.

Thank you for the replies

I am thinking to use AA NiMh batteries so 4 or 6 of them...

Is there an application note where I can learn about voltage regulators and the arduino or the atmega328? I searched the atmel site with no luck ...

Just read a datasheet of TS2940... and I see a problem: minimum current for normal operation is 10mA! I got my arduino to use as little as 200uA so it's not the best solution. Any other ideas?


If you've got a lot of extra voltage, you could just stick with the 7805..I don't know if it has a high minimum current, but it's voltage drop is steep.

Are you looking for a TH or SMD part?

All of the linear regulators I have seen that are in TH packages have dropout voltages in the 1V range. Unless your batteries have a flat discharge curve (which I do not believe NiMh do) then you will need a large input voltage to compensate for the discharge. As you increase the input voltage the efficiency drops.

If you do a 3.3V voltage system and use a 3.7V Li-ion battery (which has a flat discharge curve) with an SMD linear regulator you will be in the 80-90% efficiency range.

The most efficient conversion would be a switcher regulator. This is a bit more complicated and usually involves SMD parts.

(* jcl *)

www: http://www.wiblocks.com twitter: http://twitter.com/wiblocks blog: http://luciani.org

I am looking for TH On what basis do you decide whether a project needs a regulator?

I am going to power my toyproject with batteries...is it necessary to have a regulator?

On what basis do you decide whether a project needs a regulator?

I look at the currents -- steady-state and pulsed. I would also look at that size, cost and power constraints. Unless there is a severe cost, size or efficiency constraint I would use a regulator. More than one (at the same voltage level) is usually not necessary.

I am going to power my toyproject with batteries...is it necessary to have a regulator?

Maybe not. You need to determine the highest and lowest battery voltage the system will see. For the uC you need to determine the proper operating frequency. The A/D range will change if you use AVcc as reference. The output levels of the digital I/O will also change. You need to verify that the different devices you are driving will operate over a wide range of voltage.

If your battery voltage can drop too low you will either need a step-up converter or more batteries.

I don't recall if you mentioned the voltage/current requirements or the battery configuration you were looking at. I would start with a spreadsheet and list the various loads and then start to plug in some numbers.

(* jcl *)

If you have batteries the answer is to go for a switching regulator. However these are much more tricky and expensive than "normal" regulators.


If you have batteries the answer is to go for a switching regulator.

Not necessarily. If you run at 3.3V and use a Li-ion (which is 3.7V) you can use a LDO. You will be at 85% efficiency but you will most likely need an SMD LDO. I haven't seen any TH devices that will work with a few hundred mV of dropout.

The flat discharge curve and the small difference between the battery voltage and the regulated voltage make the Li-ion a nice solution.

(* jcl *)

If the batteries provide 3.7V it would be possible to run the Arduino without regulator at all. However things are different if there are other devices that actually require 3.3V.


Keep in mind that a Li-ion cell does not have a constant output voltage. Rather it varies with it’s state of charge ranging from 4.2vdc at full charged down to whatever safety cut-off voltage you implement (most recommend to stop discharging at 3.5vdc and cell damage can occur if allowed to discharge below 3.0vdc. Proper usage of Li cells can be quite complex and has severe safety issues if not observed.


I have used both LP38690 (http://search.digikey.com/scripts/DkSearch/dksus.dll?WT.z_cid=sp_014_0310_buynow&site=us&lang=en&%20?Keywords=LP38690DT-1.8&vendor=14) and TL720M05 (http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=296-24509-2-ND) before. These are LDO regulators and work with 4AA NiMh quite well.

Thank you for your replies!

After a lot of search I found the



is it a good choice for a battery operated perfboard arduino clone?

The only downside to the MCP1825 is the input voltage range. The absolute maximum input voltage is 6.5V and the recommended maximum input voltage is 6V.

Other than that the MCP1825 looks like an excellent part. The dropout is very low (350mV max at 500mA), the Iq is low and it is stable with low values of ceramic capacitors.

This would be a great part for a 3.3V system battery powered system since you are unlikely to get close to the 6V max. Might be a little close for a 5V system but could work.

(* jcl *)

Given that its absolute maximum input voltage is only 6V, you'll have to think about what kind of batteries to use.


4x AA cells 1.5V will give you 6V. If the cells are new even a little bit more, maybe too much. With a voltage drop of 0.2V in the regulator, you have about 0.8V to work with. Split up to 4 batteries gives 0.2V per battery! If the cells drop to 1.3V each (still quite full), there's nothing left to regulate. And you can't use more cells to compensate... or use it with a 9V cell (bad choice anyway) or car battery.

Personally I wouldn't use it. I like boost converters like the MAX756 or LT1302. Efficiency is above 85% and they can suck the batteries dry to below 0.7V (MAX756). Of course they're not exactly cheap, roughly about 10$ for all parts.

I'd get a regulator with a wider range of input voltages, unless you have the perfect 6.0V battery with a flat discharge curve at hand.

Edit: seems like it took me half an hour to write this post :-)

Thank you for your replies madworm and jluciani :)

I am still a bit confused as I am new to electronics ...a couple of possibly naive questions:

''With a voltage drop of 0.2V in the regulator, you have about 0.8V to work with''

what will happen if the voltage drops below the 0.8V (The battery cells of the example drop to let say to 1.1v) ?? The LDO will not work or will not produce the fixed 5v? If I find a LDO with higher max Input specification does it solve the problem?

I have read to a white paper about LDOs that swiching supplies produce more output noice and have slower responce to input/output transients... For an application with a perfboard arduino that need accurate ADCs are these characteristis of the boost converter a real concern?

When your Vin < Vout + Vdropout the linear regulator is no longer
linear it is saturated. The output voltage will track the the input voltage.
Transient response is reduced.

With proper layout the switching noise will not be an issue for an 8-10bit ADC
like the internal ADC on the Atmega. If you are new to electronics you may
want to go with the linear solution (or a purchased switcher). If you decide
you want to learn about the converters you may want to look at lower
frequency devices. The MAX756 and LT1302 that madworm suggests would
be good choices. IIRC the LT1302 is used in the Ladyada Minty Boost 2 and a MAX856 is used in the Minty Boost 1.

(* jcl *)

I will follow your advice and test the boost circuits along with LDOs

Just to be on the safe side what is your opinion for the following LDO (again I want to use it with batteries power supply) mic5205 http://www.micrel.com/page.do?page=/product-info/products/mic5205.shtml

I think I now follow the specifications madworm and jluciani and gave me ;)