USB charger from AA batteries.

[Alana]

I made something like this for a person who wanted to have 100 of them on as low cost as possible. Design specifications were - to charge a smartphone out of set of AA batteries - sort of emergency charger. From what i managed to measure i need at least 4.85V on the charger output with 0.4-0.5A current capability.
Circuit works but i know it can be improved. At the time i lacked time and resources to do it, now i have them and i want to make it work.

Original circuit schematic in attachment.

What i thought of is to change caps from 10uf SMD ceramics to pair of 100-200uF low esr electrolytic + 100nF ceramic - both trough-hole and if that is not enough to power it from 4x AA battery so that when batteries are fully charged converter would be switched off, but when battery voltage drops it would switch on to keep above 5V. The only problem with this idea is that new AA batteries can have up to 1.7V per battery which is bit too much for a smartphone to take safely.

Any tips?
I tried few other chips but this is the only one that worked and was available at the time.

[Strada916]

Most phone chargers are rated at 5.5V @ 1A. So I would just place a diode in series of 4 AA and hey presto. Infact I have done this my self for my Samsung S3. Plug it in and it starts charging. Only just did this the other day. So I have not tested it fully but I could not see why it would not work.

[SeanB]

3 AA cells and it will run almost all the time. One less cell and slightly lower energy available but will still charge most phones and USB devices. Add a 1000uF capacitor across the battery terminals ( before the switch) to provide bulk decoupling for high internal resistance cells and a reverse voltage diode to cover reverse insertion.

[mariush]

Have a look at ISL9111A :

http://www.digikey.com/product-detail/en/ISL9111AEH50Z-T7A/ISL9111AEH50Z-T7ACT-ND/2816339
http://www.digikey.com/scripts/DkSearch/dksus.dll?WT.z_header=search_go&lang=en&keywords=isl9111&x=0&y=0&cur=USD

http://www.intersil.com/content/dam/Intersil/documents/fn76/fn7602.pdf


It can start below 0.75-0.85v and it might run as low as 0.5v (but then your batteries will be kinda dead already).

According to datasheet, with around 3v at input, it does 5v with over 90% efficiency for up to about 250mA, and it can do up to about 400-500mA but with lower efficiency.

You could just change the 4 battery holder to have 2 pairs of batteries , each pair in series, both pairs in parallel ... so you'd have about 2.4-3.5v  x 2 capacity.
Or.. I don't know.. even run all four in parallel, but then it seems it can only do about 100-150mA (which should be enough to keep the phone running, but I don't think it would charge much)

[Alana]

Thanks.
First try - 3x AA + 1000uf cap and circuit i already designed - i have few of those chips at home so i may try it out soon.
I understand that on those kind of circuits i should keep high current loop as short as possible and as thick as possible.  Device according to datasheet works on about 280kHz. What caps at the output? 100uf + 100nf is ok here?

[Alana]

Ok, board ready and assembled.
Checked it out with bench PSU and:
input voltage 2.5-2.8V can give stable 4.91V on output at 0.5A
with low input voltages efficiency 60%, input current 1.6-1.8A
with higher input voltages in range 3.5-4.5V efficiency rises to 75%
output voltage seems to be stable

i modified schematic to allow myself to use 4x AA batteries - with new batteries D1 and D3 provide enough voltage drop to get in safe range for USB chargers - lower than 5.5V

todolist:
- test it out with mobile phone - have to get usb socket
- test real operating voltage ranges - i have crappy home made PSU with single turn pot for Vout and this does not give enough precision
- test it out with regulated constant current load - to see how much amperage i can squeeze out of it

Schematics, layout, pics of assembled device - as attachments.

[amyk]

Every time I see these things and contemplate the fact that it's a battery-powered battery charger, I'm reminded of this:

[brainwash]

There is a reason for these things. It's mainly for camping, going to forgotten parts of the world or just plain forgetfulness.

I had one of these things marked 'iPod charger' that I've never used. Just before embarking on a big trip (where sockets were at a premium and AA batteries for free) I decided to stress-test it. Well, one second of short-circuit was enough to fry the converter. I did this AFTER taking the time to mount the standard USB socket.
Attached are photos the PCB. This thing cost me around 5$ or so and it had an intricate mechanical design that allowed the user to set the thickness of the iPod so that it stayed flush and secure to the back of the unit.

I also experimented with modifying an existing car USB supply to get it to convert 1-3V to 5V. It was based on the ubiquitous MC3423 SMPS IC and used the reference schematic. I just had to reverse the diode, remove a leg from the inductor and another wiring mod. This worked in theory but efficiency was 25-40% and the inductor would get too hot to touch after spending a few minutes supplying 0.5A. It's just an idea for people to try out.

[plunger]

This girl made millions off of this little thing.

http://www.adafruit.com/products/14



$20 is a little pricey for what you're getting.

[Alana]

There is a reason for these things. It's mainly for camping, going to forgotten parts of the world or just plain forgetfulness.

That was the reason for the initial 100 - they were to be sold at a 3-day concert event.

This girl made millions off of this little thing.
$20 is a little pricey for what you're getting.

How much would you pay for such unit? And how much you think it cost? Mine is no more than 5 euros in parts.

[Analogtech]

That looks like a product I would buy after a few mods.very good job nonetheless also I don't now if this is really required,but you may want to add a small capacitor at the output just before the usb port.

[IanB]

$20 is a little pricey for what you're getting.

I think you're getting a tested, proven design that's had a lot of research put into it.

Suppose you're not a big earner and you make $25/hr. Could you design, prototype, test and debug something like that in less than an hour to make it worth your while to make it yourself?

[Alana]

Suppose you're not a big earner and you make $25/hr. Could you design, prototype, test and debug something like that in less than an hour to make it worth your while to make it yourself?

Do you count "fun" and "expirience" in "gains" column as well? Do you always count time spent on DIY in "expenses" column?

[plunger]

I think you're getting a tested, proven design that's had a lot of research put into it.

Suppose you're not a big earner and you make $25/hr. Could you design, prototype, test and debug something like that in less than an hour to make it worth your while to make it yourself?

I wouldn't be able to make it from scratch. But I do know an overpriced item when I see one 

This is how she is able to maintain an office/factory/warehouse in downtown NY. More power to her!