4AA efficiency
Moderators: adafruit_support_bill, adafruit

4AA efficiency

by af3556 on Sat Jul 11, 2009 3:37 am

I've installed a MB into a neat little 4xAA battery case available in Australia from Jaycar (part MP3083). This case seems to be intended for exactly the purpose of powering USB stuff from 4 AA batteries - it contains a PCB-mounted USB socket, switch and reverse protection diode (and bonus - a clip on the back for hipsters to attach it to their belts alongside the phone :-). I removed this board and had to perform a little plastic surgery on the case to fit the MB, and had to lay over the inductor and caps to fit the reduced height. "When I have time" I'd love to do a PCB that is a drop-in replacement.

I specifically wanted 4xAAs to maximise the energy available for weekends away with the iPhone - and further, from the LT1302 data sheet the efficiency seems to improve with greater input voltage, though the sheet only shows Vin @ 2, 2.5 and 3 V. Given the device's Vin max of 8 V, I figured it shouldn't hurt to feed it 4 x AAs (5.4 V for fresh rechargeables, 6V for alkalines) and see what happens. Below are the results with the MB fed from a variable power supply, loaded with a 10 Ohm resistor (i.e. nominal 500 mA load):

Code: Select all | TOGGLE FULL SIZE
Vin   Iin    Pin     Vout    Iout    Pout    Efficiency
2.4   1.31   3.14    4.53    0.43    1.95    62%
3.0   1.05   3.15    4.79    0.45    2.16    68%
3.6   0.80   2.88    4.82    0.46    2.22    77%
4.2   0.66   2.77    4.91    0.47    2.31    83%
4.8   0.60   2.88    5.05    0.48    2.42    84%
5.4   0.53   2.86    5.16    0.49    2.53    88%
6.0   0.51   3.06    5.51    0.52    2.87    94%

NOTEs:
* Vin/Iin were measured to only one/two decimal places (PSU meter), so all of the above should have nice fat error bars applied
* further errors would have been due to me using alligator clips to hook up my little test rig - not the best way to get a nice low-resistance connection (notice the output current is not quite the expected I = Vout / 10 Ohm)
* the MB stayed cool, warming slightly at the lower input voltages
* output voltage regulation was good, within 10% across the board, under 5% for the rechargeable input voltage range of 4.0 (flat) to 5.4V (charged)
* I measured significantly lower efficiency than the datasheet's claims, at ~500 mA: cf. 62% actual vs. ~82%
* one thing's quite clear: the circuit's efficiency improves with higher voltage, up to at least 6.0 V

Conclusion: 4xAAs are the way to go for peak efficiency. The current drawn from 4 AAs is less than half 2 AAs; indeed the cells should last more than twice as long as the effective capacity of most battery types is usually higher for lower discharge currents (e.g. Fig 5.2.1).

Rgds,
Ben

Image
Image
Image
af3556
 
Posts: 17
Joined: Sat Jul 11, 2009 2:46 am

Re: 4AA efficiency

by rseni on Sun Jul 12, 2009 5:15 pm

I'd so much want to do this but won't this fry the batteries? As in this thread -

viewtopic.php?f=15&t=7620

Read the last post.

I'd still want to try this out tho I do not have a case like yours. I do have a 4 cell charger case which is useless. I'd add 2 - 4 LEDs too to use it as a flashlight too when needed.

RS
rseni
 
Posts: 6
Joined: Sun Jul 12, 2009 5:09 pm

Re: 4AA efficiency

by af3556 on Sun Jul 12, 2009 7:32 pm

rseni wrote:I'd so much want to do this but won't this fry the batteries?


No, it won't fry anything. The LT1302 has an absolute maximum input voltage of 10V (8V operating, p2. of the datasheet). And as you can see from my measurements, while the output voltage starts creeping up as Vin goes higher, it's still only 5.5V out at 6V input - above the USB spec of 5.25 V, but unlikely to damage anything. Having said that, I'd stick with rechargeable cells for this (5.6 V maximum with freshly charged cells, gives ~5.2V out; nicely in-spec).

BTW, most of the posts in that thread suggest there is a fault with the MB's construction - there's no way the LT1302 should get hot, let alone too hot to touch. With a full load output power of 2.5 W (0.5 A @ 5V), even at the worst case of say 60% efficiency the whole circuit should dissipate about 1.5 W, which would only be warm to the touch. And as you can see from my measurements, with 4 AAs the circuit is ~80-90% efficient, or approx. 1/4 W wasted in the MB - you'd be hard pressed to feel that at all.
af3556
 
Posts: 17
Joined: Sat Jul 11, 2009 2:46 am

Re: 4AA efficiency

by rseni on Mon Jul 13, 2009 4:06 am

Thanks a lot for the detailed explanation. I'll post back after I've made it.
rseni
 
Posts: 6
Joined: Sun Jul 12, 2009 5:09 pm

Re: 4AA efficiency

by af3556 on Mon Jul 13, 2009 4:34 am

I've observed the performance of the 4AA-powered MB a bit more: one set of fully-charged 2500 mAH rated Energiser NiMH cells successfully charged my 3G iPhone one and a half times. i.e. I ran the iPhone till the battery indicator went red, then plugged in the 4xAA MB until it stopped charging (reached full charge). I then left the MB alone (i.e. didn't recharge the AAs) and the next day when the phone ran low again (red), plugged it back in. It got about half-charged when it dropped out of charging (i.e. MB couldn't supply the grunt anymore). I then discovered that one of my AA pack's cells is dud, it had dropped to under 0.3 V under load (the other's were still ~1.1 - 1.2 V). I've bought some new 2000 mAH Sanyo Enerloop cells, and am testing them as I write.

I think it could go for two full cycles: 4 x (1.2 V x 2500 mAH = 3 WH) = 12 Watt.Hrs available power in the pack, compared to the phone's 3.6 V * 1150 mAH = 4 WH; even allowing for some hefty inefficiencies there should be enough power there for just about two complete charges.

4AA FTW!

Another thing: in testing over the weekend, I found I had a poor connection on the spring terminals in the battery holder, where the springs are pressed on to the metal tabs that hold them. This seemed to be causing the MB to dropout when under highest load (a few minutes after the phone starts charging, presumably when it's in some kind of fast-charge mode) - I soldered the springs to the metal tabs and haven't had even a hiccup since.

Rgds,
Ben
af3556
 
Posts: 17
Joined: Sat Jul 11, 2009 2:46 am

Re: 4AA efficiency

by rseni on Sat Jul 18, 2009 1:29 am

It's difficult to get the LT1302 here in India and the inductors come in milliHenry here. The store people say they don't come in microHenry.
rseni
 
Posts: 6
Joined: Sun Jul 12, 2009 5:09 pm

Re: 4AA efficiency

by rseni on Sat Jul 18, 2009 3:31 am

I am using a cassette box as the case and it can hold 4 cells too. I have 2 USB female connectors like the ones at the back of P-III CPUs along with the cable which I he screwed to the cassette box. (Idea stolen from "$2 Cassette Box Cellphone/USB Charger" on instructables.com)

Since there is ample space even after putting the Mintyboost circuit, I decided to have a small circuit which could charge the Ni-MH cells within the box without removal. Can anyone suggest one small and simple circuit for the same please?
rseni
 
Posts: 6
Joined: Sun Jul 12, 2009 5:09 pm

Re: 4AA efficiency

by rseni on Sat Jul 18, 2009 3:42 am

I'm replying my own question here. I got a circuit thru google here:

http://electronics-diy.com/electronic_s ... php?id=781

Only thing is I'll need two such circuits for 4 cells and maybe some DPDT switch or else two extra USB connectors to charge each set of two cells. That becomes 3 USB connectors total.
rseni
 
Posts: 6
Joined: Sun Jul 12, 2009 5:09 pm

Re: 4AA efficiency

by edalquist on Sat Dec 26, 2009 1:01 pm

So it looks like the MP3083 already comes with some charging circuitry built in, did you remove it and replace it with the MB circuit? If so why, does the MB provide better efficiencies that the provided circuit?
edalquist
 
Posts: 2
Joined: Sat Dec 26, 2009 12:57 pm

Re: 4AA efficiency

by af3556 on Sat Dec 26, 2009 6:09 pm

So it looks like the MP3083 already comes with some charging circuitry built in ...


The Jaycar MP3083 case does incorporate a small PCB with on/off switch, reverse-protection diode and USB socket, but that's it. It has no regulation, nor charging circuitry - no real circuit at all. I reckon the only reason it has a PCB at all was to have something to mount the USB socket to. The case's PCB is not useful, I removed it, scavenged the USB socket and switch for spares, and discarded it.

You could possibly use just the MP3083 case with some USB-powered devices, but those devices would need to tolerate the wide voltage range from fully-charged (~6V) to flat (~4V) batteries. Few devices can do this - certainly not the iPhone.
af3556
 
Posts: 17
Joined: Sat Jul 11, 2009 2:46 am

Re: 4AA efficiency

by fat16lib on Sat Dec 26, 2009 7:12 pm

The main reason for the diode in the Jaycar MP3083 is to reduce the voltage from 6 V to about 5.3 V so it won't harm USB devices, not polarity protection.
fat16lib
 
Posts: 593
Joined: Wed Dec 24, 2008 12:54 pm

Re: 4AA efficiency

by edalquist on Sun Dec 27, 2009 1:56 am

Great, thanks for the clarification. I have a iPhone 3GS that I'm going to give this a try with using 4 NiMH batteries with. Thanks for the details in the above posts.
edalquist
 
Posts: 2
Joined: Sat Dec 26, 2009 12:57 pm

Re: 4AA efficiency

by af3556 on Sun Dec 27, 2009 5:09 am

fat16lib wrote:The main reason for the diode in the Jaycar MP3083 is to reduce the voltage from 6 V to about 5.3 V...


Yeah, you're probably right.

Incidentally, I've been using my minty boost with the 2000 mAH Sanyo Enerloop cells - works very well, I get almost two full charges of my iPhone 3G which gets me over a weekend with moderate use. Not bad, and though I look forward to the day a smart phone can actually go for more than a day without a recharge, the minty boost does the job in the meantime.
af3556
 
Posts: 17
Joined: Sat Jul 11, 2009 2:46 am

Re: 4AA efficiency

by 18scoobz on Wed Jun 15, 2011 9:46 am

I noticed that the freshly charged 1.2v rechargeable batteries output 5.6 volts. My understanding was that the minty boost required input voltage less than 5v. So from what you are saying I can use the 4 batteries and the 5.6v input won't damage anything?
18scoobz
 
Posts: 1
Joined: Fri May 13, 2011 8:25 am