5V supply help needed

Hey all, I'm having a hard time actually finding what I need here. I know it exist but I must be searching for the wrong terms. I have an arduino project that runs off of a 5V/1A supply. I wanted to make it portable/chargeable. These are the two options I'm going for:

  1. swappable batteries with one secondary inside to allow continued operation during battery change.
    OR
  2. all internal with an external charge port, like cell phones, laptops etc, allowing use while charging.

The project is planning on using (4) 3.7V 2600mAh batteries, hooked up as to allow for roughly 10AH at 5V. So far, my plans are to hook up 2 batteries in series, making 7.4V, then make a parallel circuit with those, doubling the mAh. I was going to use a simple vreg to reduce the 7.4 down to 5V. My question is can I build a simple charging circuit inline? Any advice?

Need more parameters. (1) How much current does the circuit actually use, and (2) How long does it need to run on battery power before recharging is needed?

My apologies, To answer those two questions, I'm not 100% on the current draw but if we follow USB standards (my devices are all usb), then 5x 500mA max, so 2.5, I would venture to say 1.5 as my real max. I'm currently using a anker astro 3 10AH battery pack. As far as runtime, I would like at least 6 hours unplugged if possible

I figured the max current would be something less than 1A if it's running on a 5V/1A supply now. So that's USB, not a wall wart or something?

For starters, I'd definitely measure/know the current requirements. Linear voltage regulators do nothing but waste power, so not good for battery powered circuits. You mentioned 3.7V, I assume that's LiPo. I might look to a circuit like this which can charge the battery and power the load simultaneously, or run on just the battery. I'd then follow that circuit with a switching boost regulator to take the 3.7V up to the required 5V. That will be a lot more efficient than a simple (linear?) regulator to drop 7.4V to 5V.

Just select a 3.7 battery that has a mAh capacity compatible with the load and run time requirements.

cool thanks, yea right now I'm testing it with this anker, it's actually 2A rated, but I think its 1A dedicated per usb port. I'm going to chop a cable now and make a current tester :D. I WAS planning on a linear vreg, I for some reason thought they were more efficient to drop down than to use a boost circuit, thanks for correcting me. And yes, I was looking at li-ion/lipo

so, this is what I for see as teh following steps:

measure current draw
get appropriate battery
use the recommended circuit to charge/use

Sounds good. The switching supplies are much more efficient whether they are boost- or buck-type (i.e. lowering voltage). Efficiencies of 90% or higher are common.

Edit: Do you know how long the Anker 10Ah will run the circuit?

man, I gutted the anker, the put white glue on top of almost every chip, only thing I know for certain is that they used 4x 3.7 2.6Ah samsung li-ion batteries. one thing I forgot to add in, I'll need to include some sort of power off switch, I was trying to find out how anker did theirs. if the current draw drops below 100mA (tuneable would be nice), then the battery is switched off, that way, all I need is a SPST-mom to turn on the battery draw.

for example, with the anker: Press power button, as long as devices start drawing power, stays on, when unplugged or drawing less than 100mA, batteries switch off. I want this too if its not too complicated.

I don't really know anything about the Ankers. Could you tell if the four batteries were connected in parallel? Sounds like it does everything you want, so why not just use it? It's pretty much gotta be a fancy version of what I described earlier, probably with some intelligence built into it to do the soft switch etc.

a few reasons for not using it, 1) its expensive, 2) It's third party, I want to be the builder of all the internals of this project. and the batteries are in parallel in two series sets.

Hmmm, well I'm not that up on LiPo battery prices, but assuming similar capacity is required, I wonder if just the batteries wouldn't rival the cost of the Anker. But I get the third-party concern. Of course if the plan is to build dozens, hundreds, etc., then lower prices could be had. But to build in some of the other bells and whistles, low current cutoff, etc., will involve extra engineering effort. How complex did the Anker circuit seem to be?

But we're still working in the dark, without knowing the actual current requirements, it's hard to know how much battery is needed.

well, i tried just a bit ago to measure current on the devices, the usb ones weren't picking up for some reason, the simple two pins were though. I did the simple cut the power wire in the cable trick.

calling it here for the night. I tried to figure out teh three chips on the top side of the anker, two of them at TI LM358 power amplifiers, the last one is a 20 pin microcontroller stamped ABOV 81f4204w. I've searched my ass off for the last 3 hours, I found it listed as mc81f4204w, but i still have no idea what it actually is, there is also a 5pin header directly beneath it, i'm assuming for programming it. If anyone can help me out, i'd really appreciate it

Just wanted a general feel for the circuit complexity, didn't mean for you to make a career out of reverse engineering it. LM358s are op amps, maybe doing voltage or current sensing of the battery or load. Not familiar with that MCU but I figured there'd probably be one at least. Wouldn't be surprised if there were a dedicated LiPo charge controller and then regulators for the output too. Did you get a current reading on your circuit?

no readings yet, I'm going to have to find a better way to read current, either the cable was bad or something was up with my meter, but there wasn't enough current to power teh device with the meter in series

Your meter is defective... possibly a blown fuse inside it, or you aren't measuring the current properly.
an ammeter should be able to pass current (Very Low resistance internally) If it doesn't it is bad or
you are using the meter improperly. Charging Li-Ion or Li-Po batteries in series-parallel isn't a really great Idea as the
small differences between cells can cause one or more batteries to not receive a full charge.
There is no way to balance or spread the charge current equally between cells.

Bob

must not be using the meter properly, but i doubt it, it worked fine on testing a 5V fan, but not usb. I'll try again, I'm going to cut a new cable, should eliminate potential error

ok, so I tested most of my parts, used datasheets for the rest, if everything pulls peak power its close to 1950mA, but not everything will be at peak at the same time, if ever, closer "run values" are more like 1400, if the main processor stays running full steam and everything else settles.

Sounding like a pretty good size battery. I'd design for at least 2A and an 8-hour runtime, to give some headroom and allow for battery aging. Total energy the battery needs to deliver to the 5V load is 80Wh. Assuming a 3.7V LiPo, that'd be about 21.6Ah. Then figuring 90% efficiency for a boost converter to bring it up to 5V and we're at just slightly over 24Ah. Time to start studying datasheets for batteries, charge controllers, and boost controllers/regulators.

Thanks for adding that up for me. I'll start digging more in depth, I'll post what I find.

So I'm thinking in the first wave of searching, 8 of these should meet those requirements. Hooked up in parallel should give me 24AH.

I'm still looking for smaller or flatter, but those are a good first round pick i think.