Large 12V Power Bank

heya,

let me first say, i am not overly electrically minded, i have done some basic electrical courses in high school (and that was over 10 years ago), but one of my friends is a prepper, and he wants me to make him, a portable power bank, which is using either 18650 3.7v cells for the power storage.

i think i have got all of the basics, 12v dc in, 2x 12v solar in, the batteries in series and parrell to make 14.4v and then the 12v output.

he wants something that will hold alot of power, self charge in a SHTF world, be able to charge radios and mobile phones, laptop etc, i have recommended commerical products, but he said that they dont hold enough power,

the diagram i created is only 14400mah, but i am thinking he wants more like 50000mah or more... lol, not my problem, he is the one paying.

what does the pros think?

cheers
joel

The circuit will need some kind of control circuit to prevent battery damage from overcharging, and diodes to protect the solar cells from back current in darkness.

But Colin is right, all he needs is a car battery with a large solar panel charger. The only advantage for Li-Ion cells is they hold their charge for longer if not used.

lol, i understand what you mean, but i need to see a diagram so i know how to do it.

what do you think of the ups style batteries?

could you please give me a diagram on how to add the diodes for the solar protection, and how to prevent overcharging.

i was going to put in 2 cheap multi meters, one set for volts, and the other set of amps, cause i can get cheap multi meters for like 5$ on fleebay

Diagram attached for solar panel diode protection.
Diode voltage rating: anything higher than battery voltage
Diode current rating: anything higher than expected current in that wire (so dependent on solar panel rating)
A 1N5401 diode is rated at 3A 100V.

Volltmeter for the system: Ebay has cheap 20V self powered panel voltmeters which would be ideal, which could be connected through a push button (to avoid continual battery drain).

Ammeter: to measure battery charge/discharge current - centre zero analog ammeter would be good, as digital ammeters require a separate power supply (cannot be powered from the system they are measuring). So a digital panel ammeter would need its own (small) battery.

Digital multimeters need their own batteries anyway. For convenience in a fixed location this could be external and switched, to avoid repeatedly switching the dial through several positions, which is a small problem, as a meter should be disconnected from the circuit until the correct range is selected.
(But can be left in circuit on the correct range if powered off by a separate switch.)

I would like to make a safety suggestion. Just because I didn't see anyone else mention it. Putting large numbers of small batteries in series gives numerous chances of bad connections. Any faults in a circuit that can discharge lots of amps in a short time needs some protection. You can use a bus arrangement so that you limit the number of batteries in series and use a fuse or other protective device. Once you have all those batteries wired together you could literally weld with them.

so what do you recommend? use extra diodes at the end of every series of cells? so have 4 cells in series, and then a diode at the end of them, before they are connected in parallel?

heya, i have quickly mocked a diagram up in paint (so much easier to use than that program i tried last night)
now how would i put the over charge protection into the circuit?

And I would consider a Schottky diode so less power is lost in the diode. Or better yet, something like one of these:

* http://www.linear.com/product/LTC4358
* http://www.linear.com/solutions/1465

Or something like the attached (V2 is a "sine" wave generator to simulate the solar panel voltage fluctuation. V1 is the battery being charged).

The problem is that we don't know what the load will be. Rule of thumb is to fuse the load at 120% of it's intended load. If you are trying to pull maximum amperage out of the batteries then 120% fuses placed where the diodes are would just let all the power flow in all conditions. Let us consider the line where all the leads from the batteries come together as the BUS and any load that comes off the bus would need that 120% fuse. The bad thing about a bus like this is that the bus is not fused so you want to use a commercially available bus is built to prevent any unwanted contact. Something like this:

http://www.fuseholders.com/fuse-blocks.php

Notice that one side has common contact while the other has the fuses to individual contacts. Make sure the wires connecting the batteries to the bus on the fuse holder are heavy enough to withstand the amps from all the batteries as they are commonly connected also that they are as short as possible. That the bus on the fuse holder is rated for enough amps and that there are enough fuses for each load you intend to run.

I understand wanting to use a Li battery, but there are larger Li batteries that could be used that would simplify this. I think if you compare cost the larger battery would probably be as cost effective as well. Most of the larger battery packs come with a circuit breaker built in and may be available with a state of charge indicator as well. good luck

heya,
i was going to use the samsung 18650 icr18650 which has the protection circuit built into the battery, i use a same battery in my torches, and when i need to rebuilt laptop batteries.
will that sort of protection circuit be enough or would i need something else?
i did look into the 32600 (i think) but they costed more, than 2 of the 18650.
ie 1x 32600 5000mha was about $10
2x 18650 2600mha was $3.70 ea
and the 18650 was samsung, as were the 32600 was a no name battery.

Here's a spec sheet on that battery:
http://www.batteryonestop.com/baotongusa/prod...atasheets/li-ion/Samsung-SDI-ICR18650-26A.pdf

In there it states:

8.2 Short-circuiting

8.2.1 Short-circuit results in very high current which leads to heat generation.



8.2.3 An appropriate circuitry should be employed to protect accidental
short-circuiting.


I see no fuse, circuit breaker, or current limiter mentioned anywhere in that spec sheet. I did see a mention of it in the advert I used to get to the spec sheet. I go by the spec sheet.

The max discharge rate is listed at 5200 milliamp so you are looking at a 20.8 amps per 4 cell stack and 83.2 amps anywhere the stacks are commonly connected. The wire required to safely carry that is #2 AWG Cu (huge). It would really pay to look at the loads and the charger and install a fuse per battery stack that is just large enough to allow normal charge and discharge. It would allow you to safely use wire that is smaller.

Those numbers are from the max discharge rate the battery has been tested to without burning or exploding, a short circuit would likely be at a higher rate of discharge.





;

so what your saying is "DO NOT ATTEMPT THIS IN ANY WAY OR YOU WILL KILL SOMEONE!!!!" lol

Not at all! You just need to be sure the proper protections are in place to use the product SAFELY

thats the bits i dont know how to put in place

unless someone is able to teach me (tell me) how to do it, without going into overkill detail, i think ill tell him that im gonna give this project a miss.

Consider this: the Power Supply that Jeol's client has requested is for a post apocalyptic scenario. So design with _that_ in mind:

* What does "saftey" look like in that world?
* Will fuses be available? Perhaps resettable circuit-breakers are in order.
* Perhaps redundant paths should be considered.
* How about EMP/radiation hardening?
* Should it be tethered to the ground (or otherwise secured) in case of rapture? If Jeol's client is taken, he won't need it, but if not, it would be good for it not to float away--even if to be dropped from some height, when flotsam is jettisoned

I jest, but what the heck, a good engineer considers all possibilities that relate to the stated requirements.

I have a 4500 watt sign wave inverter tied to 4 car batteries. (Don't try to power a widow a/c with anything but sign waves or the will squeal like a stuck pig. I even used a 10 kw non sign wave inverter for a 4 kw widow unit. Same results.) Bottom line.. with 4 1000 ah OPTIMA car batteries in parallel run time was about 6 hrs. These are charged by a bank of 12 solar panels each putting out about 1.2 Amps full sunlight. It still takes a week to recharge. The inverter sucks between 30 and 35 amps dc no load to half load. It drops out at 11.2 vdc. I have since purchased a Guardian 17 Kw whole house generator (natural gas). We have never lost natural gas during Ike or any other huricanes or power outages. I still can't compete with the main power at 10 cents per kwh; but I don't ever have to worry about any power outages either. If for some reason natural gas is interrupted; I can plug in my gas powered 8 kw generator and flip a breaker switch and power my house. It cost about $100 moth but I think of it like insurance for the next 10 to 15 years. Any one interested in my solar system?