[chbot] eBay Li-Ion cell charger experiment

[Mark Atherton]

Hi All,

Please don't do as I did and buy any of the 
charger boards I referenced from eBay (below), 
just ran one up and they appear to be rubbish. 
The unit is supposed to take 5V in, charge a 
single cell (or parallel pack), then provide 5V 
via a boost regulator on the way out.

5V input charger appear to be linear, so with 
500mA charge current, the regulator gets a bit 
warm - not too worried about this.

Replacing the Li Ion cell with a variable PSU, 
there appears to be no under voltage protection 
in the discharge path - so the output boost 
regulator could completely discharge the battery 
- this is very bad. This is quite odd since the 
board has a DW01+ "One Cell Lithium-ion/Polymer 
Battery Protection IC" along with associated 
MOSFET, so there was an opportunity to do this 
correctly. Not sue what is going on here, from 
notes below it could be a false measurement, and 
something to do with source resistance of the PSU.

Quiescent current of boost regulator with no 5V 
load attached to the output is 650uA. This will 
fully drain a 3 x AA pack (2400mA total) in about 
4000 hours (6 months) - again, bad.

After a cell has been attached, the boost 
regulator only starts after a brief charge has 
been applied, not sure why, but not too bothersome.

Charge path to cell is very sensitive to series 
resistance. Charge current dropped from 500mA to 
100mA when a Fluke 77 on the 10A range between 
the cell and charger - this is 90 milli-ohms of 
extra resistance (including leads), which with 
500mA charge current is 45mV, so the circuit is 
very sensitive to small impedance and voltage changes.

As with Volker, I started off by thinking about 
designing my own, then got lazy, bought some from 
eBay (along with some kind of unexpected 
education). Single cell is much easier to manage 
for low power, low voltage applications.

So back to the (PCB) drawing board.

Lessons learned:
- custom designed PCB may be interesting - even 
for the planned 200mA charge / discharge unit 
impedances are going to be low, so layout will be critical.
- with a 5V input, a linear charger should be 
simple and easy (?) MCP73811 appears to be 
available from Active Components for NZ$1.25 with 85 or 450mA charge current.
- would prefer a charger with a status output, so 
will keep exploring - need to be low cost and readily available
- A physical switch between cell and management 
circuit will eliminate unwanted leakage / 
discharge, may also need to disable charger with second pole.
- a mini-B USB input charge connector will remove 
the opportunity of inadvertently attaching a 12V 
charging power source from all but the most enthusiastic (!)
- as always, everything isn't always what it 
seems. Beware enthusiast Marketing types, and cheap products from eBay

-Mark

NB Fluke 77 on the 10A range, without leads, and 
doing a 4 wire measurement is 14 milli-ohms, measured using an Agilent 34401A

============

How timely.

I am pretty sure that Li Ion and Li Po have similar issues:

Been given a small pile of Li Ion batteries, and 
bought some http://www.ebay.com/itm/260841670123

Briefly, if you over charge them they do occasionally explode

If you over discharge them then they can lose capacity in a real hurry

There is a difference between 4.1V and 4.2V 
cells, beware - this is absolute max terminal voltage.

The boards above have 5V in, 5V out and cell 
protection, so the cell is taken out of circuit 
of over-charge or over-discharge imminent. I do 
plan to run one of these up carefully and see how they behave.

Li Ion can deliver large currents, so an in-line fuse is mandatory.

 From Wikipedia
         Li Ion specific energy 100–265 Wh/kg
         Ni Mh specific energy 60–120 Wh/kg
         Ni Cd specific energy 40–60 Wh/kg

The good news is there is a lot in a lightweight package

The bad news is that there is a lot in a lightweight package

If treated with respect, they behave well -- these are not for the unwary.

In terms of burning your house down, you probably 
already have lots of these units in your phone, laptop etc. already.

If you have concerns about safety, stick with 
NiMh the older types have self discharge 
problems, but I don't think they will blow up (a 
function of internal resistance and available energy).

Mr Harris is one of (or The) resident expert, and 
I await comments from him on this brain-dump.

-Mark


At 11:17 a.m. 17/05/2013, you wrote:
>How safe is it really to have LiPo batteries around the house?
>
>I was thinking of using some of the 1-cell 
>hobbyking offerings as backup for some 
>microcontrollers. Then I read the reports of 
>what these batteries do to people, like burning 
>the house down, but I don't fancy having the 
>chemical smoke of one of them's guts in the house either.
>
>What would people consider to be safe handling 
>precautions in terms of charger, charge location 
>and storage? What about charging them in-circuit?
>
>Thanks,
>
>Volker