[TriEmbed] Fwd: Lithium Ion batteries

Thu Mar 6 12:50:12 CST 2014

---------- Forwarded message ----------
From: <kschilf at yahoo.com>
Date: Thu, Mar 6, 2014 at 2:42 AM
Subject: Re: [TriEmbed] Lithium Ion batteries
To: Charles West <crwest at ncsu.edu>

Hi Charlie,

I am not an expert on batteries, but I can share a few pointers.  :-)

In general, it is not a good idea to connect voltage sources in parallel
without current management.  With large loads, you want the sources to
share the load to prevent one source from overheating.  You should also be
concerned about a higher voltage source dumping current into the lower
voltage source.  Electrons follow potential not human desires.  With
batteries this is especially critical.  If a battery shorts internally it
will happily take current from its parallel neighbors.  If you are lucky,
this will just discharge the other batteries.  If not, this large current
through the faulty battery can cause a fire.  Higher energy density
batteries are great under normal circumstances, but they also make a bigger
bang when they fail.  :-)

The simplest approach to parallel batteries to to connect them in parallel
through diodes.  This does nothing for current sharing but does prevent a
bad battery from bringing down the system.  The cost is a loss of energy =
current * the forward voltage of the diode.

If you have cells in series, charging slows/stops when the first cell
reaches its maximum charge.  On the other end, the string stops discharging
(providing energy) when the first cell is depleted.  Balancing the cells
helps to age them at a uniform rate.  You want all of the cells to have
similar charge and discharge characteristics to maximize energy
storage/delivery and provide greater reliability.

Can you charge and run at the same time?  The first requirement is a
charger large enough to provide charge current and the running current
simultaneously.  That gets you to the party.  The next question is how the
power supply in your specific devices are designed?

Be safe.  Large battery packs should not be treated lightly.  Even modest
batteries can provide large peak currents.  If you want more information on
specific charging profiles or load sharing, I would point you to one of the
semiconductor vendors that make battery charger chips (Texas Instruments,
Linear Tech, Analog Devices, etc.).  I am kind of partial to Linear Tech (
linear.com).  Any of these vendors will have a great deal of technical data.

Good luck,
Kevin Schilf

  On Wednesday, March 5, 2014 11:48 PM, Charles West <crwest at ncsu.edu>
wrote:
 Hello,

I'm trying to figure out the best way to power my pedestrian robot and both
Nathan and I are trying to determine the best way to power the
Astropresence telepresence drone.  The task is somewhat complicated by the
fact that we really want both robots to be able to recharge by docking
(rather than swapping out the batteries).

What follows is what we have learned so far mixed with questions we have
for anyone more experienced with Lithium Ion batteries.

Lithium Ion seems the way to go in terms of power density and cost.
However, charging and connecting them can be tricky and some sources on the
web seem to be in contradiction.  We focused in on 18650 cells because they
seem cheap and commonly used.

Everyone agrees that it is problematic to have the voltage on a battery get
too low or too high.  Some 18650 cells come with built in voltage
protection and most battery packs have PCBs to restrict the voltage.

Where people seem to disagree the most is charging battery packs.  Some
sources say that you have to have balancing circuits when connecting
batteries together and charging them.  Otherwise, one battery gets a lower
max voltage and the other batteries are are overvolted to try make up the
difference.  Other sources just connect the cells in parallel and leave it
be.

BatterySpace (
http://www.batteryspace.com/pcmforli-coli-mn-nibatterypacks.aspx) says that
balancing becomes important for >5 cells and that with >20 cells you should
get a more sophisticated battery control circuit.  In reading again, it
seems that they treat parallel batteries as one cell (Is that right?).
They sell voltage restriction PCBs, if anyone is interested.

It is also not entirely clear how sophisticated of a charging circuit is
needed.  I believe the consensus is that merely plugging the voltage
restricted battery pack into a similarly valued power supply is not a good
idea (Pete posted an awesome guide to the charging profiles:
http://www.digikey.com/en-US/articles/techzone/2012/sep/a-designers-guide-to-lithium-battery-charging).
They sell Lithium Ion chargers for different voltages but it is not clear
if they are suitable for batteries in parallel.  It is also not clear if it
would be possible to have an active circuit connected to the batteries
while they are charging.

It would be ideal if the robot could still be active while it was being
charged.  It is not clear whether or not that would confuse the charger
(also the charger might not be able to supply enough current to both charge
the battery and run the robot.

On a separate note, there are many 18650 cells for purchase on Ebay.  Many
sources say that the ones that are labeled 4000 mah are lying about their
capacity and often garbage.  Some of these cells have been reported to be
smaller batteries placed in larger cell packaging with conductive filler.
I believe the industry standard is on the order of 2200 mah to 2600 mah.  I
believe some of the "4000 mah" ones have been reported to have 2100 mah
capacity, so they are not all garbage (but be careful and test a few before
you order a bunch).

Another cheap source for 18650 cells is apparently used and laptop
batteries.  Many laptop batteries are composed of 18650 cells which can be
taken out (there are several youtube videos on the subject).

To some up the outstanding questions:
Can Lithium Ion batteries be connected and charged in parallel without some
special arrangement?  How important is balancing and when should it be used?

Can an active load be left in the circuit while a charger is working?

Thanks,
Charlie West

_______________________________________________
Triangle, NC Embedded Computing mailing list
TriEmbed at triembed.org
http://mail.triembed.org/mailman/listinfo/triembed_triembed.org
TriEmbed web site: http://TriEmbed.org <http://triembed.org/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.triembed.org/pipermail/triembed_triembed.org/attachments/20140306/3954f3a2/attachment.htm>