Searching \ for '[EE]: NiMH battery packs' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/power/batterys.htm?key=battery
Search entire site for: 'NiMH battery packs'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: NiMH battery packs'
2001\05\13@090639 by Olin Lathrop

face picon face
part 1 1801 bytes content-type:text/plain; (decoded 7bit)

I finally got around to doing some testing with the NiMH battery packs that
several people here received recently.

I made a little circuit with a 16F876 that can enable/disable the charge
current to a pack and also measure the pack voltage.  For an initial test I
just used a 36 ohm resistor to +5 thru a PNP as the charge circuit.  I
wanted to start with low current and with the resistor the charge current
automatically gets backed off a bit when the pack gets charged up.  I know,
this is a long way from the final charge circuit, but I wanted to get some
idea what we're dealing with before getting more aggressive.

The PIC turns on the charge current for 3 seconds at a time and takes lots
of voltage readings which get heavily low pass filtered.  It then turns off
the current, waits 100mS, then takes lots of voltage readings with the pack
open circuited.  These are again filtered.  When this is done (about 200mS
after the charge current was shut off), it starts the cycle over again by
turning on the charge current again.

It also send the charging and open circuit voltage over RS-232 where a
program on the host interprets the value.  This program saves the charging
and open circuit battery pack voltage, and uses these to compute the charge
current and the battery internal resistance.  All four values are written to
a CSV (comma separated values) file along with the elapsed charge time.
Attached is the result of my first test that ran for about 10 1/2 hours.
The stepped nature of the resistance curve is do to the quantization of the
voltages.  Still, I was surprised how low it was.

I plan on trying a differnt virgin battery pack (I've got plenty) and doing
this again with higher current.  I'll keep you all posted as I get more
results.


part 2 18031 bytes content-type:image/gif; (decode)


part 3 343 bytes

********************************************************************
Olin Lathrop, embedded systems consultant in Littleton Massachusetts
(978) 742-9014, spam_OUTolinTakeThisOuTspamembedinc.com, http://www.embedinc.com

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\13@091919 by Peter Tiang

flavicon
face
Hi Olin,

   Are you trying to build a NiMH charger ?

   If so, you should'nt be charging the NiMH battery
   this way. You should use constant current, typically
   at 1C, i.e. if you battery is 300mAH, you should
   charge at a constant current of 300mA. Therefore
   if from fully discharge, it should take around 1 hour
   to charge your battery.
   Your ON/OFF current circuit is not suitable
   for this, you probably need a PWM-controlled
   DC-DC switcher.

   Your charge curve looks more suitable for
   Li-ion batteries. For Li+ you charge using
   constant voltage (typically 4.1~4.2V per cell)
   with current limit (to prevent huge current
   during initial stage of charging).

Regards,
Peter Tiang


{Original Message removed}

2001\05\13@182813 by Olin Lathrop

face picon face
>     Are you trying to build a NiMH charger ?

Yes, eventually.  As I said, the test circuit is a long way from the final
charge circuit.  Right now I'm trying to get a better handle on how these
batteries behave under various charging conditions.

>     If so, you should'nt be charging the NiMH battery
>     this way. You should use constant current, typically
>     at 1C, i.e. if you battery is 300mAH, you should
>     charge at a constant current of 300mA. Therefore
>     if from fully discharge, it should take around 1 hour
>     to charge your battery.

1C is the maximum rapid charge current.  There are a bunch of restrictions
on that.  You can't just charge at 1C without careful monitoring and knowing
where you are on the charge curve.  The trickle charge current is apparently
.033C to .05C.  I was starting with a simple circuit at the lower end of the
allowable charge current range just to see how the battery voltage behaved.
I intend to crank up the current on successive tests.  I'm running another
test with higher charging current right now.

Eventually I want to build a charger for these battery packs.  I don't care
much how fast the charger is.  I'm more interested in doing the best quality
job.  I think that means getting the batteries as fully charged as possible
without damaging them, and being able to start with packs of unknown charge.
If necessary, I could have the charger discharge a pack to a measurable
state if that makes the charging easier or is better for the battery.

I'd like to avoid measuring the battery temperature, so I hope there is a
reliable algorithm to determine how much current to use when and when to
stop by looking just at the voltage.  This is the ultimate goal, but I'm
also curious about these things and want to see how they behave under
different conditions.  I've done a bunch of research on the web.  I found a
lot of information (the Panasonic site was particularly helpful), but there
are still unanswered questions.  For example, it's not always clear when
data sheets talk about the battery voltage at various stages whether thaYQgm
open circuit or while charging.  Does it matter?  If so, what are the
differences?  So far I haven't seen much difference, but that may change as
I get to higher current.

>     Your ON/OFF current circuit is not suitable
>     for this, you probably need a PWM-controlled
>     DC-DC switcher.

I think you missed the point.  I'm not trying to control charging current by
switching.  I am only switching the current on and off so that I can measure
the battery voltage under both conditions.  One thing I don't know yet is
whether the battery needs to "stabalize" somehow after the charging current
is shut off.  For now I am waiting 100mS after shutting of the current
before starting to measure the voltage, but I have no evidence that this is
too short or completely unnecessary.

Another thing I'd like to know is can PWM be used to charge these batteries.
At what PWM frequency, if any, will the battery "see" the average current
instead of individual bursts of high current with dead times of no current
in between?  I haven't seen any discussion of this at all so far.  PWM would
be nice, and it would also be nice not to need large inductors - but will
that work?  I'll post more information as I find it.


********************************************************************
Olin Lathrop, embedded systems consultant in Littleton Massachusetts
(978) 742-9014, .....olinKILLspamspam@spam@embedinc.com, http://www.embedinc.com

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2001\05\14@022653 by Peter Tiang

flavicon
face
> >     Are you trying to build a NiMH charger ?
>
> Yes, eventually.  As I said, the test circuit is a long way from the final
> charge circuit.  Right now I'm trying to get a better handle on how these
> batteries behave under various charging conditions.

   So far, I've only came across the constant
   current charge method. Only this way can
   you obtain the characteristic -deltaV or
   dT/dt for charge termination.

   There was another method of "pulsed" charging,
   where the battery is repeatedly charged and
   discharged, with a net increase of charge over time.
   But this is also done at constant current.

   The reason for the "pulsing" is to prevent
   crystal/dendrites formation in the cells.

> 1C is the maximum rapid charge current.  There are a bunch of restrictions
> on that.  You can't just charge at 1C without careful monitoring and
knowing
> where you are on the charge curve.  The trickle charge current is
apparently
> .033C to .05C.  I was starting with a simple circuit at the lower end of
the
> allowable charge current range just to see how the battery voltage
behaved.
> I intend to crank up the current on successive tests.  I'm running another
> test with higher charging current right now.

   I've came across higher recommendation for
   rapid charge (2C), in fact 1C is just normal
   charging rate.

>
> Eventually I want to build a charger for these battery packs.  I don't
care
> much how fast the charger is.  I'm more interested in doing the best
quality
> job.  I think that means getting the batteries as fully charged as
possible
> without damaging them, and being able to start with packs of unknown
charge.
> If necessary, I could have the charger discharge a pack to a measurable
> state if that makes the charging easier or is better for the battery.

   For the best possible charge, you should aim
   for the most accurate charge termination.

   This can only be done (as far as I know) by
   giving constant current and monitoring the voltage
   and rate of change in cell temperature for
   charge termination.

   NiMH does not suffer "memory effect" as much
   as NiCD, but a full discharge is advisable
   before charging to maximise battery capacity.
   Therefore you also need a discharge circuit.

> I'd like to avoid measuring the battery temperature, so I hope there is a
> reliable algorithm to determine how much current to use when and when to
> stop by looking just at the voltage.  This is the ultimate goal, but I'm
> also curious about these things and want to see how they behave under
> different conditions.  I've done a bunch of research on the web.  I found
a
> lot of information (the Panasonic site was particularly helpful), but
there
> are still unanswered questions.  For example, it's not always clear when
> data sheets talk about the battery voltage at various stages whether
thaYQgm
> open circuit or while charging.  Does it matter?  If so, what are the
> differences?

   Yes, it does matter.

   You need to monitor the voltage during off-current
   in order to get an accurate -deltaV.
   During on-current, there is a drop due to the
   charging current and the battery's internal resistance.

   This internal resistance is affected by temperature
   and state of charge, therefore it present a source
   of inaccuracy in your measurement.

>So far I haven't seen much difference, but that may change as
> I get to higher current.

   I'm sure you'll see a much more pronounced
   -deltaV on the off-current, assuming that
   you'll be pumping in constant current.

> >     Your ON/OFF current circuit is not suitable
> >     for this, you probably need a PWM-controlled
> >     DC-DC switcher.
>
> I think you missed the point.  I'm not trying to control charging current
by
> switching.  I am only switching the current on and off so that I can
measure
> the battery voltage under both conditions.

   If you are switching a constant current source
   on and off, it's fine. But in this case your
   current source is not constant.

> One thing I don't know yet is
> whether the battery needs to "stabalize" somehow after the charging
current
> is shut off.
>   For now I am waiting 100mS after shutting of the current
> before starting to measure the voltage, but I have no evidence that this
is
> too short or completely unnecessary.

   From charge plots that I have from 5 years ago,
   it does not really matter as the voltage difference
   between on and off current more-or-less follows
   Ohm's law.

   Though you might want to wait at least couple of
   mS to avoid overshoot/undershoot due to the switching.
   Of course, you won't wnat to wait too long as
   self-discharge or discharge thru your circuit might
   be significant.

> Another thing I'd like to know is can PWM be used to charge these
batteries.
> At what PWM frequency, if any, will the battery "see" the average current
> instead of individual bursts of high current with dead times of no current
> in between?

   That's why my suggestion for a DC-DC switcher.

   The PWM frequency depends on how big the value
   of the LC filter you choose. If you choose
   a higher PWM frequency, a smaller value would
   do and vice versa.

   The switcher will provide the battery with
   a constant current source.

   Of course to implement constant current control
   via PWM, you'ld also need a current measurement
   circuit.

   Things is getting pretty complicated ;-)

> I haven't seen any discussion of this at all so far.  PWM would
> be nice, and it would also be nice not to need large inductors - but will
> that work?  I'll post more information as I find it.

   The whole idea is to provide a constant current
   source, be it 1C or .05C , and I think only
   a DC-DC switcher will do.

   Any suggestions from other PIClisters ?

Regards,
Peter Tiang

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\14@023731 by David VanHorn

flavicon
face
>
>     NiMH does not suffer "memory effect" as much
>     as NiCD, but a full discharge is advisable
>     before charging to maximise battery capacity.
>     Therefore you also need a discharge circuit.

Can you have a positive, real number less than zero?

http://www.fridge41.freeserve.co.uk/memory.htm
--
Dave's Engineering Page: http://www.dvanhorn.org
Where's dave? http://www.findu.com/cgi-bin/find.cgi?kc6ete-9

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\14@031054 by Peter Tiang

flavicon
face
Thks Dave, learned something today, on the
actual need for a discharge before charge.

Does this apply to NiMH battery as well ?


{Original Message removed}

2001\05\14@031932 by David VanHorn

flavicon
face
At 03:13 PM 5/14/01 +0800, Peter Tiang wrote:
>Thks Dave, learned something today, on the
>actual need for a discharge before charge.
>
>Does this apply to NiMH battery as well ?

I'm not certain, but the lack of memory certainly does.

The memory effect was seen in cells that were built for sattelites.
Sats have very regular orbits, and their energy consumption is also pretty
regular.
In order to see it, they had to keep the discharge cycles to within 4% IIRC.

So Overcharge Depression is the real thing to fight, and discharging the
cells is a way to patch it, but not a way to FIX it.  Fixing the charger is
the right solution.

In my experience, it's always better to fix something that's actually broke.
Better results that way.

--
Dave's Engineering Page: http://www.dvanhorn.org
Where's dave? http://www.findu.com/cgi-bin/find.cgi?kc6ete-9

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\14@101536 by Lawrence Lile

flavicon
face
part 1 4383 bytes content-type:text/plain; (decoded 7bit)

This is really valuable info, Olin!

I'm on the way to building a charge controller too.  To summarize a bunch of
earlier posts, and some reading I've been doing:

1.  Temperature methods will probably work at charge rates above 1/3C -
1/2C.

2.  Voltage sensing methods (zero delta-V or negative Delta V) are probably
the only methods that will work below 1/3C charging rates.

3.  At low charging rates, deltaV may be difficult to detect, esp. if the
battary pack is a big stack of batteries. OTOH, the risk of overcharging
decreases some as current decreases.

4.  Maxim and National both have nice charging circuits designed for these
purposes.  I'm probably going to build the Maxim circuit.
http://dbserv.maxim-ic.com/quick_view2.cfm?pdf_num=1666  They have a kit,
but it is expensive.  Doesn't look like an expensive list of parts at all.

5.  It is best to discharge to 1.0 V per cell (valid for up to 6 cells in
series) or (# Batteries -1)*1.2V for more than 6 in series.  Although NiMH
are not supposed to have the memory effect, Panasonic sez that repeated
discharging to 1.1 V can "reduce capacity", whereas full discharging to 1.0V
per cell is best.  (Sounds like memory effect to me).

6. FAQ: "DO we use negative deltaV, Zero DetaV, or DeltaTemperature
charging?"   Maxim says the following:

FOR NIMH BATTERIES  at charge rates > C/2use negative DeltaV and
temperature, Max713 charger
Charge rates < C/2 use Zero DeltaV (Max712), temperature may not work.

7. Avoid charging at very cold or hot temperatures.  <0C and > 40C

8.  According to Panasonic and Maxim, NiMH batteries definitely DO display
negative voltage slope at full charge.  Several posts have debated this one
way and another.   According to more authoratative sources such as Piclister
Bob Blick, negative deltaV may be hard to detect in a battery stack.  I'm
going with a Zero DeltaV cutoff, because I think it is safer and might
increase battery life at the cost of less fully charged batteries.  Check
the attached graphs.

9.   Panasonic battery specs:
www.panasonic.com/industrial_oem/battery/battery_oem/chem/nicmet/nicm
et.htm

See Especially their PDF on NiMH charging methods



-- Lawrence Lile




{Original Message removed}
part 2 8671 bytes content-type:application/octet-stream; (decode)

part 3 131 bytes
--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\14@102149 by Lawrence Lile

flavicon
face
> Another thing I'd like to know is can PWM be used to charge these
batteries.
> At what PWM frequency, if any, will the battery "see" the average current
> instead of individual bursts of high current with dead times of no current
> in between?  I haven't seen any discussion of this at all so far.  PWM
would
> be nice, and it would also be nice not to need large inductors - but will
> that work?  I'll post more information as I find it.

Maxim has a circuit all designed for thier MAX713 which uses a switching PWM
charging circuit, and works for NiMH batteries.  See my other post today for
URL:

Lawrence Lile

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\14@111640 by David Cary

flavicon
face
Dear Olin Lathrop,

Olin Lathrop <olin_piclistspamKILLspamEMBEDINC.COM> on 2001-05-13 08:03:04 AM wrote:
> I finally got around to doing some testing with the NiMH battery packs

> For an initial test I
> just used a 36 ohm resistor to +5 thru a PNP as the charge circuit.  I
> wanted to start with low current
...
> I know,
> this is a long way from the final charge circuit, but I wanted to get some
> idea what we're dealing with before getting more aggressive.

Start simple. Good idea.

> The PIC turns on the charge current for 3 seconds at a time

Sounds like an interesting project. For a long time I've wanted to develop
something that could test batteries to find their true characteristics, rather
than rely on 3rd hand rumor and innuendo. Too many things to do, not enough time
...

I suspect that the "best" way to charge batteries does not involve completely
discharging them. However, it sounds like you don't mind testing a couple of
batteries to destruction. In that case, consider adding a "discharge" circuit,
so you can automatically run charge/discharge cycles over a few weeks
(simulating charging-on-charger / discharging in user application). (This test
would run until the battery is not just "dead", but completely killed -- unable
to recharge).

That way you can measure the things that are really important: how does the
battery act in the user application ? I.e., how much total energy can I suck out
? How quickly can I suck it out ? Given that my RC car runs a race in 10
minutes, how many races can I run with this battery before I should throw it
away and buy a new one ? (What other important characteristics am I leaving out
?). How do I optimize these 3 numbers, by changing the characteristics of the
charger and/or my application circuit ?

> This program saves the charging
> and open circuit battery pack voltage, and uses these to compute the charge
> current and the battery internal resistance.

Say we test a battery by charging for 5 seconds, discharge (through ~ 36 Ohm)
for 2 seconds, and open-circuit for a few seconds in-between ... repeating until
the battery is charged.

Is the battery internal resistance the same when charging as when discharging ?

...
>The stepped nature of the resistance curve is do to the quantization of the
>voltages.  Still, I was surprised how low it was.

Resistances of 0.2 Ohms are pretty small, and difficult to measure accurately.
At that level, the resistances of your wiring and connectors start to become
significant.

>One thing I don't know yet is
>whether the battery needs to "stabalize" somehow after the charging current
>is shut off.  For now I am waiting 100mS after shutting of the current
>before starting to measure the voltage, but I have no evidence that this is
>too short or completely unnecessary.

Looks like you could just modify your software to collect lots of data over a
couple of charge/rest/charge/rest cycles, then (instead of averaging) just dump
this ~10 seconds worth of raw data to your PC (marking the exact instant the
charger was turned on). That would answer this question. Could you post this
data as well ?

> I'll keep you all posted as I get more results.

Thanks, this sounds fascinating.

--
David Cary

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2001\05\14@122642 by Peter Tiang

flavicon
face
> That way you can measure the things that are really important: how does
the
> battery act in the user application ? I.e., how much total energy can I
suck out
> ? How quickly can I suck it out ? Given that my RC car runs a race in 10
> minutes, how many races can I run with this battery before I should throw
it
> away and buy a new one ? (What other important characteristics am I
leaving out
> ?). How do I optimize these 3 numbers, by changing the characteristics of
the
> charger and/or my application circuit ?
>

   Smart Battery, http://www.smbus.org, and it's already
   used in most late-model notebooks.

   It can give you information like "If I'm to
   draw 2A for 30s would the battery be able to
   supply ?".

   Useful when notebooks are trying to spin up
   harddisk and need a prediction on the power
   available from battery packs to avoid
   catastrophic failure when battery packs
   suddenly ran out of juice.

   The idea behind smart battery is that instead
   of the charger trying to be "smart",
   let the battery control the charging algorithm.
   The charger effectively becomes just a programmable
   current/voltage source.
   This way, when new battery chemistry comes out,
   the charger won't be obsoleted.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


More... (looser matching)
- Last day of these posts
- In 2001 , 2002 only
- Today
- New search...