Searching \ for '[EE] Making my own high current battery charger' 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: 'Making my own high current battery charger'.

Exact match. Not showing close matches.
PICList Thread
'[EE] Making my own high current battery charger'
2010\12\23@140552 by V G

picon face
I want to build my own high current battery charger for Ni-MH battery types
(specifically Eneloops).

* The device will be designed to charge anywhere from 1-8 cells in parallel..
* Must charge at high currents PER channel, up to 4 A per channel
(adjustable via some sort of switch selector, maybe in 500 mA increments).
So that means a maximum output of 4 * 8 = 32 A. I'm going to be using an old
300-500 W computer power supply for this. It can output on 5 V and 12 V
rails, and more than enough current.
* As for charge termination, I'm thinking primarily a negative delta V
detection. Secondary mechanism would probably be a dT/dt mechanism, but
that's not a major concern.

I'll be using precision ADCs for voltage monitoring, need millivolt
precision. The system will be microcontroller controlled with displays for
voltage and other information.

1. My biggest question here is: how do I make a high current (up to 8 A)
constant current supply from the 5 V/12 V power supply output?
2. How do I switch large currents (up to 8 A)? Would I use MOSFETs? The
currents need to be controlled via a microcontroller.
3. How do I make this constant current supply adjustable? As in 250 or 500
mA increment selectable, for example?
4. Some additional functions I want to include are connections to a
computer, so I can do charge/discharge voltage and time profiling on the
batteries. Any ideas on this

2010\12\23@142454 by Isaac Marino Bavaresco

flavicon
face
Em 23/12/2010 17:05, V G escreveu:
{Quote hidden}

Step-down switching converter.


> 2. How do I switch large currents (up to 8 A)? Would I use MOSFETs? The
> currents need to be controlled via a microcontroller.


Yes, MOSFETS.


> 3. How do I make this constant current supply adjustable? As in 250 or 500
> mA increment selectable, for example?


Measure the effective current and adjust the PWM in software.


> 4. Some additional functions I want to include are connections to a
> computer, so I can do charge/discharge voltage and time profiling on the
> batteries. Any ideas on this?


UART (easier) or USB (more modern and compatible).


Best regards,

Isaac

__________________________________________________
Fale com seus amigos  de graça com o novo Yahoo! Messenger http://br.messenger.yahoo.com

2010\12\23@152637 by Olin Lathrop

face picon face
V G wrote:
> I want to build my own high current battery charger for Ni-MH battery
> types (specifically Eneloops).

Did you even look at the link I pointed you at a week or two ago where I
have done something similar?

> * The device will be designed to charge anywhere from 1-8 cells in
> parallel.

Bad idea.  Handle each cell individually.  It also makes the circuit per
cell simpler as it has to handle less current.  Once you have one
breadboarded and debugged, it's not big deal to replicate it 8 times.

> * Must charge at high currents PER channel, up to 4 A per channel
> (adjustable via some sort of switch selector, maybe in 500 mA
> increments).

There should be no user choice.  The firmware should control the charge
current.  Up to 1C maximum current is quite safe, but there is no need it
has to be that high.  There is nothing wrong with charging NiMH cells at 1/2
C, for example, just that it takes longer.  Eneloops are rated for 2.5 A-h,
so I would make the max current per cell in the 2 to 2.5 amp range.  There
is no need to push it, especially if you can handle a bunch of cells.
Arrange it so that some will always be fully charged and ready for use.

> So that means a maximum output of 4 * 8 = 32 A.

Again, 4A per cell is excessive.

> * As for charge termination, I'm thinking primarily a negative delta V
> detection. Secondary mechanism would probably be a dT/dt mechanism,
> but that's not a major concern.

I used a combination of dV/dt and elapsed time.  It worked fine.

> I'll be using precision ADCs for voltage monitoring, need millivolt
> precision.

You don't need "precision A/Ds" beyond the A/D built into a PIC.  Even a 10
bit A/D is fine.  Again, take a look at what I did.  It used a PIC 10 bit
A/D and worked fine.

Maximum voltage of a NiMH cell is around 1.5V.  Let's say for sake of margin
you scale 2V to the PIC 10 bit A/D range.  That gives you about 2mV cell
voltage resolution, which is more than enough.

> The system will be microcontroller controlled with
> displays for voltage and other information.

Getting information out can be useful for understanding the system and
debugging, but it not really needed for normal operation.  I gave each
channel a single LED.  You can do a lot with different blink patterns to
indicate the state of a channel.

A display may be nice, but it will be a lot easier to dump telemetry data to
a PC via a serial port and display it there.

> 1. My biggest question here is: how do I make a high current (up to 8
> A) constant current supply from the 5 V/12 V power supply output?

The simplest answer is a linear regulator with current feedback.  It will
get warm, but you can arrange to dissipate most of the power in the current
sense resistor.  If you limit each channel to 2A, then there is 10W going
somewhere per channel.  If the battery is at 1.5V, then it gets 3W and the
other 7W is heat.  If you use a 3 Ohm current sense resistor, then it
dissipates 6W and the transistor only 1W, and leaves the transistor with
500mV headroom to perform the regulation.  A TO-220 transistor with a small
heat sink can handle over 1W in air easily.  You can get power resistors to
almost any power.  10W should be findable.  That can also be a bunch of 2W
resistors in the right paralell/series combination.  It's probably cheaper
that way as you can buy a pile of 100 2W resistors of the right resistance
and string them together to get 8 of what you want.  Getting or making a 3
Ohm 10 Watt power resistor isn't that hard.

There are more finesse ways to do this, but they are all more complicated.

> 3. How do I make this constant current supply adjustable? As in 250
> or 500 mA increment selectable, for example?

I would use a opamp to drive the pass transistor, which is a NPN in emitter
follower configuration.  The opamp circuit regulates the current based on a
analog input signal, which is generated by low pass filtering a PIC PWM
output.

> 4. Some additional functions I want to include are connections to a
> computer, so I can do charge/discharge voltage and time profiling on
> the batteries. Any ideas on this?

Serial port will be the easiest.  That is what I did.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000

2010\12\23@153929 by Olin Lathrop

face picon face
Olin Lathrop wrote:
> If you use a 3 Ohm current sense resistor,

I meant to say have the current sense resistor drop 3V, which at 2A would be
1.5 Ohms, not 3 Ohms.  Sorry for the confusion.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000

2010\12\23@155822 by V G

picon face
On Thu, Dec 23, 2010 at 3:40 PM, Olin Lathrop <spam_OUTolin_piclistTakeThisOuTspamembedinc.com>wrote:

> I meant to say have the current sense resistor drop 3V, which at 2A would
> be
> 1.5 Ohms, not 3 Ohms.  Sorry for the confusion.


Thanks for the reply. I made some mistakes in my post.

I meant to say that charging is not technically in parallel, but can charge
8 cells at a time, so 8 different channels. So yes, I'll just build a
circuit and replicate it for 8 times.

I want it to be capable of 4 A per channel because Eneloops aren't the only
types of Ni-MH I'm charging. But I suppose 2 A max is fine.

I'll take a look at your link

2010\12\23@160112 by V G

picon face
On Thu, Dec 23, 2010 at 3:27 PM, Olin Lathrop <.....olin_piclistKILLspamspam@spam@embedinc.com>wrote:

> Did you even look at the link I pointed you at a week or two ago where I
> have done something similar?


Is this what you were talking about?

http://www.embedinc.com/pic/bat

2010\12\23@170124 by Olin Lathrop

face picon face
V G wrote:
> Is this what you were talking about?
>
> http://www.embedinc.com/pic/bat/

Yes.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000

2010\12\24@134251 by Brooke Clarke

flavicon
face
Hi V G:

The advice that's been given will get you 80 percent of the way to max charge, but there are some things that work better.
This may be controversial, but I believe that pulse charging methods do in fact work better than constant current/voltage methods.  I saw this when working with the Propel BB-590 Battery Charger.
http://www.prc68.com/I/PropelBB590.shtml
It uses what's called Burp Charging, see:
http://www.prc68.com/I/BatChg.shtml#Pu

Also note that you need to use a temperature sensor when charging Ni-MH cells but not when charging Ni-Cad cells because of the differences in the chemistry, for more on that see the "Handbook of Batteries" page 29-22 in the 3rd edition (Ni-Cad is endothermic, Ni-MH is exothermic).  In fact the manufacturer data sheets for some batteries uses temperature change as the key termination method followed by a topping up charge.

The Triton 2 charger- discharger allows programming pretty much any parameters within its power ratings:
http://www.prc68.com/I/Triton2.shtml

For AA size batteries the Maha MH-C401FS 4-channel charger is hard to beat.
http://www.prc68.com/I/MHC401FS.shtml

For analyzing AA batteries the Maha MH-C9000 Charger Analyzer is very capable:
http://www.prc68.com/I/MHC9000.shtml
This charger supports topping up.

You can use a modified capacitor ESR meter to measure the internal resistance of a battery.
http://www.prc68.com/I/BatTst.shtml#IR

I haven't updated my Eneloop page for some time, maybe next year.
http://www.prc68.com/I/RTU-Batt.shtml

-- Have Fun,

Brooke Clarke
http://www.PRC68.com

More... (looser matching)
- Last day of these posts
- In 2010 , 2011 only
- Today
- New search...