Super Simple DIY Li-Ion Rechargeable Charger

djpark

Enlightened
Joined
Nov 5, 2003
Messages
445
Location
SJ, Malaysia
There is a new post with slightly improved design here . However, all information in this thread is still all valid.


DIY Li-Ion Rechargeable Battery Charger HOWTO
by D. J. Park
First created on 21 Aug 2004
Last updated on 13 Sep 2004
[Edit] Output capacitor added as per MrAl's reccomendation and led status explain respectively.

This document describes my experience on making a home made DIY (Do-It-Yourself) charger for Li-Ion rechargeable batteries.


Table of Contents

1. About this document
2. Basic design using LTC4054
3. Advanced design ideas using LTC4054
4. Basic design with MCP73843
5. Other designs



1. About this document

This is the first version of HOWTO document to describe my experience of making a home made li-ion rechargeable battery charger. Though there are many commercial chargers already available, the only chargers for li-ion battery are one PILA charger and another one for Nikon camera. When I acquired some 16750, 14430 and 14500 battries, I thought it is about time to make a charger for myself. Candle Power Forums ( CPF ) has been my main source of information, idea and knowledge and so this document is presented on CPF with hope to benefit those who need a li-ion charger but didn't know where to start to make one. (I don't put buying a ready made one as an option since I am a DIY guy.)

There are just too many great members on CPF who inspired me to do this diorectly and indirectly and I find it impossible to mention the names. Ok, just blame me for simply being lazy to write down all the names, some day I might do. Or I may be a chiken not to do so in case I miss someone accidently.

I try not to get into the technical detail of the operation and theory or rule of charging a li-ion rechargeable cell. There are many CPF thread discussing these subjects and I will not try to act as an expert in the area. But I will describe how I made a charger and how it performs. I believe it is more beneficial and practical to most members and that is the main purpose of this HOWTO.

I am focusing on a single cell charger particularly AA (14500) size. But it can be easily adapted to charge different size batteries such as AAA or 17650. This document is not completed and will be updated as necessary -- correction and/or additional chapters, etc..

I do not guarantee the unit I produce will meet your requirement. I disavow any potential liability for the contents of this document. No responsibility is accepted by me for any loss or damage caused in any way to any person or equipment, as a direct or indirect consequence of following these examples. Use of the concepts, examples, and/or other content of this document is entirely at your own risk.

All copyrights are owned by their owners, unless specifically noted otherwise. Use of a term in this document should not be regarded as affecting the validity of any trademark or service mark. Naming of particular products or brands should not be seen as endorsements.

The text contents of this HOWTO is NOT copyrighted except the pictures and the data analysis charts. Whatever I write here is purely reproduction of what is described in the datasheet of the charger control IC I used and I just added my observation only. You are free to quote in any way. But you need my written consent to use the pictures and charts if you want to include in your publishing. Also I always appreciate any correction or suggestion to be made to this document (when I can find the time to do so).


2. Basic design using LTC4054 (Super Simple Li-Ion Charger)

2.1 This design uses a charger controler LTC4054-4.2 (datasheet is found here ). To get a working li-ion charger, only 1 capacitor and 1 resistor is used with optional led and 1 resistor for status indicator.

Here is the circuit I used.
[Edit] This is the 2nd circuit to be used. The 1st one is here.

LTC4054-basic-2.gif


2.2 The power is supplied from a computer USB connector (type A) and the pcb is made to fit in the half of 2AA battery holder. I used Type A male both end and cut to half to get 2 cables. There are 4 wires plus 1 shield. Usually there are red and black wires which can be used to provide the power. Join black and the shield to connect to the ground and the red wire to Vcc.

USB-cable.jpg


2.3 I don't have any means to produce a nice pcb, so I use a perforrated board. This is the bottom view, the small pieces are to hold the sot23-5 chip and convert it to PDIP size.

ltc4054-basic-pcb.jpg


2.4 Cut the small pcb following the patten and solder the chip. This will effectively convert the SMT chip to a PDIP size. 3 leg is left hand side, a short wires are added to the top 2 legs.

LTC4054-pdip.jpg


2.5 This is the finished parts arrangement. The bottom view is flipped upside down.

LTC4054-parts.jpg


I wanted to provide a parts arrangement with bottom wiring, but I don't know how to do. Perhaps someone can help me. Meantime, a text version here, I hope ypu can figure out. From left,

LTC4054 module (X,Y) = 3,2 to 6,4
3.9K ohm resistor (X,Y) = 8,2 to 8,3
LED (X,Y) = 10,1 (cathode) to 10,2 (anode)
330 ohm (or 1K) resistor (X,Y) = 12,2 to 12,4
10uF capacitor (X,Y) = 15,3 (-) to 15,4 (+)

[Edit] Resistor values are slightly changed to be more reasonable operation -- 3.9K=>3.3K 330=>470 ohm. Refer to the advanced section for controlling the charging current.

[Edit] The output 10uF capacitor is not shown in the photograph as it is added in later. But there is enough space to add it, you may need to move all the components 1 or 2 position to the right to make space for it.

2.6 This the complete unit. I thought having the pcb and the battery on the same side will be easier to handle, but the reality is that it is actually harder to put in the battery and take out. So I suggest to use the battery holder where the pcb and the battery go to the opposite sides.

LTC4054-basic-charger.jpg


2.7 I recorded the charging performance of this charger.

LTC4054-charge-14500.gif


The battery was drained to 2.5V on load, rested some time and the open voltage recovered to 3V, then repeat again a few times. With LED as a load, I was not able to discharge below 2.5V. That means, we don't need a low voltage protection circuit to direct drive an LED.

As soon as the charging starts, the battery voltage jump up above 3.5V within a few seconds and I hardly managed to catch a few seconds of trickle charging before the constant current charging starts. It is interesting to know that the charging current starts dropping linearly at 2.4 hours even though the battery voltage is only 4.08V, and the current starts free falling when the battery voltage reaches 4.2V.

The initial trickle charge was done at 10% of the programmed charge current and the charge cut-off at completion was also when the charging current dropped to 10% of the programmed current. This is exactly the way how it is supposed to be done.

For those who wish to measure and log themselves, please note that the charging current can be calculated by Ibat = Vprog / Rprog using Vprog measured between Rprog and ground. The low impeadance measuring equipment can cause erratic behavior of the charger and I suggest to use a 10K resistor in series to measure any voltage especially the battery voltage.

2.8 It is important to know the behavior of the status led. According to the datasheet, this /CHARG pin acts as an open drain with strong pulling to sinks max 10mA during the charging cycle, so it turns on the led. When the charge is complete, it is in 20uA weak pulldown condition and the led is turned off, but the led may show very (really) tiny bit of light if a bright led with clear acryl is used.

[Edit] The led status behavior is corrected with the addition of the output capacitor.

When there is no battery connected, the 1st version without output capacitor will light up the led thinking it is charging possibly due to unstability. With output capacitor added as the current design, the led will produce dim blinking as trickle recharge cycle. This is achieved with MrAl's suggestion of adding the output capacitor.

The datasheet suggests to use a Micro-P to read the pin and determine the status, but I think it is an overkill for this basic design. But I think it can be an idea to have dual color led to indicate the status -- red=charging, green=cmplete.


3. Advanced design ideas using LTC4054

As you see in the picture above, I made one set with 7805 regulator to accept power from a DC supply instead of the computer USB cable. You need to supply at least 6.5-7V to get the 5V output at the regulator. The regulator becomes very hot during the charging without a heat sink. But so far it seems it can take the heat from 10V supply and max charging current of 425mA on that unit.

The datasheet provides many examples of configuration such as input reverse polarity protection, combining wall adaptor and USB power and more.

Since the charging current can be programmed using a common resistor value, it can be a good idea to put a toggle or slide switch to change the charging current for fast/slow charger, or AA or AAA or even 17650 cells.

[Edit] Charging current calculation added. I thought I included in the original post, but can't find now. So I added now.

The max constant charging current can be set by changing the resistor between PROG pin and GND. The value can be calculated using this formula Rprog = 1000 / Ichrg.

Initially I used 3.9K to program, so the max charge current of 1000/3900 = 256mA. Using 3.3K will set the max charge current to 1000/3300 = 303mA and it seems more reasonable for 650-700mAh li-ion cell at 0.5C charging.

Some recommended values of Rprog:

1.3K => 769mA suitable for 16750
3.3K => 303mA suitable for AA
6.6K => 151mA suitable for AAA

I am planning to make a box containing a 7805 with heat sink which will provide power to 2 battery slots, each slot independantly having separate battery holder for AA/AAA and 16750 cells. A 3 points slide switch for each slot to select which battery holder the charging current go and also select appropriate charge programming resistor, one of the point is no resistor which will set the charger to shutdown mode. If a small PIC is used, it may bring moer meaningful status indicator and also software control of the charging current... No time to prepare the circuit diagram, but it is a simple one. Can someone draw?


4. Basic design with MCP73843

I made this charger using MCP73843-4.2 from Microchip before I made LTC4054 version. This IC requires a P-channel FET, and also the current sense resistor needs to be a small value such as 0.15 ohm.

The information of the chip is here and you can request for the sample from the same web page.

Beside I could get the free sample of the chip, the controller also handles the status led the way we expect - on for charging, off for none or complete, blink for error.

Here is the photo of the charger and the charging cycle data chart.

MCP73843-charger.jpg

MCP73843-charging.gif


5. Other designs

I have seen a design on the web using LM317 regulator. It is out of my scope for this document to mention if this is suitable for li-ion charging or not. But I expect quite a number of people are actually using this design or something similar. I even used to charge li-ion cells using a constant voltage / constant current bench power supply to charge te li-ion cells without trickle charge stage.

I have no idea how many or what other designs are available on the net. But I will not list or describe those designs since it is not the purpose of this document to list them, but to provide some ideas for others from what I experienced. Also I don't intent to try all other types to be expert in this area, either.


-- end of HOWTO


I hope you find it helpful to you.

-- dj
 

4sevens

Flashlight Enthusiast
Joined
Feb 29, 2004
Messages
2,876
Location
Atlanta, GA
dj,

Absolutely beautiful! And you always make such nice graphs!

I particularly like the fact that when discharging a Li-ion with
an led, you really can't over discharge the cell because of the V
need to drive an LED. /ubbthreads/images/graemlins/smile.gif

great! /ubbthreads/images/graemlins/smile.gif
 

shiftd

Flashlight Enthusiast
Joined
Nov 17, 2002
Messages
2,261
Location
CA
i like them
totally awesome

so Dj, in your post, you did not mention which one you like the most, the LTC or the MCP.


a bit correction here
on the fourth paragraph of number 1, you said
"I am focusing on a single cell charger particularly AAA (14500) size. "
don't you mean AA?
 

djpark

Enlightened
Joined
Nov 5, 2003
Messages
445
Location
SJ, Malaysia
[ QUOTE ]
shiftd said:
i like them
totally awesome

so Dj, in your post, you did not mention which one you like the most, the LTC or the MCP.


a bit correction here
on the fourth paragraph of number 1, you said
"I am focusing on a single cell charger particularly AAA (14500) size. "
don't you mean AA?

[/ QUOTE ]

Thank you, it is corrected now.

I like MCP because I could get the IC free (sample) and the LED status is more meaningful.

But I still prefer LTC because it uses common value resistor for current programming, does not need FET, and I can measure the current flow by measuring the program voltage.

With MCP, I had to measure the voltage drop at DMM in series to the battery.

Now, someone may want to organize a group buy of LTC4054?

-- dj
 

djpark

Enlightened
Joined
Nov 5, 2003
Messages
445
Location
SJ, Malaysia
[ QUOTE ]
4sevens said:
dj,

Absolutely beautiful! And you always make such nice graphs!

I particularly like the fact that when discharging a Li-ion with
an led, you really can't over discharge the cell because of the V
need to drive an LED. /ubbthreads/images/graemlins/smile.gif

great! /ubbthreads/images/graemlins/smile.gif

[/ QUOTE ]

David, I made a simple data logger using PIC12F675. It collects data every second and I feed them to Excel. When I find time, I will post a HOWTO make the logger.

It is indeed a pleasant surprise to find the white LED not being able to overdischarge a li-ion. But you will want to check red or yellow with Vf below 2V.

-- dj
 

MrAl

Flashlight Enthusiast
Joined
Sep 9, 2001
Messages
3,144
Location
New Jersey
Hello there,

Very nice work dj! Thanks for posting your results too.
I might want to try this type of charger myself.
It would be nice if you looked into the max charging
current too for such cells as the 18650, which can
take 1 amp charging current for quick charge.

Take care,
Al
 

vacuum3d

Enlightened
Joined
Feb 26, 2004
Messages
553
Location
Bay Area, CA
dj,
Thanks to your excellent instructions, I just placed an order for 10 LTC4050. I'm gonna locate the rest of the parts and possibly put a few kits together to share.

ernest
 

djpark

Enlightened
Joined
Nov 5, 2003
Messages
445
Location
SJ, Malaysia
[ QUOTE ]
MrAl said:
Hello there,

Very nice work dj! Thanks for posting your results too.
I might want to try this type of charger myself.
It would be nice if you looked into the max charging
current too for such cells as the 18650, which can
take 1 amp charging current for quick charge.

Take care,
Al

[/ QUOTE ]

The max charging current of LTC4054 is 800mA and the current programming resistor shall be 1000 / 0.8 = 1250 ohm.

When the temperature of the chip goes up, the current will be automatically reduced.

-- dj
 

MrAl

Flashlight Enthusiast
Joined
Sep 9, 2001
Messages
3,144
Location
New Jersey
Hello again,

Thanks dj...this sounds like a part i would be interested
in trying too.

Take care,
Al
 

moraino

Enlightened
Joined
Jan 20, 2002
Messages
447
Location
B.C. Canada
Hi DJ,

It is excellent post. Thanks you so much for sharing.

Now I need a magnifying glass to get busy.

Henry
 

Mgz

Newly Enlightened
Joined
Mar 26, 2004
Messages
59
Location
Alberta, Canada
Hi djpark, do you know that Maxim also offer free sample of Battery Chargers ICs, plz take a look at there http://para.maxim-ic.com/compare.asp?Fam=Batt_Chrg&Tree=PowerSupplies&HP=PowerSupplies.cfm&ln=

Order sample at here http://dbserv.maxim-ic.com/samplescart.cfm?Action=Step1&ln=en

I have no idea about EE, but I think a free sample Maxim IC might further improve the charger, and save us poor flashaholcis couple bucks /ubbthreads/images/graemlins/frown.gif /ubbthreads/images/graemlins/broke.gif
 

djpark

Enlightened
Joined
Nov 5, 2003
Messages
445
Location
SJ, Malaysia
[ QUOTE ]
Mgz said:
Hi djpark, do you know that Maxim also offer free sample of Battery Chargers ICs, plz take a look at there http://para.maxim-ic.com/compare.asp?Fam=Batt_Chrg&Tree=PowerSupplies&HP=PowerSupplies.cfm&ln=

Order sample at here http://dbserv.maxim-ic.com/samplescart.cfm?Action=Step1&ln=en

I have no idea about EE, but I think a free sample Maxim IC might further improve the charger, and save us poor flashaholcis couple bucks /ubbthreads/images/graemlins/frown.gif /ubbthreads/images/graemlins/broke.gif

[/ QUOTE ]

Thanks for the info. You are actually the second person mentioned that IC. Now I look at it again carefully, I find that MAX1555 seems much interesting -- 100mA fixed charging current with USB power and 280mA from DC power -- good for AA/AAA or even bigger cells at slow charging. The price is also cheaper, half of LTC4054.

When I pump up the charging current with LTC4054 and 7805, the heat made the charging current drop and the constant current charging phase was a few minutes only. So slow charging with a low charging current may be better for the battery and charger.

I did get some others samples from Maxim before. They came by the regular air mail and during the last Christmas season, it took a month to reach me sitting in the local post office warehouse. Micochip always send by FedEx and TI send by UPS.

Talking about the sample request, I find that TI is the most generous of all. /ubbthreads/images/graemlins/thumbsup.gif Linear Technology raised many questions through the email and in the end my request for just 2 items was turned down. /ubbthreads/images/graemlins/thumbsdown.gif

-- dj
 

4sevens

Flashlight Enthusiast
Joined
Feb 29, 2004
Messages
2,876
Location
Atlanta, GA
dj,

I'm trying to lift a SMT part from a circuit and was wonder
what was the exact marking on the LTC4054 package.

thanks
david
 

djpark

Enlightened
Joined
Nov 5, 2003
Messages
445
Location
SJ, Malaysia
[ QUOTE ]
4sevens said:
dj,

I'm trying to lift a SMT part from a circuit and was wonder
what was the exact marking on the LTC4054 package.

thanks
david

[/ QUOTE ]

Mine says "LTH7".

Fro the datasheet,

"LTH7" for LTC4054ES5-4.2 (800mA)
"LTADY" for LTC4054XES5-4.2 (800mA, no trickle charge)
"LTAFA" for LTC4054LES5-4.2 (10-150mA)

Regards,

-- dj
 

Klaus

Flashlight Enthusiast
Joined
Sep 6, 2001
Messages
1,998
Location
Germany
Nice one dj - I have a similar one setup for the smaller R123s - CPF-er CM was introducing the LTC4054 a while ago and I named this USB-charger the CM-4054 DIY Kit then - at 1.x USD a pop for the chip the total cost is quite reasonable I think - I think the toughest part for most will be to solder down the pretty tiny LTC chip itself - if a kit would be done this part could be pre-soldered to a small board already like yours - to add the other parts is cake I think - thanks for sharing.

ARC4Li.jpg


Klaus
 

djpark

Enlightened
Joined
Nov 5, 2003
Messages
445
Location
SJ, Malaysia
[ QUOTE ]
Klaus said:
Nice one dj - I have a similar one setup for the smaller R123s - CPF-er CM was introducing the LTC4054 a while ago and I named this USB-charger the CM-4054 DIY Kit then - at 1.x USD a pop for the chip the total cost is quite reasonable I think - I think the toughest part for most will be to solder down the pretty tiny LTC chip itself - if a kit would be done this part could be pre-soldered to a small board already like yours - to add the other parts is cake I think - thanks for sharing.

Klaus

[/ QUOTE ]

Klaus,

Truly I want to say that I was inspired by your charger to make it in the battery holder, and also it is your post which let me know about LTC4054. /ubbthreads/images/graemlins/thumbsup.gif Thank you!

Putting SOT23 chip on a perforated board wasn't that bad, but if a pcb with more area for gound was made, the heat dissipation would be better.

-- dj
 

moraino

Enlightened
Joined
Jan 20, 2002
Messages
447
Location
B.C. Canada
I was very happy that I finally got one done and working.

Unfortunately, I screwed it up big time. It was charging the battery untill 4.05 volt then I took the battery out and hook up the wires trying to measure the amperage.

The meter showed some reading then 0. I smelled the smoke while I was wondering why. I then relized I connected the wire reversed.

Now, I am very upset and feeling tired. I am going to bed now and hopefully tomorrow I can figure out which part burnt out and can be replaced. Wish me luck will you, my fellow CPFer.

Henry
 

djpark

Enlightened
Joined
Nov 5, 2003
Messages
445
Location
SJ, Malaysia
I wonder which one went wrong, could that be your DMM probe fuse?. I wish you a good luck finding. I wasted quite a few 4054 till I got things working fine (or rather I found it working, I was tricked to see the led on when there is no charging).

If you want to measure the charging current, you can follow this portion taken from my original post.

[ QUOTE ]
For those who wish to measure and log themselves, please note that the charging current can be calculated by Ibat = Vprog / Rprog using Vprog measured between Rprog and ground. The low impeadance measuring equipment can cause erratic behavior of the charger and I suggest to use a 10K resistor in series to measure any voltage especially the battery voltage.


[/ QUOTE ]

If you insert DMM between charge output and the positive battery terminal to measure the charging current, you will likely get wrong value. With my DMM, I got just about 30mA with 400mA range and 300mA with 10A range. Dont trust those value and calculate it from the suggestion.

BTW, I requested sample charger controller ICs from Maxim and TI. I will put them up when I get them.

-- dj
 
Top