Best homemade lipo charger design

[av8n]

I have seen a number of homemade lipo chargers and was hopeing to get some opinions on which one seems to be the most reliable, failsafe, fastest.
So far I have built the Scott Henion designed one, used it for about six months with great success. I have plans for the charger that incorporates two LM317's, one that uses a L200. I'm sure there are others.

Any opinions or facts appreciated.

Paul

[Happy Hobit]

Hi Paul,

Well, I have 5 rules for charging a LiIon battery

1.) Limit voltage to 4.2 volts per cell,

2.) Limit current to 1C or below. Charge a 500 ma hour battery at 500 ma MAX. (Or manufacturer specks)

3.) If the battery voltage is below 3 volts per cell, charge at C/10 until the voltage is above 3 volts per cell but below 4.2 volts per cell. (Lithium Ion Batteries should NOT be discharged below 3 volts per cell.)

4.) Do not charge a battery if the ambient temperature is above 45 degrees C or below 0 degrees C. (Or manufacturer’s specs).

5.) Charge in a safe location. (A damaged Lithium Ion Battery may burn violently.)



All the homemade charges you listed all follow rules 1 & 2 but none of them follow rules 3, 4 &5.

It is the users responsibility to enforce these Rules. Some commercial charger may enforce rule #3.

Jay

[JRuggiero]

Happy Hobit is correct on all points. That said, I built a charger using the L200 chip, and it works fine. It's simpler than the 2 LM317 design.

As for HH's Point 3: I check my pack with a voltmeter to determine the state of charge before charging. If it ever got below 3 volts per cell (it never has), I'd do the C/10 charge. Re Point 4: ambient temp check is not hard to do. Point 5: I charge my packs outside on the driveway.

Jim R.

[Foamaholic]

Good idea. Not in the sun I hope.

[JRuggiero]

Happy Hobit might also have added a 6th point: a simple charger can't determine the state of imbalance among cells in a pack. Would charging at C/10 reduce this imbalance?

[Happy Hobit]

Hi Jim,

(I could add this as a rule but it’s getting kind of wordy)

6.) If there is a voltage imbalance in a series pack, the only safe thing to do is parallel charge the cells. Series charging an ‘Imbalanced Pack’ will result in overcharging the ‘good’ cell and LiIon cells don’t handle this well.

Here’s an interesting article I got the graph below from.

http://www.planetanalog.com/printableArticle.jhtml?articleID=12802557


When NiCad and NiMh cells are fully charged they dump the additional current as heat.
LiIon cells just overcharge. See fig2.

Jay

[KillerWatt]

So based on the links various charger descriptions, most if not all of the current DIY LiPo chargers are of the simplest design, slowest to charge, Current Limited + Constant voltage types that taper-off the charging current steadily and well below the battery's allowable charging current as the battery's voltage approaches it's finishing voltage......This " reduced and tapered-off charge current before 4.2 volts per cell" effect is probably due to the rather high ohm value, fixed, series connected resistor that is calculated to determine the charger's max charging current ..It probably doesn't harm anything, but there's got to be less total charge ma's inputed to the battery over same 2 hrs when compared to chargers that automaticly reduce their internal impedance to a much lower value and remain at a true CC (like C1) till the 4.2 v is reached........ kw

[soren]

Hi

The L200 works fine.
Take a look for a diagram at my homepage:
http://users.cybercity.dk/~ccc16084/
follow the "electronic" link

Soren

[KillerWatt]

The L200 by itself wired-up in the CV and self-current limiting 2 amp applications without any < 2 amp output series resistor added does work great....but in the smaller LiPo battery size, say the 340 ma for instance, the required added current limiting series resistor of over 1 ohm is going to make keeping the charging current anywhere near 340 ma even at 4.0 cell volts impossible to achieve....The initial 340 ma charging current at 4.00 cell volts is probably going to have dropped to about 200 ma, not exactly following the full 340 ma CC charge current all the way to 4.20 cell volts charging curve ......... kw

[Happy Hobit]

Hi KillerWatt,

The article was interesting, but not particularly well written.

The writer of the article tried to describe different charging techniques but to my knowledge there is only One CORRECT way to charge a LiIon (LiPoly) battery. (See rules above)

I can’t speak for the L200 regulator, although I think it works the same, but the charging curve for the Dual LM317 is below.

The charger outputs MAX current (1C) (Rule #1) until the voltage reaches 4.2 volts. Then rule #2 kicks in.

The battery is not truly at 4.2 volts now. It’s probably at 4.1 or 4.0, but in order to pump in 1C current the charger voltage would have to be higher.

The current starts dropping now as the battery fills up.

This is the only way I know to charge LiIon cells.

The most accurate description would be ‘Current Limited / Voltage Limited’.

The battery has a natural internal resistance (ESR) not something added


P.S. Here’s Part 2 of the article. http://www.planetanalog.com/printableArticle.jhtml?articleID=12802558

P.P.S. Based on the article writer the battery is fully charged in 60 min. (Charge current drops below .1C)

Jay

[zagisrule!]

What about this:

http://www.rcgroups.com/forums/showthread.php?t=212232&highlight=zagisrule

My design uses computer-controlled switching to determine state of pack, then delivers a CC charge, switching to CV at the manufacturer's suggested switch point. Turns off CV charge when the battery peak is reached.

I have gone no farther with the design yet, no need to here and not enough time to spend.

For some reason nobody really had any interest in the unit, while both I and my friend ("Roger Doger" on the forum) use the chargers and find them to work just like they should.

Here is a picture of my buddy's unit, up and running.



-Matt

[Happy Hobit]

Hi Matt,

I read your ‘Superb DIY Lithium Charger’ thread and it’s a nice design, but other that shutting off at end of cycle, there’s not much new.

And not everyone feels comfortable with microcontrollers. There’s a pretty steep learning curve. Learning a programming language (even basic), learning microcontroller architecture and then flashing the chip. (I’ve been playing with micros for 30 years and I sometimes forget this to.)

A few suggestions.

Get rid of the LM317 and the LM350 and use PWM to control the voltage and current.

Check-out the ‘LM1577 Step Up Voltage Regulator’. It can handle 3 amps and step-up 12 volts to 20+ volts.

Add a ‘Low Voltage Pre charge’. If the battery voltage is below 3 volts per cell the charger starts at C/10 until the battery voltage is above 3 volts.


Jay

[soren]

Hi Matt,
Check-out the ‘LM1577 Step Up Voltage Regulator’. It can handle 3 amps and step-up 12 volts to 20+ volts.
Jay

Use the LM2587 step up voltage regulator - it can handle up to 5A - the current in the regulator is not the samte as the output current...

Soren

[zagisrule!]

That was my plans for the new design, using MOSFET's to switch PWM to control voltage and current. I am trying to figure out how to do a reliable auto-cell count, but have not come up with anything yet. The start-up trickle charge would be easy...

Step up might be easier because I have found some great transformers that could do 40V at several amps from 12V...perhaps too much voltage? Because if the PIC or MOSFET's were to fail that would start a big fire :D

Wait....current can be controlled in with a feedback loop consisting of shunt resistor and ADC. If the PIC senses less than requested current, simply increase duty cycle a small amount until the current requested = the current delivered. If current is too high, then the charger lowers duty cycle to maintain requested current.

Constant voltage would just sense voltage output with ADC and adjust duty cycle accordingly, to deliver the voltage requested.

Hahaha! I have it figured out..we just need some active feedback and then we don't have to worry about math and we will have a spot-on output! Bam! :cool:

LCD is very easy to interface, so that can be included easy. Would a charger such as this sell in kit form? So people don't have to worry about the learning curve associated with MC's? Or just pre-programmed PIC's and PCB's?




-Matt
....thinking now....

[zagisrule!]

BTW: Soren, are you that same "Soren" on diyaudio.com? I knew I saw that name somewhere else! :) RC Aircraft and Audio...the two best hobbies on the face of the earth! :)




-Matt

[soren]

BTW: Soren, are you that same "Soren" on diyaudio.com? I knew I saw that name somewhere else! :) RC Aircraft and Audio...the two best hobbies on the face of the earth! :)
-Matt

...Sorry thats not me.
I love: RC aircraft, RC boats and flyfishing for Atlantic Salmon....the three best hobbies on the face af the earth..

Soren

[Happy Hobit]

Hi Matt,

I am trying to figure out how to do a reliable auto-cell count, but have not come up with anything yet.

You can make a fairly accurate cell count guess for 1, 2 and 3 cell packs, but it’s only a guess. A pack that has been discharged to below 3 volts per cell will fool you. You’ll need to provide a PB for manual cell selection. If you go above 3 cells it becomes pure speculation


LCD is very easy to interface, so that can be included easy. Would a charger such as this sell in kit form? So people don't have to worry about the learning curve associated with MC's? Or just pre-programmed PIC's and PCB's?

An LCD display would be nice. You could display a menu to select voltage, capacity and start the charging cycle. Then during the charging cycle it could display Voltage, Current and the Total Milli-Amp hours that have been pumped into the battery.

Will it Sell? Yes, but how many. It has to do something that the dual LM317 charger doesn’t and somebody has to want those features.

Programmable for Packs from 1 – 4 cells and 50 – ?? ma (1200, 2400, 3000)
Charge 3 or 4 LiIon packs from? (a 12 volt car battery)(110 VAC).
Display battery charge statistics.


….Keep thinking…

Jay

[zagisrule!]

Charge current could be infinantly variable from say 10mA to 5A.

Depending on the supply, packs of more than 6 cells could be done.

LCD could do all that, perhaps even display the time elapsed and estimated time remaining?

I would aim for a price of $50 for a kit.





-Matt




-Matt

[zagisrule!]

Schematics and source codes would all be open-source and free of course. I just would like to make a few $$$ for the whole endeavor, to buy more planes and batteries to charge ;)





-Matt

[Happy Hobit]

Hi Matt,

Charge current could be infinantly variable from say 10mA to 5A.

Depending on the supply, packs of more than 6 cells could be done.

LCD could do all that, perhaps even display the time elapsed and estimated time remaining?

I would aim for a price of $50 for a kit.


That would be an impressive charger for $50. I’d buy one. :)

Charging a 6S packs would be nice, but add the boost circuit so you can do it from a 12 volt car battery.

5 amps might be a bit high for the 6S packs. With a boost circuit it would take over 10 amps input. (as Soren noted).

Jay

[zagisrule!]

Of course input current increases as a function of the ratio of voltage increase assuming 100% efficiency. It would be roughly 85% efficient, so it would take even more amps to produce the power.

It would be better to over-design than to do something that does 2A or something and be wishing for more. 5A is attainable and a good goal. I am sure that as new cell technologies arrive, charge currents will increase and that 5A may eventually be hit. Point is that it is better to over-build than under-build in this case. If you don't use 5A at the ouput, you use less than xxA. So, if your 12V source is the weak link, limit charge current of the charger.




-Matt

[quax]

Hi,

I can add a simple design with the MC34063.

http://home.versanet.de/~b-konze/lithium_loader/lilo_bko.jpg

My homepage ( http://home.versanet.de/~b-konze/ ) is in German but the schematic should be international.

Very important is a good output voltage control. I set it to 4.15V as good as I could. R1 and R5 limit the INPUT current. The values have to be evaluated with the power supply that is used with the loader.

The good thing about it is the efficiency. I have 13V input voltage, 600mA output current and nothing becomes hot. My Duo-LED starts with mixed green/red and is saturated green, when the cell is full.


QUAX

[Happy|Harry]

I would aim for a price of $50 for a kit.

-Matt


this is something that i would buy too! :cool:

phil

[zagisrule!]

Great, I am almost done with hardware design! :D

Software comes next...should be straight-forward.

I also have transformers in stock and a schematic for a SMPS to boost the 12V input to roughly 25V or a little more so to charge 5S packs from 12V. Current available would be several amps. This would be an "add-on" PCB.

I also plan on incorporating Ni-MH and Ni-CD, so all types currently flown could be charged.

-Matt

[Soeren]

Hi HH,


[...] but the charging curve for the Dual LM317 is below.

Just wondering... Why use two, when a single one (plus a small signal BJT) is enough ?
Far cheaper, as it only needs one heat sink :)


[...] The most accurate description would be ‘Current Limited / Voltage Limited’.

Normally, this constellation is called "current limited constant voltage" (or sometimes "constant voltage/constant current" but that is not a very correct term from a physical viewpoint :) )


The battery has a natural internal resistance (ESR) not something added

Err...
"ESR" = Equivalent Serial Resistance is a term used about inductors and capacitors. Talking about internal resistance in battery cells, the correct term is "Ri"


Regards,

Soeren

[David T]

For what its worth I've been using the L200 design with an LM2577 voltage boost circuit to charge 3 cells from 12v. In theory it can do up to 9 cells from 12v although it would probably need a higher spec regulator to get a decent charge current at higher voltages. Photos and design here http://www.flyelectric.ukgateway.net/lithium.htm#charger
David.

[soren]

For what its worth I've been using the L200 design with an LM2577 voltage boost circuit to charge 3 cells from 12v. In theory it can do up to 9 cells from 12v although it would probably need a higher spec regulator to get a decent charge current at higher voltages. Photos and design here http://www.flyelectric.ukgateway.net/lithium.htm#charger
David.

...or look here: http://users.cybercity.dk/~ccc16084/ follow the electronic link, and then the LIPO link.
There you can find another L200/LM2577 charger this one with a LED showing the charging status.

Soren

[mattijs321]

zagisrule where's the update?

I would really like to make one of your chargers!

[KenSt]

I built a dual charger using two Texas Instr. bq2057c chips. Satifies rules 1 thru 4 from Happy hobit, to which I also agree. Two circuits easily fit on a radio shack board. Parts are available thru Digi-key. I believe this chip is designed for cell phones and therefore is quite small. I soldered them to Surfboards for easier connecting to the Radio Shack board. Features include built in voltage limit (4.2), LED "i'm charging" light, .3 voltage drop out for low heat dissipation, only 7 external components, Etc... Sorry, i can't do pictures and haven't posted drawings, or schematics... haven't learned yet. I used the circuit from the Texas Instr. Application info, page 7. from their site, www.ti.com. I didn't use the temp. control, but it's there if you want it. :) Rule 5: I use a gutted air conditioner connection box. It sits on my metal cellar steps which leads to the outside. No combustibles inj the area. The box is available at any hardware store for about $10 :D

[Happy Hobit]

Hi Soeren,

Sorry about being so slow to respond but I was on vacation.


Just wondering... Why use two, when a single one (plus a small signal BJT) is enough ?
Far cheaper, as it only needs one heat sink :)
Scott Henion design? The regulators can be insulated (See http://www.rcgroups.com/forums/showthread.php?t=130567&page=21&pp=15 post #308.

All of the Lithium Ion chargers have advantages and disadvantages. The ‘Dual LM317 Chargers’ use simple readily available components and the circuit is easy to understand. It’s simple. (KISS)

I personally don’t think much of charging batteries over 1 Amp with ‘Linear Regulators’ (Too much wasted heat). I think ZagisRule idea of a microcontroller based switching charger a fantastic idea. In fact while on vacation I started designing one using an ATTiny-15 microcontroller. I'll probably order some chokes next week.

Normally, this constellation is called "current limited constant voltage" (or sometimes "constant voltage/constant current" but that is not a very correct term from a physical viewpoint :) )
I like to call it ‘Current Limited and Voltage Limited’ because that is what the batteries require.


Err...
"ESR" = Equivalent Serial Resistance is a term used about inductors and capacitors. Talking about internal resistance in battery cells, the correct term is "Ri"
Thank you for that correction. I had been talking with someone about overheating capacitors in a power supply and ESR stuck in my head.

Jay

[zagisrule!]

I have been busy with work, and school starts again Monday here (I'm a Senior in HS now! Finally!) so I have not gotten much done. Hopefully I can do some work on it but at the moment I have other things going on.

I still use primarily Ni-cd and Ni-mh so compatibility for those cell types would be built in. Once the hardware is designed the only thing to make the charger work for those types of cells is the firmware.

I have some transformers for SMPS's that we can use for supplies for our chargers when we charge larger cell counts. They were custom made for some car-amps and I have a ton of extras we can use for power supplies. I have smaller units good for 25V or so from 12V, as well as some other which will do 35V from 12. The small unit can push maybe 150W with the larger unit easily capable of 500W. They must be run in closed-loop regulated mode with feedback from the final DC ouput going back into the chip to adjust PWM to keep output where it should be. We can use the TL598 to do that job...




-Matt

[rroback]

good luck senior yet, I just graduated last june.. so I'm off to college. I'm such a electronics novice, I'm having trouble following your plans ( or any for that matter)...

Rhett

[Larry3215]

I want one too!

Larry

[Hovertime]

I built a dual charger using two Texas Instr. bq2057c chips. Satifies rules 1 thru 4 from Happy hobit, to which I also agree. Two circuits easily fit on a radio shack board. Parts are available thru Digi-key. I believe this chip is designed for cell phones and therefore is quite small. I soldered them to Surfboards for easier connecting to the Radio Shack board. Features include built in voltage limit (4.2), LED "i'm charging" light, .3 voltage drop out for low heat dissipation, only 7 external components, Etc... Sorry, i can't do pictures and haven't posted drawings, or schematics... haven't learned yet. I used the circuit from the Texas Instr. Application info, page 7. from their site, www.ti.com. I didn't use the temp. control, but it's there if you want it. :) Rule 5: I use a gutted air conditioner connection box. It sits on my metal cellar steps which leads to the outside. No combustibles inj the area. The box is available at any hardware store for about $10 :D

Ken, please give us some more info, this looks very interesting, small part count project!! Plus it uses chip special engineered for this purpose .

[KenSt]

Hi hovertime! Too many projects- not enough time! :( I am working on the schematic in AutoCad so I can post it. I'm also working on a components layout drawing since I can't do pic's. One of the guys on the forum said he could convert it to a PFD file for better viewing.

The nastiest part of the circuit is soldering the tiny chip onto the Surfboard. The EE that helped me with this at work did the soldering for me under a magnifying light. The chip is about the size of a tick. :eek: The rest is simple resistors and jumpers. The resistor to control amp output is a low non standard value. Digikey is essential to this project. That's were the EE got the chip, Surfboard, and resistor. Other stuff is Radio Shack. Stay tuned, I''ll try to get this done soon and post a thread. Thanks for patience. :)

[Dr. Robert]

Hope I am doing this reply right! I design computers; I only use them under duress!
I have been using a charger per cell set for several months now, and most LiPo battery people that I have talked to concur that this is the best and safest way to charge a Lithium battery. Some LiPo manufacturers DO NOT RECOMMEND charging more than 3 or 4 LiPos in series. But multiple chargers is not the cheapest way to do the job!

It takes a + and - lead from each cell set (1, 2 or 3 lithium batteries in parallel = one cell set) connected to an electrically independent charger which is a max current, max voltage (4.2 volts) design. The various designs posted on this forum will do fine, just needs one of these, set to 4.2 volts max, per cell set. A two voltage regulator circuit will need 10 voltage regulators to charge a 5SXP LiPo pack.

I am currently running the 4th generation of this circuitry. The parts cost w/o PCB is <$10 per cell set, including a $1.25 wall wart for each charger. My PCB cost is $7 per charger.

If interested in my design, my next post will be the schematic, parts sources, PCB layout and pix of the chargers.

Robert

[KenSt]

Do it Dr. :) ! I posted my charger in another thread called "My Lipo charger". It too charges cells individually using a lipo battery management chip to control everything :D I'm not thrilled about spending big bucks for chargers with adjustable current and cell counts or automatic detecting. Like you, I want 4.2 volts, period! My current can only be changed my replacing a resistor. Nothing to adjust means no mistakes. No fires at my house can be blamed on overcharging or cell imbalance!

[Dr. Robert]

The circuitry is similar to those published in the National Semiconductor Handbook in 1980. I changed the voltage regulator to a newer 5 amp Low Voltage Dropout IC available from Digikey for $1.80 each. A 7.5 VDC 1.3A wall wart (p/n PWR1021) for each cell set charger is available from BG Micro (1-800-276-2206) for $1.25 each. I had the single cell set PCBs made by ExpressPCB for about $7 each in small quantities (20). A PCB with 6 of these chargers on it would cost about $25 in small quantities (10).

Each LiPo charger is set by the 50 ohm potentiometer to 4.200 volts output with no load. The current regulating IC is set to 1.25 volts by using a 1 ohm resistor (or two 2 ohm resistors in parallel.)

My operational LiPo charging system is a 6 cell charger with separate leads and connectors for 2 cell, 5 cell, and a pair of connectors for 3 cell LiPo packs. My LiPo packs are from 1200 MAH to 4500 MAH, for motors and receivers. I use Molex connectors, but anything handy will do fine. The Tanic packs with their JST connectors should be ready to go; I am waiting to try one here. These linear regulators get quite hot, so I have a small 12V fan from an old PC switching power supply blowing over the boards to keep them cooler. The 6 wall warts plug into an 8 socket computer power strip from Office Max.

This is the 4th generation of this charger over the last 6 months and will top off and balance a LiPo pack in 1 to 4 hours, depending on the size of the pack. Just plug in the LiPo pack and each red LED comes on very bright, indicating about 1.25 amp of charge going into a cell set. When the voltage at the cell reaches 4.200 volts, the current will decrease and the LED will start to dim. When the LiPo battery is fully charged, the LEDs will be almost but not quite completely out due to about 5 MA drawn by each 4.200 regulator. Avoid charging multiple packs simultaneously on the same cell charger unless at the same state of discharge. When fully charged, a LiPo pack can be removed and stored, or left on the charger until used. Don't try to economize on wires and connector! Use 2 wires, + and -, to each cell set otherwise a failed line or connector could put you back to the old cells in series situation.

Robert

[Hovertime]

I posted my charger in another thread called "My Lipo charger". It too charges cells individually using a lipo battery management chip to control everything :D
If someone is trying to find it - here is the link : http://www.rcgroups.com/forums/showthread.php?t=279926

[jeffs555]

I have purchased several different aftermarket cellphone car chargers at a local $1 stores. Some were just a dropping resistor and are worthless. Most of the others have the same MC34063 chip and basically the same circuit that QUAX posted on page 2 of this thread(without the end of charge led). They were set for about 1 amp and 6 volts out. The nokia phone uses a single LiIon cell, so apparantly has a 4.2v regulator in the phone. They use no heat sink, and barely get warm charging the cellphone at 1 amp. I plan to use several of these $1 chargers for my LiPo packs by changing the current sense and voltage divider resistors. My current LM317 based charger has a large heat sink, and still gets very hot.

Mouser has an NJR dip version of the MC34063 for 64 cents, so this looks to be a fairly cheap charger to build. http://www.mouser.com/index.cfm?handler=displayproduct&lstdispproductid=340071&e_categoryid=265&e_pcodeid=51313

[quax]

The circuit of my MC34063 LiPoly loader is indeed a standard application.

The only changes I made are:
1.) A very high output capacitor for little voltage ripple and better end-voltage control.
2.) A transistor powered LED as load control.
3.) Standard resistor values for voltage adjustment.

I love simple designs :) .

QUAX

[Andy2No]

Hi,

I can add a simple design with the MC34063.

http://home.versanet.de/~b-konze/lithium_loader/lilo_bko.jpg

My homepage ( http://home.versanet.de/~b-konze/ ) is in German but the schematic should be international.

Very important is a good output voltage control. I set it to 4.15V as good as I could. R1 and R5 limit the INPUT current. The values have to be evaluated with the power supply that is used with the loader.

The good thing about it is the efficiency. I have 13V input voltage, 600mA output current and nothing becomes hot. My Duo-LED starts with mixed green/red and is saturated green, when the cell is full.


QUAX
Does anyone have a copy of that schematic? It's not there any more.