A little A123 NicdAdapter Charger

[thanhTran]

I've been using A123 batteries for flying for a while, and like many other, I love these cells. Ive been using a Lab DC regulated power supply to charge thse cells using CC & CV method. But I can't bring that heavy lab power supply to the field or with my heli on vacation so I have to do something ;). I thought a portable & cheap charger for these cells would have some interests, so I hope this project is useful for anyone wants a quick and easy to do device.

I saw the the Lipo dapter http://www.slkelectronics.com/lipodapter/index.htm from SlkElectronics. And being a DIY guy, I want to build one myself. I have small Lipo charger that can charge 14 Nicd cells, so that would be perfect for my A123 cells.

It took a few hours to convert a software that I've written for another project to make it an A123 charger. I have 2 differnet packs on my Swift 16 heli, 5cell, 4 cells to combine it to become either 9 or 10 cell pack to fly. So I need the auto cell count function. The current limit function is provided by the Nicd charger. Infact, these A123 cells can charge as high as 10A so current limit function is not really needed. The auto cell count & voltage monitor to decide when to cut off the charge are done with my circuit & software. That's pretty much it for a functional A123 NiCd adapter charger, I guess. So here it is:

The circuit:
http://tjtrc.com/pictures/A123ChargerAdapter/s_a123charger.jpg

It has a PIC that does ADC and detects the number of A123 cell in the pack. The software can detect up to 10 cells. The voltage reference of the PIC's ADC comes from the Vcc of this chip, so I use a 0.5% LDO regulator. There is an opto to isolate the high voltage battery pack from the PIC. A FET acts like a switch to charge and cut power to the pack. When the battery is first plugged to the circuit, the cell count LED will blink the number of cells it detects. If It sees the cell < 0 and > 10 it will report the error on the error LED. The "charging" LED turns on when the circuit finishes detecting the cell count and turns on the circuit to charge. When the charge is done (3.7volts per cell,) the current will be cut and the full charge LED will turn on.

The PCB of the PIC, ADC and monitoring LEDs:
http://tjtrc.com/pictures/A123ChargerAdapter/IMG_9849.JPG

The finished charger:
http://tjtrc.com/pictures/A123ChargerAdapter/IMG_9847.JPG

Charging a pack:
http://tjtrc.com/pictures/A123ChargerAdapter/IMG_9844.JPG

Here is the Software (http://tjtrc.com/pictures/A123ChargerAdapter/A123ChargerCutoff.hex)

I went on vacation for 2 weeks with this charger and it worked fine for me.

Well I hope it's useful!

Thanks for reading :)

Thanh

[litespeed]

Wow! Your quite talented!

Very impressive indeed.

Tom

[phil007]

NICE...did you make any extras? ;) Can it be programmed to go higher than 10 cells...12-14?

Phil

[thanhTran]

Thanks for the kind words guys. Phil, I didn't make any extra. I may build some more for charging more packs at the same time. The software can be changed for higher cell count easily. The problem with higher cell count 12-14 cells is in the regulator. Anyone knows which regulator works with higher than 40v?

Thanks

Thanh

[Acetronics]

Hi, Thanh

A LM317 HVK ( rated 57v ) as a pre-regulator ( 0.5-1% requirements ! ) ... may be an external ballast for Higher currents.

Alain

[rooster]

Hi Thanh

I agree with the others... excellent idea. I'm sure quite a few of us have nicad chargers that are sitting around that need to feel usefull again!

One question though, on the nicad charger, are you using a specific setting, for example setting it to it's highest cell count, for the different size A123 packs?

Or are you changing the settings on the charger for different size A123 packs?

Thanks

[mmormota]

For high cell count, auto detection is not reliable. A 10 cell pack with low charge can be lower voltage then a 9 cell pack with higher charge. As the cell count is very important, manual setting seems necessary.

the cell count LED will blink the number of cells it detects

A simple method is: set the voltage threshold to a level when the device autodetect always counts the correct cell number or 1 cell lower (but never higher). Then there are 2 pushbuttons:
1: cell count is OK, the user accepts it
2: increase the cell count by 1

[thanhTran]

Hi, Thanh

A LM317 HVK ( rated 57v ) as a pre-regulator ( 0.5-1% requirements ! ) ... may be an external ballast for Higher currents.

Alain

Thanks very much for this info Alain. It looks big & heavy. But using the LM317, I found this one, smaller and cheaper :)
http://catalog.digikey.com/scripts/partsearch.dll?Detail?name=LM317HVT-ND

Thanks, I'll be getting these to try out.


...One question though, on the nicad charger, are you using a specific setting, for example setting it to it's highest cell count, for the different size A123 packs?

Or are you changing the settings on the charger for different size A123 packs?


Thanks :). On my Nicd charger, the only configurable setting is the current. So I just set the current at, says, 5A, and hit the start button. It tries to maintain the 5A current rate. If the Nicd charger is not capable of putting enough voltage, it would not charge, I believe.


For high cell count, auto detection is not reliable. A 10 cell pack with low charge can be lower voltage then a 9 cell pack with higher charge. As the cell count is very important, manual setting seems necessary.
...
A simple method is: set the voltage threshold to a level when the device autodetect always counts the correct cell number or 1 cell lower (but never higher). Then there are 2 pushbuttons:
1: cell count is OK, the user accepts it
2: increase the cell count by 1


Thanks for your insight. I've thought about posibility of lower cell count detected (due to too much discharged on the pack). My thinking is that it's won't hurt anything. For example, if I have a 9 cell pack, and the auto detection detects 8 cells, the charger will cut off really soon when it sees the voltage reaches 8 cell limit. If that happens, we can charge it one more time, and the second time will sure have correct cell count. So far, I've not seen this problem during my 1-month testing. I fly with 9 A123 cells, ESC cuts off at 12v (because it won't work if voltage goes below 12v). And I've never seen a case which the charger detects lower cell count than the pack is. But I charge 4 - 5 cells at a time due to limitation of my NiCd charger. I'll have to test it with a 9S pack. Your solution for an input from user, indeed, a great way to fix the problem if it happens. Thanks

Thanh

[rcboosted]

Any problems charing it without balancing it?

[thanhTran]

Any problems charing it without balancing it?
These cells can take beating and doens't need to balance well. If I don't balance them, I get about 0.1v difference between cells when fully charged. I balance them once a while to have more capacity :)

Thanh

[lowflyer101]

you need make and sell them here for us ThanhTran, like your lipo strobe and alrm kits.. they are great stuff :)

[rcboosted]

Then the next thing we need, it to find the cheapest NiCD charger with the highest amperage output.

[thanhTran]

Then the next thing we need, it to find the cheapest NiCD charger with the highest amperage output.

I don't think we even need a NiCd charger. We can just use any power supply that can provide enough voltage to charge the pack (higher than pack voltage). The NiCd chargers just act like an adjustable current source. Without the NiCd charger in the picture above, the A123 cells will be subjected to as high amp as the power supply can pump to it. It's not a good / safe way to charge LIPO as we mostly want to charge at 1C; but with these A123 cells, we can pump in as high as 10A (limited by the current the FET can provide, and the input voltage, of course). The nice thing about NiCd charger is that it could have an internal voltage boost so that we can charge off a 12V car battery.

Thanh

[Acetronics]

Hi, Thanh

I had forgotten this one ...

http://pdf1.alldatasheet.com/datasheet-pdf/view/28835/TI/TL783.html

Might be enough !!!

Alain

PS: at these voltage levels ... dissipation becomes a real problem and a pre-regulation on mains ( before entering the trafo ) looks rather interesting. ;)

[mrbisset]

Thanh,
the circuit photo in the 1st post isn't working for me? Is anyone else having that problem? I would love to make one of these!

thanks

[thanhTran]

Are you saying the circuit doesn't work?

[mrbisset]

no, sorry, must have been my computer, the photo wasn't displaying. Will have to see if I can build one of these. I have a 10s dewalt that I use with the dewalt AC charger...I would love to get this working with my old astroflight 110 deluxe DC charger!

[thanhTran]

the photo wasn't displaying. Will have to see if I can build one of these. ...get this working with my old astroflight 110 deluxe DC charger!

Don't know why you can't see the picture :). I don't know the Astoflght 110, but to charge 10S, that charger should at least generate 36v.

Thanh

[thanhTran]

I got an email from Achim Jendrian Germany(?) and he made a beautiful PCB board for this. I think it's very nice and should be easy to build. Here is his web site:

http://motorschrauber.de/Motorschrauberde/Adapter/NICD-LIFE-Adapter.html


http://motorschrauber.de/Motorschrauberde/Adapter/2.jpg
http://motorschrauber.de/Motorschrauberde/Adapter/1.jpg

And the Eagle board: http://www.motorschrauber.de/Motorschrauberde/Adapter/Li-Fe-NICD-Lader2.brd

The rest of the infomration can be found there.

Thanks very much Achim :)

Thanh

ps. I will make some changes that would flash the cell count after the charge is finished instead of blinking all over the place :D

[GeorgeP]

Thanh...

Do you know (or have plans for) if a board is available? I did check the above sight last week and didn't see one (I can't read German). I can point to point it, but a board is always nicer.

Also, do you have the newer hex file with the change for the cell count blinking yet? The sight above has your v15.1.07.

tnx
George

[kookboy]

Do you think it would be possible to use a power supply such as this one, change the cell count to 10S @ 7.5 amps using Thahn's gizmo and in effect by-passing a lipo charger ? :cool:

http://www.iotaengineering.com/dls75.htm

http://www.iotaengineering.com/images/dls75_10.jpg

[hermperez]

I use a 36v 400 watt power supply to charge 10s A123 packs.. it has a +/- 10% voltage adjustment that I use to tweak the charge current. When the batteries are low, I lower the voltage to about 35v.. this results in around a 10amp charge current, after 5 minutes or so I increase it to 36-37 volts.. in about 5 more minutes the pack is charged. That power supply you are linking does not have a voltage adjustment.. it would really charge the cells hard when they are low.

Do you think it would be possible to use a power supply such as this one, change the cell count to 10S @ 7.5 amps using Thahn's gizmo and in effect by-passing a lipo charger ? :cool:

[kookboy]

I use a 36v 400 watt power supply to charge 10s A123 packs.. it has a +/- 10% voltage adjustment that I use to tweak the charge current. When the batteries are low, I lower the voltage to about 35v.. this results in around a 10amp charge current, after 5 minutes or so I increase it to 36-37 volts.. in about 5 more minutes the pack is charged. That power supply you are linking does not have a voltage adjustment.. it would really charge the cells hard when they are low.

Are you charging 2 5S packs in parallel at 10amps ?

I was thinking of doing that with my 75A PS and TP1010C charger.

How long is it taking you to charge a 5S at 10 amps ?

[hermperez]

I am charging two 5s packs in series.. the charger can only do 10s packs.. it takes 10-13 minutes to charge the two 5s packs in series.. probably could do it faster but it would exceed the limits of the batteries and the ps.

[Rad Racer]

Thanh, thankyou for the excellent design/project. It works perfect.

I used your project to try out my latest toy. We were getting rid of this CNC board cutting system at work....so I bought it, thinking to modify it for cutting airplane parts. After messing with the board design/cutting software I have reconsidered and decided to keep the whole machine together.

Thought you might like to see some pictures..... enjoy.

Wayne

[thanhTran]

Wow Rad, That looks really nice. I guess it's my first time seeing a CNC machine milling those PCB tracks (I've heard, but never seen how it's done.) Thanks very much for the pictures.

Thanh

[norbique]

Thank you Thanh for making this possible!
Hats off to you!

[Cryofix]

Is anyone making these? I sort of know what to do and can almost read the schematics, but would be too nervous to make one of these.

A123 say 10C so that is 23 amps of charge, theoretically I could charge 1 10S at a time giving me 23 amps x 33 volts = 759 watts which is right about the wattage limit of the power supply.

I can get 14 volts out of the power supply so 14 volts x 55 amps = 770 watts

I know that at 10 amps these 10S packs will charge in 14 minutes.

If anyone can make these I would love 2 of them, if not could someon tell me where to buy the programmed PIC

Thanks,
Ed

[thanhTran]

...
I can get 14 volts out of the power supply so 14 volts x 55 amps = 770 watts

I know that at 10 amps these 10S packs will charge in 14 minutes.
...Ed

This charger requires a power source that has votage as high or a little higher than the voltage of the A123 pack to charge. 10S A123 fully charged is at 36v (charger cuts off at 37v). I'm not sure how you plan to charge your pack with 14volt power supply.

If you need programmed PIC, shoot me an email.

Thanh

[andres.c]

any updates on this project I would like it to charge a 14sp1 A123 pack

I have to split them right now and charge to packs at 7sp1 packs it would be nice to be able to charge them all at the same time

manual cell count selection is highly reccomended plus or minus one cell to compesate for low voltage packs as others have stated

[Richard Ingram]

I wonder if a non regulated power supply would work ? If so you could rewind a microwave transformer and use a bridge rectifier. Depending on the number of turns you could set your desired voltage. I just rewound one for 24 volts to use as a cnc power supply. This would be cheap to do since lots of dead microwaves are lying around with good transformers.

[Rad Racer]

"Is anyone making these? I sort of know what to do and can almost read the schematics, but would be too nervous to make one of these."

I have a couple of extra PCB's and/or completed units. (see the PCB photo's shown earlier in this thread.) PM me if your interested.

Regards,
Wayne

[kgfly]

any updates on this project I would like it to charge a 14sp1 A123 pack

I have to split them right now and charge to packs at 7sp1 packs it would be nice to be able to charge them all at the same time

manual cell count selection is highly reccomended plus or minus one cell to compesate for low voltage packs as others have stated

You might find this thread interesting: http://www.helifreak.com/viewtopic.php?t=43454&highlight=mastech

One of these might solve your problem:

Mastech HY3010E-3 Dual 30V/10A - charge two A123 packs at the same time (up to 8s) or a single pack up to 15s at up to 10A (USD$278 + shipping)
http://cgi.ebay.com/MASTECH-TRIPLE-DC-LAB-POWER-SUPPLY-0-30-V-0-10A_W0QQitemZ120141377186QQihZ002QQcategoryZ32720Q QrdZ1QQcmdZViewItem

[johnandles]

Hello, is the regulator in your plan important ?,is it a low drop out, or is it also a voltage reference?.I tried building your schematic and substituted a 5 volt ldo and i dont count the number if cells correctly. any help?
Johnno

[roewe]

Hey,

i´m Malte from germany and very intresseted on building your adapter.
My problem: I wanna load a 12S A123 Pack.
Do you have any idea how i can change your adapter easily to load more than a 10S pack?
In this thread a 57V transistor is called. Can i exchange "your" one? It´s easy to adapt the software?

best regards
malte

[thanhTran]

Hello,

I didn't see that there were some requests for higher cell counts and questions. Sorry for that.

@KGfly, yes, the LDO regulator is important if you want to charge 2 cells. Its accuracy (5v) is also important too.

@Johnno & Malte, I'll work on an update for 14S and will post back soon.

Thanh

[nsoward]

Are the pictures and code still available for this project?

Nathan

[nsoward]

bump

[thanhTran]

Hi Nsoward, they should be still available now.

Sorry for being away for awhile. I've been really busy moving to a new place.
I got the new board and new firmware ready before the move. But then I was busy and didn't have internet until now.

Here is the new design that supports up 15S A123.

http://www.liglow.com/pictures/A123ChargerAdapter/a123charger_1_1.jpg

Because of the regulator I use (not LVC) the minimum number of cells that it can charge is around 3 cells (4 cells if they are too empty)

The new software (ver 1.1)

http://www.liglow.com/pictures/A123ChargerAdapter/1_1_A123ChargerCutoff.hex

I've done some testing, but it may need more testing. If you try out, please let me know how it works.

The new software slows down the blinking rate so that it's easier to count the number of cells. Reduce the number of counts from 3 to 2 and reduce delay, thus the charge is ready faster. At the end of the charge, it show the cell count repeatedly. The new software is still compatible with the original hardware (just that the older hardware doesn't work with high cell count unless you replace the regulator)

Cheers!

Thanh

[SPINFORCE]

How long times for full charge, when i used this circuit with A123 3S(2300mA) connect by my Battery 's car?

[thanhTran]

Spinforce, this device is not intended to use with car battery to charge A123 since it doesn't have the ability to reduce the charge current to the A123 cells. It's supposed to sit between a NiCd charger and A123. The NiCd charger can then be connected to your car battery. When you do that, the time of charge depends on how empty and which current you set on the NiCd charger. if you charge at 2.3a, it supposed to full in about an hour.

Hope I answer your question

Thanh

[wwldo]

Does anyone have the circuit boards for sale...?

If not what file and software do I need to print the board layout to make myself..?

Thanks,

Larry

[thanhTran]

I don't have the whole circuit. My set up is like one of the pictures in the thread which I used one PCB to hold the PIC, regulator, and LEDs. Another board to hold the FET and the opto.

I use express PCB (free) to make the board.

Post #19 http://www.rcgroups.com/forums/showpost.php?p=6798240&postcount=19 has very nice board which should work with Eagle Layout

[wwldo]

I don't have the whole circuit. My set up is like one of the pictures in the thread which I used one PCB to hold the PIC, regulator, and LEDs. Another board to hold the FET and the opto.

I use express PCB (free) to make the board.

Post #19 http://www.rcgroups.com/forums/showpost.php?p=6798240&postcount=19 has very nice board which should work with Eagle Layout

Ok thanks......will check it out

A couple questions.......

What are the specs on the Leds..?

What are the specs on the veriable resistor, and the procedures for adjusting the voltage..?

Thanks,

Larry :)

[thanhTran]

Hi Larry
Any small LED should work. The ultra bright one will be brighter.
If you could pick out any wire wound variable resistors that would be great. The one I was using was just a small thumb wheel pot. Before putting the PIC into the circuit, I would adjust the voltage on pin 2 (the output) of the regulator to have 5.0v. The PIC uses this voltage as the reference for the ADC.

Thanks

Thanh

[norbique]

Hi Thanh,

Spinforce's question made me wonder, if the FET (or paralleled FETs) can handle the high current, we could use the car battery (AKA ZIP charge) charging method. What do you think?

[thanhTran]

Hi Norbique,

Sure, it should work. Adding more high amp FET would allow high current and Zip charge should work. Only the first version of the hardware would work with 2 to 3 cell though.

Best regards

Thanh

[norbique]

Why do you say the first version of the hardware would only work with 2-3 cells? don't get it. I thought the first version would work up to 10 cells in series?!

Or did you mean that Zip charging works only with 3 cells in series?

[thanhTran]

Hi Norbique,

There are 2 hardware versions and 2 software versions. The first software version works with up to 10 cell, but the first hardware version works up to the max voltage that the regulator can work with. The MIC2950 in the first hardware versions works fine with voltage as high as 30v. 30v is about 8 cells. The first software works with up to 10 cell if someone replace the MIC2950 regulator with any regulator that can works with up to 40v (which is plenty out there)

The second version of software and hardware works with 15 A123 cells. but the second hardware version has the limit of 4 cell minimum.

Hope that makes sense

Thanks

Thanh

[norbique]

Yes it makes sense, thank you!

[Joat]

I know this kind a bring up from the dead, but I was wondering if anyone thinks I could convert a "power tool nicad" charger with this thing ? I am only gonna run 4s, the chargers output is 3 amps max - ctc318- snap on charger