Li-Ion battery low voltage cut off circuit needed for project

[jeff]

Hello,
This is a current project I'm working on and need some advice.
I have a solar powered Li-Ion battery circuit utilizing a MCP73831 charge management controller chip. This battery is going to be powering a 1watt LED that draws 300-350ma @ 3.5vdc.  As a test I have the LED connected to the battery using a 5 ohm 1/2 watt resistor and a 1000uf capacitor, and even though the current draw is only 150ma (dropping to around 90ma at the end of the run cycle) it still seems plenty bright for the application. This allows the LED to operate for about 11 1/2 hours until the battery voltage drops to around 2.95 volts.  The final addition to the circuit will be to use a LM1084IT-ADJ regulator to keep a constant current on the LED.
The question I have is there a simple circuit (using a transistor?) that I can add to disconnect the LED when the LI-Ion battery reaches a low of around 3.0 volts to protect it?
Any help would be appreciated.

[amspire]

I do not really understand the concept right now. How are you going to control the current to the LEDS's with a voltage regulator with a 1V dropout voltage?

Also, at what voltage does the LED just turn off itself? Many of the Cree leds stop conducting at not much below 2.9V.

[jeff]

I found an article on Instructables that stated the LM1084IT-ADJ regulator could be used to control current also. After looking at it more closely it has a 2.4v dropout voltage. I guess that won't work.
I will try to determine the voltage that the LED shuts down but is that a safe option so as not to damage the battery or cause an unsafe condition?

[IanB]

You seem to have two independent questions. One being current regulation for the LED and the other being low voltage battery protection.

The simplest solution for battery protection is to use a battery with an attached protection circuit. You didn't say what kind of battery you are using, but lithium ion cells in the 18650 size are readily available with electronic protection in the specialist flashlight arena, for example here:

http://www.lighthound.com/AW-18650-Protected-2900-mAh-Rechargeable-Lithium-Battery_p_3671.html

The above cell is from AW, but the same AW batteries are available from other sources too. Other quality sources of protected lithium ion cells are Redilast and Callie's Kustoms.

Regarding current regulation, do you need it? The classic solution for low drop-out linear current regulation is the AMC7135 chip, but if your resistor is working, it is simpler and cheaper.

[jeff]

Yes, I am using a 18650 but it is the unprotected type.  I may get a few of the type you suggested, although I wanted to keep the costs down as it is for a LED lantern project for developing countries. After all is said and done this may be the way to go. 

I did run the LED down to 1.75 volts and it still had a low level of illumination so it obviously doesn't cut off like the Cree's. They were only $.29 ea so they are pretty basic.

Thanks for the help everybody.

[IanB]

IMHO you can't keep the costs down with lithium ion chemistry without endangering safety.

Those protection circuits that are installed in the protected cells can be bought and installed separately if you are creating a complete and integrated device.

However, any device that includes a lithium ion cell or battery must include one of those protection circuits. To do otherwise is very irresponsible. Exploding lithium ion cells are a danger to life and property. You will not be thanked for giving people incendiary bombs to use, however needful they are.

For the developing world it might be better to use a power source that is more readily replaceable like NiMH cells, perhaps? How will someone replace the lithium ion battery when it wears out? Will they just throw the whole device away?

[jeff]

I'm not trying to take any shortcuts as far as safety is concerned. That's why I was inquiring about a low voltage cut off circuit along with the charge management controller chip. Because the locals will be trained in maintaining these lamps, swapping out a lithium battery is possible if we go with something other than a 18650 type. I've seen lamps using  cell phone batteries that are available almost everywhere. It's another option we will be looking at.

[IanB]

The lithium ion protection circuits do more than just low voltage disconnection. Generally they also do high voltage disconnection in the event of inadvertent over charging and high current disconnection in the event of inadvertent overload or short circuit.

It would be more practical to look for such a circuit as a standard module rather than trying to design and build one I think.

A cell phone battery would normally include such a protection circuit inside it, as would for example the kind of battery used in laptops or cordless power tools.

[RJSC]

I've got cells from a laptop with a defective motherboard that I have put on a LED camping light.

All the 6 cells in parallel give me almost 4 days continuous runtime and I charge them with my special R/C aircraft charger that does the required CC/CV charge, but I have no low voltage cut-off.
Since I'm running white LEDs straight from the Li-Ion cells (current seems to stabilize) I guess when it reaches 3v the LEDs will go at least dim, but I'm going to make a circuit like this (not designed by me):



Choose any suitable p-channel mosfet and design the R1 ans R2 voltage divider to reach 2.5v when your battery is at the desired cut-off voltage (3v for a single cell).

[metalphreak]

I looked into this a while ago and found it much more difficult than expected. You can run into issues with the mosfet due to the low gate voltages required since it's running off the battery.

Plenty of places sell the protection PCBs fairly cheap:
http://www.all-battery.com/protectivepcbsforli-ionandli-polymerbatterypacks.aspx
http://www.batteryjunction.com/pcb.html

This one is on many of the small rectangular li-ion packs you get from hong kong: http://www.all-battery.com/ProtectionCircuitModulePCBfor3.7VLi-ionBatteryPack-32088.aspx

If you look at the datasheet it has a schematic and heaps of info: http://www.all-battery.com/datasheet/32088_datasheet.pdf

Getting the ICs in small quantities is almost impossible. For small quantities you are better off incorporating one of the premade PCBs into your design, or removing the ICs off them and placing onto your own board. It's a great IC to use though, you get over charging/discharging protection, as well as current limiting.

[jeff]

Thanks again for the info.
I bought a couple of these http://www.batteryjunction.com/ultrafire-18650.html to play with. Anybody have any experience with using them or know if they are reliable and safe? Some of the brands I looked at were close to $22US so if these $7.95 batteries perform and are reliable, the price is well within the cost parameter for the project. They also incorporate the button top so they are easily replaceable.

[w2aew]

One of my favorite little chips is the old LM10 from National Semi (oops, now TI).  It's VERY low cost, 8-pin DIP, operates with supplies below 1.5V, and contains an opamp, a reference and a reference amplifier.  Awesome little part, and would work great in this application.  Check out some of the circuits on page 18 of this dataheet http://www.ti.com/lit/ds/symlink/lm10.pdf and adjust/tweak for your application.

[IanB]

Thanks again for the info.
I bought a couple of these http://www.batteryjunction.com/ultrafire-18650.html to play with. Anybody have any experience with using them or know if they are reliable and safe? Some of the brands I looked at were close to $22US so if these $7.95 batteries perform and are reliable, the price is well within the cost parameter for the project. They also incorporate the button top so they are easily replaceable.

In general, no they are not reliable. They may work according to spec, they may not. One batch may perform differently to another batch. They may wear out and stop performing after relatively few cycles. It is hard to know exactly what to expect from any individual cell in your hand.

The trouble is those batteries do not have a pedigree. You cannot trace them back to an original manufacturer, with a known quality standard and a proper datasheet describing their performance. Buying them is a bit of a lottery. (Incidentally you can probably get Ultrafire brand cells direct from China for about $4 each. That low price should give you an indication of the expected quality.)

In contrast, the other batteries I mentioned from AW, Redilast or Callie's Kustoms are first quality cells from a manufacturer like Panasonic, made in Japan with a datasheet and a predictable performance.

If I wanted a battery I could trust, I wouldn't want to go near any of the Ultrafire, Trustfire or similar brands of cell.

[metalphreak]

The Trustfire "Flames" batteries are well regarded as the best of the hong kong/china cells. Just be careful where you buy from because there are a few places selling fake ones.

I got mine from ManaFont: http://www.manafont.com/product_info.php/protected-trustfire-18650-37v-2400mah-rechargeable-batteries-flame-2pcsset-p-5560

Low voltage cutoff works fine. My flashlight that I use them in will just cut out suddenly and won't turn on again until the voltage rises up.


All the ones that are $4-8 for a pair, are absolute rubbish. For most, its not that they are unsafe, they just don't live up to the mAh rating on them (not even close).

[jeff]

I'll order a few Redilast 2600's and see how they perform with the 1 watt load. I'm still tossing around the idea of using Ni-MH. 

[sonicj]

i use 14500 trustfire's in my flashlight. definitely some variation from one order to the next... last pair i received have been excellent. previous pair were about a MM too long (which makes a difference in my light).

heres 2 x 18650 protected for $10usd shipped. TrustFire Protected 18650 3.7V True 2400mAh Rechargeable Lithium Batteries (2-Pack)
-sj

[metalphreak]

DX hasn't always been the most reliable for batteries. Check out the latest thread on that item...

[sonicj]

yea, LiIon from china can be a hit or miss type of thing. most of those discharge tests were at 1A, where as the OP is only looking for 350-ish. the difference will affect the peukert at least to some degree. if you don't need the higher discharge rates, $10 for 2 protected 2000 or 2200mAh cells still isn't a bad deal...
-sj

[jjrosent]

I ripped apart a Sparkfun battery once and found this board inside it and have been buying them for my unprotected projects since ($2 for the whole circuit):
http://www.google.com/products/catalog?hl=en&safe=off&authuser=0&q=tenergy+protection&gs_upl=4352l5972l1l6092l15l9l2l0l0l0l193l1171l1.8l11l0&bav=on.2,or.r_gc.r_pw.r_cp.,cf.osb&biw=1233&bih=705&um=1&ie=UTF-8&tbm=shop&cid=7389256412521339460&sa=X&ei=aGIUT5IppKCIAufk6NEN&ved=0CF8Q8wIwAA

But I've also searched for battery protection chips and there aren't many out there for single cell. The few that exist can include fun features like fuel gauges, but are $7 and still need $1.50 in fets and current resistors!
http://search.digikey.com/us/en/products/DS2762BE%2B/DS2762BE%2B-ND/1197518

The lone cheap solution, and the one most assuredly used in that Tenergy board, are this family from sii-ic.com
http://www.sii-ic.com/en/param_chrt.jsp?subcatID=5

I'm working on adding it to my project now, in fact.

[metalphreak]

It's in the PDF datasheet I linked. Seiko S8261 management IC. SII-IC is Seiko Instruments.

The one in particular they use is S-8261ABJMD-G3JT2. 4.28V and 3.00V limits for overcharge and over discharge.

http://www.findchips.com/avail?part=S-8261ABJMD-G3JT2

Only 2 places that have them in stock. Never used either places before.

By the time you order those, and you get the dual mosfet that goes with it, its about the same price as buying a pre-made board. Can't beat economies of scale and bulk buying

[jeff]

Ok,
After weighing the pros and cons I have decided to abandon the Li-Ion and go to Ni-MH batteries. I still need a low voltage cut out circuit so the individual cell voltage only drops to .9-1v.  I think that is a good number to use. Here is a schematic of what was suggested to me. I just finished building it and it hasn't shut down. The battery pack voltage is 2.87 and while the LED is dim it is definitely still on. Anyone have any ideas?
Thanks

[IanB]

You are probably seeing a combination of the leakage current through the zener diode, the gain of the transistor, and the sensitivity of the LED. (You do have the zener diode the right way round?)

I suspect you will get better results with a Schmitt trigger kind of design so there is a hard and fast shutoff when the battery voltage goes too low.

By the way when you have 4 NiMH in series it is probably better to aim for a 1.0 V or even 1.1 V per cell cutoff voltage to compensate for uneven discharging between the cells. It is not nice to reverse polarize a NiMH cell due to discharging a pack too low. The tortured cell will degrade in performance and weaken the whole pack.

[jeff]

The zener arrow points to the B+ as shown in the drawing. I think that is the proper direction. The  Schmitt trigger may work for this project as long as there aren't too many components involved.  These will be assembled by people with limited soldering skills so the fewer parts the better. Could you point me to a schematic?

[IanB]

The zener arrow points to the B+ as shown in the drawing. I think that is the proper direction. The  Schmitt trigger may work for this project as long as there aren't too many components involved.  These will be assembled by people with limited soldering skills so the fewer parts the better. Could you point me to a schematic?

Sorry, it was just a general idea. I don't have an actual circuit to show you. However, in your circuit with the zener diode what happens if you add a resistor between the base of the transistor and the battery negative, to pull down the base voltage when the zener is not conducting? A value like perhaps 1 k. This should help to ensure the transistor is turned fully off and does not conduct.

[jeff]

Now we're getting somewhere.  At 3.99v it's on but at 3.72v it's off. Now I'll just have to find the correct value resistor to turn off at 4.4 as suggested by IanB.  I think I'll swap out the zener with a 3.6v one and then find the right resistor.  Thanks everyone for your help.