$10 1200mA HIGH QUALITY Li-Ion charger, want one?

[VanIsleDSM]

Because of the lack of a suitable charger for a suitable price I'm building my own.

The Ultrafire chargers are cheap enough, but it shows.. they only charge at 600mA (or is it 500mA?) wayyyy to slow for 18650s. And I've heard many complaints about them over charging batteries right up to 4.35V. The charger I am making does not do this. There is no endless trickle charge. The charging will completely halt when finished.. then the IC monitors the battery's voltage.. once it starts to fall the IC will automatically start recharging the battery.. and then stop again.. Keeping it topped up in an intelligent way.. unlike the endless trickle charge you never have to worry about how long you leave your batteries in the charger.. they'll just be perfect every time you take them out.

I've finished my battery holder that will hold 18650, 17670, 14500, 16340, and any other battery not longer than 70mm and not wider than 21mm. Since I don't have any 16340s or 18650s yet I made some wooden dummies for fitment.

http://i105.photobucket.com/albums/m203/shaft0/HPIM1405.jpg

http://i105.photobucket.com/albums/m203/shaft0/HPIM1406.jpg

The ICs I'm using for these will charge to 4.2V within 0.5%. (easily modified to charge to 4.1V if desired)

For heavily depleted cells (under 2.9V) the charger will "precondition" them with a trickle charge of 10% the fast charge value... This is much better for your cells than hitting them with 1.2A when they're low on power.. also creates a lot of heat.

Which leads me to the temp monitoring. Each cell will be individually and continuously monitored for temperature. If any cell is bellow 0C or above 50C charging will stop.. and automatically start once temp is back within parameters. This, as far as I am concerned, is an indispensable feature for anyone who doesn't want to babysit their charger.

There will be 2 status LEDs. 1 to indicate charging, and blinks upon completion. And another to indicate a fault situation.

There's also a safety timer for the whole operation. But I'm not too certain what to set it to yet. at 1200mA how long will it take to charge an 18650? 3 hours, maybe 4? maybe I'll set the safety to around 5 hours, what do you battery gurus think of that?



So. My whole deal here is this:


If I purchase 1-24 of these ICs I need to make the charger they're almost $3 each.

If I purchase 25 they're $1.70!! so.. if I go it alone I'll just grab 6 and make myself a 6bay charger... but I thought there would be some other folk around here that might be interested, so I thought I'd throw this out there.

If I can get at least 12 charger bays accounted for then I can let them go for $10 each.

So is anyone interested? each charger bay is completely independent from one another. And for power supply I'm going to add a molex connector so the charger will plug right into your computer's power supply. (or you could buy a used computer PSU for ~$10 and just use it as a desk power supply too)

One thing I haven't decided on yet.. and need a little help to do so... is how I want to be able to adjust the charge current. A pot is a simple way to go about it.. make it adjustable down to 500mA or so.. but then you won't know what the charge current is exactly unless you test it... So I'm kinda thinking just a little jumper to take it from 1200mA to 500mA would be good. Then removing the jumper all together would knock it down to 100mA (if you ever wanted a charge that low for some reason)

So.. if there is some interest in this I'll order the parts right away and hopefully have everything built within a couple weeks.

And just to be clear, this isn't for profit.. with the electronics, and the actual battery holder itself (not to mention my time machining them on the lathe) the price comes to $9.67.

If you want all the info on the IC chip I'm planning to use for the charger, the datasheet is here:

http://ww1.microchip.com/downloads/en/devicedoc/21893c.pdf 0.11751955


The list:

michelkenny - 2
zapper - 2

4 of 12 confirmed.


Over Sized Images Removed

 

[MorePower]

There's also a safety timer for the whole operation. But I'm not too certain what to set it to yet. at 1200mA how long will it take to charge an 18650? 3 hours, maybe 4? maybe I'll set the safety to around 5 hours, what do you battery gurus think of that?

At 1200mA, an 18650 should be done charging by 2.5 hours. I believe typical capacities are around 2400mAh, nominally.

 

[VanIsleDSM]

That quick? right on.. I knew the capacities were ~2400mAh, but I thought the charging was less efficient than that. I suppose not.

Thanks for the reply.

 

[jtr1962]

One thing I haven't decided on yet.. and need a little help to do so... is how I want to be able to adjust the charge current.

I would probably use a header which allows several positions for a jumper, with each position connecting a different value resistor to Vss. I'd say a good set of values in mA might be 250, 500, 750, 1000, and 1200, with no jumper of course being 100 mA. This is a relatively low cost way to allow as many charging currents as you want, although of course I don't really see much need for more than about 5 or 6.

 

[Sampolainen]

I'd be interested. How much for shipping to Finland? And what about payment, paypal?

 

[Mr Happy]

Just one little suggestion about the charging indicator light: It's a very common user interface pattern to have the indicator light blinking while "in transit" and steady when complete. So I'd suggest reversing your pattern so that the light is blinking during the charge and steady when charged. Another "mind's eye" picture for this is that blinking lights indicate activity. Compare for example the disk drive light on a computer or the lights on a modem.

 

[Handlobraesing]

I read this somewhere... I think it was the Battery Handbook.

You can't really fast charge a lithium ion if you want a full charge. You'll hit a brickwall that you can't get past.

Li-ion is charged in CC/CV fashion, so you start charging at xA, then once you hit 4.20v per cell, you hold it at a constant voltage. Charging current will decrease with gain in charge. It's considered done when the current stabilizes / -dI/dt becomes very low.

At higher charge current, you hit 4.20v at lower charge percentage than at a lower current and you spend more time finishing off the second phase.

The ceiling voltage of 4.10v is used when life is more important than capacity. and 4.15v for something in between. In any case, going above 4.20v is absolute impermissible.

 

[VanIsleDSM]

You're right jtr, I think the header/jumper method will be best, so you're using a preset charging current. A trimpot is actually lower cost.. but not as good in this case.

However, instead of using a jumper I think a female socket type header would be best. Then you could just move a solid core wire to different sockets for different charge currents. This would eliminate a jumper that would need tweezers to move and something you could possibly lose.

Using a 5 position female header I could use the the 1200, 1000, 750, 500, and 250 scheme that you suggested.


Sam,

Paypal works, although we'll wait until I've got them all built and working first. I just want to see if there's enough interest for me to build 25 instead of 6 for myself. If 12 bays can be spoken for I'll go through with it.

I just checked shipping to Finland, surface shipping should be about $10, and air shipping $15-20.. depends on the weight.. how many bays are you interested in? Let me know and I'll start a list.

 

[VanIsleDSM]

These posts came up while I was writing..

Just one little suggestion about the charging indicator light: It's a very common user interface pattern to have the indicator light blinking while "in transit" and steady when complete. So I'd suggest reversing your pattern so that the light is blinking during the charge and steady when charged. Another "mind's eye" picture for this is that blinking lights indicate activity. Compare for example the disk drive light on a computer or the lights on a modem.

It's not my configuration, it's preset into the charging IC. And I prefer it that way. Blinking all the way through a charge would be annoying as it catches your attention.. however I want my attention to be caught when done.. blinking does this well.. just another way to arrange your logic to perceive the LEDs.

I read this somewhere... I think it was the Battery Handbook.

You can't really fast charge a lithium ion if you want a full charge. You'll hit a brickwall that you can't get past.

Li-ion is charged in CC/CV fashion, so you start charging at xA, then once you hit 4.20v per cell, you hold it at a constant voltage. Charging current will decrease with gain in charge. It's considered done when the current stabilizes / -dI/dt becomes very low.

At higher charge current, you hit 4.20v at lower charge percentage than at a lower current and you spend more time finishing off the second phase.

The ceiling voltage of 4.10v is used when life is more important than capacity. and 4.15v for something in between. In any case, going above 4.20v is absolute impermissible.

If you check out the link for the IC that I posted you can get all the info you crave.

This chip uses a CC/CV charging method. During the CV charge at the end the current is monitored, once it drops bellow 8% of initial current the charge is terminated and battery voltage constantly monitored.. topped up as needed.

Edit: reread your post and I see now more what you mean. Cheap chargers just switch to a lower charging current in CV mode, usually 10% or 8% (and keep trickling that charge in without stopping) This takes a long time to finish off. The IC I'm using can vary it's charge current.. once 4.2V is reached and it switches to CV mode the charge current will still be high at first.. not a predetermined 10% a balance between keeping the cell bellow 4.2V but still charging it quickly is kept. The highest charge current possible will be used that will not push the battery over 4.2V. Check out the graphs in the datasheet.

 

[Handlobraesing]

These posts came up while I was writing..



It's not my configuration, it's preset into the charging IC. And I prefer it that way. Blinking all the way through a charge would be annoying as it catches your attention.. however I want my attention to be caught when done.. blinking does this well.. just another way to arrange your logic to perceive the LEDs.




If you check out the link for the IC that I posted you can get all the info you crave.

This chip uses a CC/CV charging method. During the CV charge at the end the current is monitored, once it drops bellow 8% of initial current the charge is terminated and battery voltage constantly monitored.. topped up as needed.

Edit: reread your post and I see now more what you mean. Cheap chargers just switch to a lower charging current in CV mode, usually 10% or 8% (and keep trickling that charge in without stopping) This takes a long time to finish off. The IC I'm using can vary it's charge current.. once 4.2V is reached and it switches to CV mode the charge current will still be high at first.. not a predetermined 10% a balance between keeping the cell bellow 4.2V but still charging it quickly is kept. The highest charge current possible will be used that will not push the battery over 4.2V. Check out the graphs in the datasheet.

Actually... no...
When the battery is depleted, you apply a voltage and the circuit limits the amount of current. Under this condition, voltage will continue to rise.

Once you hit 4.2v, the circuit will no longer allow the voltage to escalate and holds it at 4.2v. The battery just eats less and less current as it picks up charge and charger just keeps holding 4.2v constant. Now, you'll eventually reach a point where current will no longer drop and the amount of current the battery will continue to draw at 4.2v after a full charge depends on the cell capacity. Once you reach 4.2v, the only thing the charger can do is hold 4.2v and nothing else. Battery decides how much current it will draw. The final decision for the charger at this point is when to stop.

That's why I said perhaps it uses when relative -dI/dt flattens out.

Figure 6-2 on page 17:
solid line is current.

So, you start off with a warm up current.
2.
Then, charger jumps to the fast charge current.
I = constant provided that v <=4.20 or 4.10v
Charger holds the current, then the reaction in the battery dictates the rise in voltage.

Figure 6-3's curve represent the charger backing off fast charge current in response to overheat signal by the connected sensor.

3.
v = 4.20v I = will decrease as charge level is built.
When does it turn off. When whichever condition occurs first:

Iterm = between 8.5 and 90mA (you need to set it with a resistor)
Timer = ? (you need to set it with a capacitor)

 

[tpchan]

I would be interested in one or two bays as well. Would your planned charger be able to recharge a 10440 cell ? 250mA would be a good level for recharging the 10440 Li-Ion cells. I would like to be able to charge 18650, 17670, 14500, 16340 and 10440 cells all from ONE charger instead of the three units I currently have to use. Thanks!

 

[VanIsleDSM]

Handlobraesing, look more carefully. You said exactly what I did in a different way.

If you want to see how different types of Li-Ion chargers work.. just start reading through all the different datasheets for different ICs.. I've been through dozens now.

I see the point you're trying to make.. however it's not very valid when we're talking about a 0.5C rate for an 18650.. hardly near a brick wall. I wanted something faster than the cheapo 500-600mA units. And something that would not overcharge my batteries no matter what.. plus temp monitoring for the added safety.

tpchan,

A 10440 will fit fine. Anything longer than a 340 (34mm) will fit.

If you're in for sure let me know and I'll start a list in my first post.

 

[Mdinana]

Sounds interesting so far... Will this be able to charge 2 different battery types at the same time (I see some AAA/AA NiMH chargers... can this be done with Li-ions?).

If it can, I'd be interested; maybe a 4-bay charger?

 

[Handlobraesing]

Handlobraesing, look more carefully. You said exactly what I did in a different way.

If you want to see how different types of Li-Ion chargers work.. just start reading through all the different datasheets for different ICs.. I've been through dozens now.

I see the point you're trying to make.. however it's not very valid when we're talking about a 0.5C rate for an 18650.. hardly near a brick wall. I wanted something faster than the cheapo 500-600mA units. And something that would not overcharge my batteries no matter what.. plus temp monitoring for the added safety.

tpchan,

A 10440 will fit fine. Anything longer than a 340 (34mm) will fit.

If you're in for sure let me know and I'll start a list in my first post.

So, assuming you're charging a 2.4Ah cell what is the time to fully charge from ~0% to 100% at 1.2A vs 0.6A?

In primitive math, you'll about half it.. but at 1/2C, you'll just hit 4.2v sooner and spend more time in CV state whereas at 1/4C, you have more state of charge by the time you hit 4.2v and spend less in the final phase.

http://www.batteryuniversity.com/partone-12.htm

So my argument is that, if you want to race from 0-50%, I'm guessing the 1/2C will be nearly twice as fast, but I really don't know how much time you'd actually save if you want to do 0-100%, or charge a cell with 75% state of charge to 100%. It's the last 20% or so that takes a long time to charge regardless of what charger you use and the battery decides how fast you get to do it.

 

[VanIsleDSM]

Sounds interesting so far... Will this be able to charge 2 different battery types at the same time (I see some AAA/AA NiMH chargers... can this be done with Li-ions?).

If it can, I'd be interested; maybe a 4-bay charger?

Yes. Each battery bay is completely independent. You can have an 18650 in one bay and a 10440 in another. Each bay will have it's own status LEDs. You could use 1 bay at a time, or all 4.

So, assuming you're charging a 2.4Ah cell what is the time to fully charge from ~0% to 100% at 1.2A vs 0.6A?

In primitive math, you'll about half it.. but at 1/2C, you'll just hit 4.2v sooner and spend more time in CV state whereas at 1/4C, you have more state of charge by the time you hit 4.2v and spend less in the final phase.

http://www.batteryuniversity.com/partone-12.htm

So my argument is that, if you want to race from 0-50%, I'm guessing the 1/2C will be nearly twice as fast, but I really don't know how much time you'd actually save if you want to do 0-100%, or charge a cell with 75% state of charge to 100%. It's the last 20% or so that takes a long time to charge regardless of what charger you use and the battery decides how fast you get to do it.

Indeed. I don't see it as an argument though, I never stated the contrary. Only that we're no where near a brick wall. If you want to use more crude math, charging would be roughly 25% faster in your scenario.

Something else you have to take into account. If you have a 600mA charger, sure CV will kick in later.. but this doesn't matter. Why? because if you're using a 1200mA charge it will hit CV sooner.. however during the first bit of CV the cell will still be taking in close to 1200mA.. by the time the cell is charged to a point where it would only pull 600mA is the point at which the the 600mA charger would hit CV.. later yes, but that's no better.. the early CV portion of the charging with 1200mA will be faster than CC charging at 600mA. (until the point which the 1200mA charger declines to 600mA in CV mode)

Assume you could use a 10C charge rate, the charger may hit CV very quickly, but the CV charge will still be pumping out huge current as it declines.

Again, I see what you're saying, but it doesn't take much effect yet.. if you put in more and more power you do get diminishing returns.. but it's still well worth it right up to the 1C max recommended rate.

 

[Handlobraesing]

That quick? right on.. I knew the capacities were ~2400mAh, but I thought the charging was less efficient than that. I suppose not.

Thanks for the reply.

It has nothing to do with efficiency. When you're fueling your car and you are nearing full, you slow down on rate of filling. The charging efficiency remains unchanged, because even though slowed down, every ounce coming out of the nozzle is still going into the car.

That's exactly the way lithium ion is.

If you have a laptop, pop open the battery indicator while charging. You ever notice that it picks up the charge pretty quickly from near empty, but the last 20% or so takes a very long time to top off?


So, I guess it'd be nice to have state of charge gained vs time at 1/2C and 1/4C

 

[VanIsleDSM]

Ok, the first discussion we had was good.

Now you're nitpicking my thread.

You're car analogy is wrong. Why do you think a battery heats as it's charged? It's not 100% efficient like the transfer of gas from a pump to your car (neglecting evaporation) or there would be no heat created.




Back on topic... I'm surprised there are no graphs out there comparing different charge rates vs time. Though I'll easily be able to tell the difference once I'm done building these. Should be roughly 30-35% faster I suspect.

 

[Handlobraesing]

Ok, the first discussion we had was good.

Now you're nitpicking my thread.

You're car analogy is wrong. Why do you think a battery heats as it's charged? It's not 100% efficient like the transfer of gas from a pump to your car (neglecting evaporation) or there would be no heat created.

Actually, the charge efficiency of lithium ion is pretty darn close to gas fillups
http://www.batteryuniversity.com/partone-12.htm (99.9%)

Lithium ions hardly get hot.

 

[tpchan]

Sure I'm in for two bays @ $10 each. If the price goes up much more than that, then one bay will do fine for me. Thanks.

 

[zapper]

OK.

I think there is a lot of good information here that is getting chewed up as useless heat in theorizing.

Build one first then test it next to the standard and let the chips fall where they may. Minimize your time and investment. Theory to practice is a normal testing method.

Once the results are known and if they are a positive improvement then you can add $2-3 each and I'm sure plenty of people will line up.

I'm in for at least two as part of the R&D buy in for this project to get things going ASAP.

I'll even go in halfsies with you to get the cheaper components till the rest of the people I really think will buy into the project.

Let's see, ~5V *1200mA= 6W at max draw, therefore you should run out of Molex connectors before hitting the 350W PS capacity ? :thinking:

PM me if you want to hash out the details.

:twothumbs

P.S. I would chime in for 4.2V terminal charge as there is lots of power between 4.1V and 4.2V and still gives plenty of lifetime charge cycles. Info to this affect can be found with some searching and lots of reading as well!