-- THE SINGLE CELL CHARGER POST --

[Doctorbass]

But that one says only 5% voltage adjustment.

3.3V+ 5% would give 3.47V.. si the 130mV would be easy to tweak with the resistor divider bear the pot or comparator IC

maybe 3.7V...

 

[Beagle123]

The only consideration I can think of is that you probibly don't want to charge too fast. A123s can be charged very fast, but it will shorten their lifespan. I think 0.5C is a good charge rate, and it should be pretty fast. Doc, if you're usig a 2p configuration, charging at 4 amps = 0.8C which would be good. Assuming that the charger doesn't perform quite as well as advertised, the 25W model might be a good choice.

Those are great finds. I can't wait to see how this works out.

 

[Ypedal]

Nice finds... personally i want a 30 minute charge, 1 hr max.

With LiFe cells, honestly, i don't expect to get 2000 cycles in real life use, i rather use and abuse a bit and move on to new things after 2 to 3 years...

The way i ride the packs will suffer crash damage before cycle life comes into play ! lol..

 

[GGoodrum]

The only consideration I can think of is that you probibly don't want to charge too fast. A123s can be charged very fast, but it will shorten their lifespan. I think 0.5C is a good charge rate, and it should be pretty fast. Doc, if you're usig a 2p configuration, charging at 4 amps = 0.8C which would be good. Assuming that the charger doesn't perform quite as well as advertised, the 25W model might be a good choice.

Those are great finds. I can't wait to see how this works out.

In the RC world, charging a123 packs at 10A, which is a 4.3C rate, is the norm, not the exception. Even a123Systems recommends charging these at 10A per cell. Some guys are charging packs at 30A, with no apparent issues (they don't even get too warm...) I've got packs with hundreds of charges on them, still going strong, and many of these have been charged at even higher rates. You are not going to increase their life by charging them at such low rates as you suggest.

With the bulk charging add-on I'm working on right now, to the 2A VoltPhreaks chargers, I'm trying to get the CC mode up to 20A, which for my 10s4p packs means it is still only charging at about a 2C rate. Ideally, I'd like it much higher, but finding inexpensive 40V supplies that will handle 40-50A is not easy. Bob found some relatively inexpensive 17A surplus supplies used by the phone companies at remote sites, so we're going to try that first.

-- Gary

 

[EMF]

With the bulk charging add-on I'm working on right now, to the 2A VoltPhreaks chargers, I'm trying to get the CC mode up to 20A,

-- Gary

Hurry! :wink:

 

[Beagle123]

Wow. I wasn't aware that they reccomaend charging at 3.3C. That's incredible. That means that you could charge any bike to almost full in about 15 minutes. That's really revolutionary.

 

[Jozzer]

Beagle Twigs :wink:

 

[dermot]

It's a pity that there seem to be no single-chip chargers designed specifically for LiFePO4 cells, there are plenty for Cobalt and Manganese chemistries.

 

[Doctorbass]

I really feel that the final conclusion of that post will turn around many independent power supply.. most switching type) that are adjusted to 3.6V and that have a current limiter to around 5 to 10A... 8)

Are you curious to know what capacity on dishcharge you can get with A123/LiFe at many different charging current?

I will see what i can do :wink:


With A123 M1 cell


0.5C
1C
2C
4C


soon...

 

[steppedup]

Can anyone post pictures of how to charge up a single cell straight out of the Dewalt pack using a battery charger like the one below?

Will I have to completely disassemble it (including BMS etc) to do so?

Unfortunately, I discharged one of my Dewalts too low and am getting the triple blink on the charge - and it measures around 12 volts. Tried putting it in the charge 30 to 80 times like some have recommended with no success.

Thanks either way!!!!

 

[OLIVIER]

Hi,
I am new to this forum and this is my first post.
I just received a Goldenmotors bike conversion kit (36V, 500W) and I am searching for a battery pack.
I want LIFEPO4 cells for a total of 36V.
It seems to me from everything I read on E-S that cells in parallel are best for charging and that cells in series are best for discharging. In my case I would need 12 cells.
One high amp single cell charger could charge AND ballance a parallel battery pack, without any overcharging.
The controller would prevent over-discharging of pack in series. If single cells drop too low or invert, the should be changed anyway.
The trick is to wire the pack with a toggle to go from series to parallel. Each time the EV is turned off or the throttle backed off, the pack would switch to parallel to equalize the cells or for charging (or regen). It would go back to series for driving.
I know this would require lots of large fast cheap relays but with each cell equiped with relays, in case of a crash, they could all be isolated instantly.
What do you think?

Thanks,

 

[Doctorbass]

Welcome Olivier
(Salut mec... Bienvenue!)

 

[Doctorbass]

I really feel that the final conclusion of that post will turn around many independent power supply.. most switching type) that are adjusted to 3.6V and that have a current limiter to around 5 to 10A... 8)

Are you curious to know what capacity on dishcharge you can get with A123/LiFe at many different charging current?

I will see what i can do :wink:


With A123 M1 cell


0.5C
1C
2C
4C


soon...

Like i promissed here is the link to my lastest test about capacity/energy vs charge current :

http://endless-sphere.com/forums/viewtopic.php?p=38018#38018

Doc

 

[Beagle123]

the pack would switch to parallel to equalize the cells or for charging (or regen). It would go back to series for driving.
I know this would require lots of large fast cheap relays but with each cell equiped with relays

Hi Oliver:

I've been working on that exact theory. I'm going to switch from parallel to series using plugs made from andersens connectors. Each of my plugs will have 7 battery packs on them. When the two plugs fit together, the packs are in series. When the plugs are unplugged everything is disconnected, so I can make charging plugs to charge through the same plugs. I think it will be a good advancement.

There is some discussion on my thread below, and I will post pictures of my batteries/plugs/charger in the next couple of weeks.

I'd like to see if you can invent another solution. I think if we can find a way to solve this, it will be a big step forward.

 

[catalystTGJ]

I want to throw out a real simple line of questions, and it may be cheesy, but i just have to put this out here... forgive me if this is too junior!

If one were to hook up a series of single cell chargers, say in the neighborhood of 20 to 24 units, and if those chargers were all in the 8 amp range, what sort of current tolerance is going to need to be on the AC line that this current is drawn from? Also, what sort of AC power cord is going to be needed? Isn't this going to throw breakers?

Am I right in figuring 20 chargers X 8 amp @ 160 amps?? This can't be right.

But if this is correct, this is going to be a bit too much for the typical office cubicle outlets, for those thinking of carrying their mega-power bricks up to work and plugging them in. Most 110 home outlets are going to be double digit only amperage, or is the low voltage the offset? Is it really only overall wattage needed and so not an issue? If that is true, what about the inrush power on each of the chargers? Could that be enough to pop a breaker with so many of these coming on line all at once?

Hope this wasn't a waste of packets for folks.

 

[Ypedal]

It's all about the watts.

Your AC wall outlet usually = 15 amp breaker at 120v = 2000w

Your chargers at 8 amps x 3.2v = 25w

So as long as you don't exceet 1000w + at the wall outlet ( AV voltage.. not charger voltage ) you are ok

another example:

My 72v 20ah LiMn pack uses 2 chargers

2 amp 36v = 72w
4 amp 36v = 144w
Total : 216w.. i could charge 10 bikes on one outlet ! * would be reallly pushing it tho )

 

[Doctorbass]

the pack would switch to parallel to equalize the cells or for charging (or regen). It would go back to series for driving.
I know this would require lots of large fast cheap relays but with each cell equiped with relays

Hi Oliver:

I've been working on that exact theory. I'm going to switch from parallel to series using plugs made from andersens connectors. Each of my plugs will have 7 battery packs on them. When the two plugs fit together, the packs are in series. When the plugs are unplugged everything is disconnected, so I can make charging plugs to charge through the same plugs. I think it will be a good advancement.

There is some discussion on my thread below, and I will post pictures of my batteries/plugs/charger in the next couple of weeks.

I'd like to see if you can invent another solution. I think if we can find a way to solve this, it will be a big step forward.

I just wonder what would be the best tradeoff between the advantages of your idea Fechter and the problem of multiple charger.

Beagle123 method:

pos
+ simple to built
+ great balancing results
+ 100A+ <5V power supply easy to find in some surplus for less than 100$

neg
- high resistance loss using multiple connectors to join in serie all the 1p group
- multiple big wires needed to each connectors ( sometime 40+ 12gau wire to the charger)


Now the multiple single charger method:

pos
+ smaller wires goes to the battery
+ more flexible harnes for handling...
+great balancing results
+ smaller multipins connector (cheaper)
+no resistor loss during high current demand

Neg
-higher cost when buying 10+ single power supply (125$+)
-all these power supply can take alot of place.
-the need of adjusting the voltage precisly to match to each others.


Beagle123, I think your idea is great and simple, the only thing that would worry about is the high resistance loss in the multiple connectors for serie wiring at 40A+.. have you calculated the loss?

-

 

[Beagle123]

Hi Doc:

You left-out two major advantages to my method:

1) That its easier to diagnose problems and replace bad cells.

2) You don't have to connect a bunch of charging leads that can potentially short your pack.


I would also say that it gets better than "great balancing results." In fact, it will give perfect balancing results because all the batteriess will actually be connected together when charging. This is a slight advantage over individual chargers whose voltages vary slightly. However, if an individual charger burns-out, or malfunctions, this would be a major issue.



As far as the added resistance from the connectors goes, I'm going to test that very soon. I'm hoping that my 45 amp andersens connectors have almost no resistance. I'm going to have the current go through 28 of them, so it is critical that I don't lose any voltage on each connection. I think it will work fine. I have calculated my power losses due to the additional length of wire. I'm planning on using about 20 feet of stranded 10 guage copper wire. This length of wire should have 0.0204 ohms according to this online calculator:

http://www.stealth316.com/2-wire-resistance.htm

The corresponding voltage drop will be 0.408 volts at 20 amps. That should be fine too. So hopefully, the voltage drop from the wires and connectors will be in the neighborhood of 1 volt.

I'm not sure how to calculate what kind of power loss that would represent. Here is an example:

20 amps @ 50 volts = 1000 watts = 2.5 ohms

So, if the resistance of my wires and connectors is 0.05 ohms it owuld be about 2% of the toal resistance of 2.5 ohms. (0.05/2.5) So, it would use about 2% of the power or 20 watts. Do you think that's correct?

 

[Doctorbass]

Hi Doc:

You left-out two major advantages to my method:

1) That its easier to diagnose problems and replace bad cells.

2) You don't have to connect a bunch of charging leads that can potentially short your pack.


I would also say that it gets better than "great balancing results." In fact, it will give perfect balancing results because all the batteriess will actually be connected together when charging. This is a slight advantage over individual chargers whose voltages vary slightly. However, if an individual charger burns-out, or malfunctions, this would be a major issue.



As far as the added resistance from the connectors goes, I'm going to test that very soon. I'm hoping that my 45 amp andersens connectors have almost no resistance. I'm going to have the current go through 28 of them, so it is critical that I don't lose any voltage on each connection. I think it will work fine. I have calculated my power losses due to the additional length of wire. I'm planning on using about 20 feet of stranded 10 guage copper wire. This length of wire should have 0.0204 ohms according to this online calculator:

http://www.stealth316.com/2-wire-resistance.htm

The corresponding voltage drop will be 0.408 volts at 20 amps. That should be fine too. So hopefully, the voltage drop from the wires and connectors will be in the neighborhood of 1 volt.

I'm not sure how to calculate what kind of power loss that would represent. Here is an example:

20 amps @ 50 volts = 1000 watts = 2.5 ohms

So, if the resistance of my wires and connectors is 0.05 ohms it owuld be about 2% of the toal resistance of 2.5 ohms. (0.05/2.5) So, it would use about 2% of the power or 20 watts. Do you think that's correct?

The Anderson 45A powerpole connector have 0.0006ohm each so 28 of them would have 0.0168ohm. + the 0.0204ohm of your wires that give you a loss of 0.0372ohm. In this case, That will dissipate P=R*I^2 = 0.0372*400 = 14.88W total without taking acount of the soldering between wires and connector. It's a bit less than 1V loss. That give 0.744volts.

 

[GGoodrum]

While it is true that connecting all the cells in parallel will balance all the cells, but it will be to the weakest one. With individual cell chargers, each cell/block can be charged to whatever maximum voltage each cell can handle. With healthy cells, it probably doesn't matter all that much, as the differences will be smaller.

-- Gary

 

[Beagle123]

Thanks Doc. That's good info. I was wondering the formula to compute power loss. thanks.

Gary, I disagree that my charging method will charge to the weakest battery voltage. I'm going to use a power supply that will force all the batteries to charge to the exact same level. The weakest cells won't be able to drag the others down because the power supply will force the voltage at all the cells.

 

[Ypedal]

What i think may happen is the weakest cell in the string will drag the voltage of the others down to it's level once you disconect the charger.. :?

 

[Doctorbass]

What i think may happen is the weakest cell in the string will drag the voltage of the others down to it's level once you disconect the charger.. :?

That would mean that the weakest cell in BOTH case would limit the pack to his caracteristic.

-if it could drag the voltage of the others down to it's level in paralleling charging after disconecting the charger, or

-if charged seperatly with individual charger, the weakest cell will empty faster than other and would limit the capacity of all other cell in serie to his capacity anyway...


What to remmember:

---MATCH YOUR CELLS--- and make sure when discharging, that they have all the same temperature. By this way, they will have a closer internal resistor from each others and will have a balanced charge and discharge current in when in parallel.

We all know that parallel cell have all the same voltage... BUT they can have different current depending of each of their internal resistor!

Suppose that all cells of a parallel group have the exact same capacity.
The lower cell Ri will be drained faster and will empty faster than the higher Ri cell.. ! and finally, the higher Ri will be the lasts cells to supply the current of this parallel group!

This is not dangerous for your cells because they will have the same voltage and will cut at the same time.. buy their current sharing will be different.

Doc

 

[Malcolm]

Interesting discussion here:
http://www.convertthefuture.com/bbs/viewtopic.php?t=15335&sid=41955a36b1cf4dde1bdccdd8e1a1060b

Lee Hart: You don't need a $1000 BMS for a scooter. But you need *something* to keep them healthy, or you'll soon have many dissatisfied customers and maybe a fire or two!

A minimal BMS would be something like my zener-lamp regulators, which are under $5 per cell. But you'd have to use something other than a zener at 3.6v/cell, like a TL431 zener equivalent.

For a low-voltage shutdown, use something like my Batt-Bridge circuit. It doesn't tell you which cell is low, but it will tell you that *some* cell is low. Use it to warn the operator and shut down the scooter.
Again, this is only $10 or so worth of parts.

One other option that I find interesting as an engineer. We recently
talked about boost converters instead of buck converters for the motor controller. For a scooter, you can use a single large lithium cell and a boost converter to step this up to 36v or so. There is no need for a BMS with a single cell.

It's a nice idea, taking the single cell solution to the extreme. The downside is that you increase I2R losses before the controller, so it would only be practical for relatively low-voltage, low-power applications.

 

[Doctorbass]

Interesting discussion here:
http://www.convertthefuture.com/bbs/viewtopic.php?t=15335&sid=41955a36b1cf4dde1bdccdd8e1a1060b

Lee Hart: You don't need a $1000 BMS for a scooter. But you need *something* to keep them healthy, or you'll soon have many dissatisfied customers and maybe a fire or two!

A minimal BMS would be something like my zener-lamp regulators, which are under $5 per cell. But you'd have to use something other than a zener at 3.6v/cell, like a TL431 zener equivalent.

For a low-voltage shutdown, use something like my Batt-Bridge circuit. It doesn't tell you which cell is low, but it will tell you that *some* cell is low. Use it to warn the operator and shut down the scooter.
Again, this is only $10 or so worth of parts.

One other option that I find interesting as an engineer. We recently
talked about boost converters instead of buck converters for the motor controller. For a scooter, you can use a single large lithium cell and a boost converter to step this up to 36v or so. There is no need for a BMS with a single cell.

It's a nice idea, taking the single cell solution to the extreme. The downside is that you increase I2R losses before the controller, so it would only be practical for relatively low-voltage, low-power applications.

I toke that from the same forum and post:
We have a very simplistic BMS on the KillaCycle at this time. Steve
Ciciora calls it the "Opto-Bio-Mechanical Feedback" BMS. There are
two LEDs associated with each parallel group of cells. You turn on
the charger and watch the lights like a hawk. When a green LED come
on, this means that parallel bank of cells is "full enough". If the
red LED comes on too, it means that that bank is "completely" full
and you need to turn down the knob on the charger. The operator turns
down the knob when he sees any red LEDs, while waiting for all the
green LEDs to come on. When all the green LEDs come on, the operator
turns off the charger.

The system is simple, lightweight, reliable, but labor-intensive. It
works great for drag racing.

While this is just fine for a drag racing application where you have
a person that will watch the pack and turn the knob, it is not
suitable for a daily-driver EV application.

The very first pack of A123 Systems M1 cells that we ever
built is still working perfectly. We ran it for 18 months of racing
in the KillaCycle, lent it to Shawn Lawless to set a record in a
dragster, and lent it to John Wayland to set a NEDRA record in his
White Zombie street-legal EV. We have never replaced a single cell in
it. It is still at 100% its original HP output.

We have no idea how many cycles we will get from this
original racing pack. We estimate that is would last for about five
years of racing every weekend, but that is just an estimate. We won't
really know until we actually wear it out.

Bill Dube'