Simplest BMS

[317537]

I tried those cheap 12v to 60v boosters off my solar system , the non-isolated type from ebay, without too much sucess, even with resistors I found it dropped from the converter output set voltage, almost to pack voltage over the resistor to get it to work, meaning 60v to 53v @ 2amps, I think about 6v i was reading , a waste of 12watts, which isnt terrible but not ideal considering I was already running over 25% losses in the converter. Comparable to a 12v to 220v AC inverter and ping battery charger losses though.

Im going to try again with a brushed controller on the output and run it under 1.8amps which is fine for over night charging off my 350ah 12v flooded lead acid..

Im not sure .1ohm is enough resistance, and like my test system you will note that when you achive the output current you need the drop occurs, you might see, more like a .5v drop @ 15A and this is 7watts per series cell, unacceptable IMO.

 

[317537]

The issue presneted with my DC booster experiment, is that limiting the input current saw too much V drop from the DC converter output to the Vbatt input, and the booster would shutdown to low voltage protect mode. However limiting the DC output with a 48v ebike brushed controller should be an efficient easy fix.

With the buck conveters the issue is the low voltage outputs not being very high for an easy solution.

So this is a BMS right, I suggest series the output then uses balance resistors up to 2 amp between first + terminal primary ground cell and VCC - terminal header cells and directly connect the first and last cells to the DC converters series string. Maybe. << Doesnt work.

 

[317537]

Ahhhh I see the light.

The beauty of LI battery charging, is that cells are already (at optimum DOD) only 1v at under its hot 100% SOC. Its the v difference between the charger output and battery terminals that allows current to flow to charge and internal resistance expense to keep it afloat

So you will only see a 1v drop over any resistor per cell at any currnet rating while charging, and less if you dont take the risk of over discharge to the bottom limits of the cells spec. Not much in the last .5v anyway. Limit it to 2 amps say and you only lose 2 watts per resistor at most 1watt at best. Try a 1.8 ohm resistor on each output.

 

[Jeremy Harris]

Ahhhh I see the light.

The beauty of LI battery charging, is that cells are already (at optimum DOD) only 1v at under its hot 100% SOC. Its the v difference between the charger output and battery terminals that allows current to flow to charge and internal resistance expense to keep it afloat

So you will only see a 1v drop over any resistor per cell at any currnet rating while charging, and less if you dont take the risk of over discharge to the bottom limits of the cells spec. Not much in the last .5v anyway. Limit it to 2 amps say and you only lose 2 watts per resistor at most 1watt at best. Try a 1.8 ohm resistor on each output.

In practice, even a completely discharged cell, at, say, 3.5V, will very quickly increase to about 3.8 V to 3.9 V, within a minute at most normally. If you have a 0.1 ohm resistor in series with a 4.2 V output from DC DC converter that means the voltage drop across the resistor will be around 0.3 to 0.4 V for the majority of the charge period. The charge current is then limited to around 3 to 4 A with the 0.1 ohm resistor. I use 0.05 ohm resistors on my chargers (two 0.1 ohm 5 W resistors in parallel) and this gives me around 8 A max charge, dropping to around 4 A as the cells get towards 4 V. The charge current continues to taper off as the cells reach full charge, just as the cell manufacturers generally recommend.

On the LiFePO4 charger I don't use any current limiting resistors, as I've found that the resistance of three feet of 12g wire is enough to keep the DC DC converters from going into hiccup current limiting mode. That charger does run at around 20 A per cell group though.

 

[317537]

Yes Jeremy, I think we are on a winner here, I took a closer look at that data sheet,

To provide protection in a fault (output overload) condition,
the unit is equipped with internal current-limiting
circuitry and can endure current limiting for an unlimited
duration. At the point of current-limit inception, the
unit shifts from voltage control to current control. If the
output voltage is pulled very low during a severe fault,
the current-limit circuit can exhibit either foldback or tailout
characteristics (output current decrease or
increase). The unit operates normally once the output
current is brought back into its specified range.

Looks Good to me as the voltage aint going to get pulled down too much anyway.. These converters are usually very reliable in so many situations.

If one can get something that can handle over 25 amp to the paralleled DC converter inputs, EG a welder trasnformer, there is no reason these DC converter wont work for high current battery chargers.

 

[amberwolf]

Amber, perhaps you can move these into a new thread, there seems to be interest and I'm sure others want to share the data about near-perfect chargers.

What say all the rest of you? I would like to move all of this DC-DC discussion over to the thread Jeremy linked:
http://endless-sphere.com/forums/viewtopic.php?f=14&t=2824
as it seems the most appropriate place for it.

 

[Alan B]

Amber, perhaps you can move these into a new thread, there seems to be interest and I'm sure others want to share the data about near-perfect chargers.

What say all the rest of you? I would like to move all of this DC-DC discussion over to the thread Jeremy linked:
http://endless-sphere.com/forums/viewtopic.php?f=14&t=2824
as it seems the most appropriate place for it.

Near-perfect.

 

[deVries]

DC-DC converters are not generally designed to charge batteries. They may not have the proper current limiting to avoid damage to themselves and provide significant current when the batteries are low. They may not be designed to be connected to a source of power without having power input. Depending on their design, they may work, or not, or they may blow up, and they may leak current when not powered. It depends on how the particular converter is designed. Also not all converters are isolated. Isolation is required for this use.

You need to find the right ones, and what you are seeking is not in the specs.

Good luck in your search,

And this belongs in a separate thread.

Comments from a known battery expert:

Individual cell charging is the most safe method, and the method that enables any state of imbalance to be corrected quickly and easily and safely. You can have fully charged cells in a string with 100% empty cells and they all end up safely at 100%.

Isolated DC/DC converters powered by a meanwell or server supply is generally the best method.

Seems to me this IS the Simplest BMS, which is what the OP is asking for too in the Subject OP: Simplest BMS.

 

[Alan B]

If you don't mind having zero information about the state of your batteries, maybe. I'd like to see more in the way of data as well as checks and balances, so to speak. If one of those converters fails, or becomes disconnected, you will not know, and the pack will quickly be damaged when you try to use it and one cell group has little charge while the rest are full. Pack voltage will be pretty far from empty when those cells are ruined. DC converters may be a good component of a system, but more is needed. They don't protect against overdischarge either, so they are not a Battery Management System at all.

Individual cell charging doesn't necessarily mean DC-DC converters either.

This thread was specifically about a weak balancing design, which is also not a BMS in that sense.

 

[deVries]

If you don't mind having zero information about the state of your batteries, maybe. I'd like to see more in the way of data as well as checks and balances, so to speak. If one of those converters fails, or becomes disconnected, you will not know, and the pack will quickly be damaged when you try to use it and one cell group has little charge while the rest are full. Pack voltage will be pretty far from empty when those cells are ruined. DC converters may be a good component of a system, but more is needed.

Celllog8 or want more data? ... get it in the logger version & play with it on the computer too!

This thread was specifically about a weak balancing design, which is also not a BMS in that sense.

Individual cell charging doesn't necessarily mean DC-DC converters either.

Sure, use Voltphreaks isolated output chargers, but LFP specifically recommended the DC-DC converters in the quoted previous post. True enough, one has to search around to find the right match, and it's definitely not turnkey retail or kit form. :wink:

What BMS will not kill a pack if it is not charged periodically? I think Fechter/Goodrum were solving that, so it would take more than a year to kill some packs???

 

[Jeremy Harris]

If you don't mind having zero information about the state of your batteries, maybe. I'd like to see more in the way of data as well as checks and balances, so to speak. If one of those converters fails, or becomes disconnected, you will not know, and the pack will quickly be damaged when you try to use it and one cell group has little charge while the rest are full. Pack voltage will be pretty far from empty when those cells are ruined. DC converters may be a good component of a system, but more is needed. They don't protect against overdischarge either, so they are not a Battery Management System at all.

Individual cell charging doesn't necessarily mean DC-DC converters either.

It's true you don't know when full charge for every cell has been reached with a DC DC charge system, but neither do you know this with a low capacity shunt system like the circuit suggested here, either. It's also true that there are other individual cell charge solutions, like the Voltphreak units.

When there is a DC DC converter failure it will result in a cell group not getting charged, something easily checked with a post-charge voltage check with a Cellog or similar. A component failure in the weak shunt described here may result in a battery fire, from a cell group going over voltage on charge.

This thread was specifically about a weak balancing design, which is also not a BMS in that sense.

It's not a balancing system at all, it's simply a discharger that gets all cells to the same terminal voltage, and actually ensures that cells in a pack are unbalanced, as manufacturing tolerances guarantee that not all cells in a pack will have the same capacity. You're right, it's not a BMS, in that it neither ensures full charge for all cell groups nor does it provide low cell voltage cut off.

At least the individual cell group charge method does pretty much guarantee that charge is managed safely and cells are balanced, even if it doesn't provide an LVC capability.

 

[Skippic]

Hey guys,

more than a year of bulk charging every day I measured all cells and they are perfectly balanced both charged and discharged (+/-0.005V). The battery capacity seems to be somewhat smaller, but that's unrelated to the balancing issue. I'm very happy with the BMS.

 

[Alan B]

Glad it is working for you.

I've put over 3,000 miles on my Borg in a year with no BMS at all (aside from me checking balance periodically with a meter) and balance is still good on the pack.