DIY Electric Car Forums banner

1 - 20 of 94 Posts

·
Registered
Joined
·
102 Posts
Discussion Starter #1
I've been pondering BMS designs and issues and have a thought for a bit different approach. Cell balance is a universal function of BMS systems. Top balance is easily automated by shunts but not really desirable. What if the BMS included automated BOTTOM balance? Automated bottom balance could be achieved by having per cell charging circuitry that responded to the "average" cell voltage, charging any cells below the average while idling for the rest. Another issue with most, if not all, existing BMS designs is parasitic loading and the in ability to easily " disconnect" the BMS when the pack is "idle".
My solution: I'll call it charge only active balancing. Per cell chargers based on economical buck converter IC's and pulse transformers (TI's LM46000 and Pulse Engineering's PA 3855.004). This combination is capable of charging one cell with up to 6 amps! The cells are transformer isolated from the rest of the circuitry and when powered down the parasitic loading, per cell, will be measured in pico-amps! All the chargers are powered by a tap on the pack that can be between 4 and 16 cells (12 to 60 volts). A 4 wire " bus" connects the modules, the wires are 1: power ground, 2: power, 3: cell average (at 3 times the per cell voltage), 4: charger shut down ( an open collector or of all the modules set to activate when any cell hits 3.6 volts). The costs? about 20 milli-watts per cell when not charging and BOM <$10 per cell. I'm planning on building modules that serve 4 cell, or less, groups to minimize interconnections (module dimensions, about 3"x4"x1/2"). Anticipated performance is all cells remain within +/20 milli-volts of each other. I'm struggling with a couple of decisions; Should I include LEDs to indicate status ( they cost energy but add comfort), If yes, what should be indicated? Simple possibilities include power on, charging, fully charged.
Anyone interested?
 

·
Registered
Joined
·
102 Posts
Discussion Starter #3
So you have to empty your battery to balance...
NO! NO! NO! it is being continuously kept balanced throughout the entire process of charging or discharging with per cell balance curents of up to 6 amps.

a good battery doesn't need much balancing, so I don't see any positive things about this idea. Please tell me where I'm wrong..
Agreed, but this assures that it'll stay good and even as cell capacity mismatch happens the absolute maximum capacity will remain available. Every cell will remain within +/-20 milliolts of the others to the extent that 6 amps can keep it that way.
 

·
Registered
Joined
·
1,478 Posts
i like the idea of being able to disconnect the parasitic drain of the BMS from the cells--i have studied this in ryobi 40V power tool packs and found that to be the biggest source of pack "failure". Most "dead" packs had fully functional and good cells, but the BMS drained them. 5 Ahr 18650 packs of a smaller size but same principles apply.
 

·
Registered
Joined
·
102 Posts
Discussion Starter #6
i like the idea of being able to disconnect the parasitic drain of the BMS from the cells--i have studied this in ryobi 40V power tool packs and found that to be the biggest source of pack "failure". Most "dead" packs had fully functional and good cells, but the BMS drained them. 5 Ahr 18650 packs of a smaller size but same principles apply.
To get to the very low drain this "BMS" would need to be "turned off" by removing the power. However, even with out powering down the parasitic load on the "cell tap" can be below 5 micro amps per cell module when cell voltage falls below 2.1 volts, assuming a 12 volt (4 cell) power situation.
 

·
Registered
Joined
·
565 Posts
I've seen OEM modules (VW, Tesla, Mitsubishi, Kia) that were on the shelf for sometimes 2+ years, without inbalance...so the parasitic drain in 'off' state if obviously -very- low, and at least 'well balanced' on a good bms...

You'r free to build anything you want, but the only usecase would be a battery pack made from different capacity cells / modules....but...

when you bottom balance like this...what happens when you charge the pack fully? yes...smalles capacity cell arrives at 'charged' voltage first...and your 'active' bms has to drain this cell into the other cells of the module.

Also, the different modules (groups) become unbalanced from each other and how are you going to tackle that?

Luckily, more and more OEM bms systems become usable (Tesla, Mitsubishi, Volt?, etc) so everyone can use high quality stuff for very little money.
 

·
Registered
Joined
·
102 Posts
Discussion Starter #8 (Edited)
I've seen OEM modules (VW, Tesla, Mitsubishi, Kia) that were on the shelf for sometimes 2+ years, without inbalance...so the parasitic drain in 'off' state if obviously -very- low, and at least 'well balanced' on a good bms...

You'r free to build anything you want, but the only usecase would be a battery pack made from different capacity cells / modules....but...

Just how good is the capacity match of commercially available cells?

when you bottom balance like this...what happens when you charge the pack fully? yes...smalles capacity cell arrives at 'charged' voltage first...and your 'active' bms has to drain this cell into the other cells of the module.

NO! NO! I clearly said this "BMS" can CHARGE ONLY so while pack charging is occurring extra charge current,up to 6 amps, is being given to the "stronger" cells. Other than tacking power from the cell tap there is simply no way to discharge anything with this structure.

Also, the different modules (groups) become unbalanced from each other and how are you going to tackle that?

Again NO! NO! The modules are not independent AT ALL! They share a 4 wire "buss" with power ground and an average cell voltage wire and a signalling wire to stop charging when any cell hits 3.6 volts.

Luckily, more and more OEM bms systems become usable (Tesla, Mitsubishi, Volt?, etc) so everyone can use high quality stuff for very little money.
Frankly I do believe this is better than any OEM system available. The retained effective and availale capacity after lots of use should be substantially higher. And almost all the OEMs insist on using more fragile and dangerous chemistry. LIFEPO4 doesn't have the highest energy density but is far and away the most forgiving and safest.
 

·
Registered
Joined
·
102 Posts
Discussion Starter #9
I'm surprised by the lack of response. I'll offer a bit of explanation of my reasoning in suggesting "charge only" active balancing. From a pragmatic view it is much easier to build, but does it work without ever needing to remove charge? Here is why it does: In use the pack is either being charged or discharged, during pack charging this BMS will add additional charge to the "stronger" cells making them fill to their actual capacity wile during pack discharge it will add charge to the "weaker" cells boosting their effective capacity allowing the stronger cells to deliver their actual capacity. Basically forcing all cells to have the same voltage through out the entire cycle of pack charge and discharge strong cells get added charge during the pack charge and weaker cells get added charge during pack discharge. Once I realized this the design became much simpler. It's easy to add an active rectifier to a "flyback" type of DC to DC converter if it's "half wave" and only transfers energy in one direction, much more challenging to get a truely bidirectional energy transfer. However, it still was a challenge to adapt available IC's to do it. they are all made to deliver an output voltage that is fixed by design and do not support the very necessary function of "tracking" an input voltage (the cell average). But after much head scratching I've "invented" a way to do it. I'm now thinking that the extra wire to stop the charger should go away as any pack charger will have a pack voltage setting. Or is redundancy really important? My 4 cell module is 4"x 2 1/2" x 1/2"


.
 

·
Registered
Joined
·
102 Posts
Discussion Starter #11
Well, do the math... see attached image and calculate how many amps you have to pull on your (12v?) bms supply if you have 100 cells and you need to charge 99 of them with up to 6A
It's true that the 6 amps comes from the cell tap and limits the available charging of all cells combined. This is why a higher voltage cell tap is desirable. With a tap at 16 cells, still within the components ratings, there would be 6/16 amps or 0.375 amps per 6 amp load allowing 15 cells to simultaneously receive 6 amps.
The effect is the cell tap power limits the number of number of places the full 6 amps can be applied (bad cells) to 3 for a 4 cell volt tap 7 for an 8 cell tap and 15 for a 16 cell tap any of which is far better than traditional BMS or either top or bottom balancing!!
 

·
Registered
Joined
·
565 Posts
ok I'll try something else:

when a weaker cell drops voltage...you can charge it with 6A, but say your car wants 50A...no-go...

and now you say you -do- use cell groups to get the power from... so group1 can get out of balance with group x

Plus, when charging 15 cells from a 16 cell group...well..when not counting the losses...you're basically discharging the 1 cell. Losses make it so that lots of heat is put out and this cell-group will drop below other groups.

An active BMS -might- make sense in a 4 cell battery...but there also the cells should just be the same, not needing active balancing.
 

·
Registered
Joined
·
102 Posts
Discussion Starter #13
ok I'll try something else:

when a weaker cell drops voltage...you can charge it with 6A, but say your car wants 50A...no-go...

In this situation if the cell can supply 44 amps it's voltage won't drop. if it cannot supply 44 amps it's voltage will drop to where it can supply the 44 amps, you raise a good point, I'll see if i can " alarm" on this situation.


and now you say you -do- use cell groups to get the power from... so group1 can get out of balance with group x

Not really,because it is part of the pack average of cell voltage and is charging itself to maintain the average.


Plus, when charging 15 cells from a 16 cell group...well..when not counting the losses...you're basically discharging the 1 cell. Losses make it so that lots of heat is put out and this cell-group will drop below other groups.

You are absolutely correct that charging 15 cells in a 16 cell group is effectively discharging the remaining cell however, the losses, in the chargers, are quite surprisingly small it takes about 2 watts to charge at 6 amps so this situation would result in a total of about 30 watts of dissipation to which we must add the heating of the cells.

An active BMS -might- make sense in a 4 cell battery...but there also the cells should just be the same, not needing active balancing.
The active balancing systems I've seen are much less capable, most offer balancing currents in milliamps. This should cost about what the "Mini BMS" cost.
 

·
Registered
Joined
·
102 Posts
Discussion Starter #15
Thank you boekel!
I have a way of "alarming" when any cell is pulled below say 2.5 volts and in my mind this is a much more valuable alarm than a charger cutoff. To have both a low cell and a high cell alarm would add more cost but can be done. Are both needed?
 

·
Registered
Joined
·
565 Posts
Yes you need both, or you'll destroy cells the same way many diy projects without these functions have destroyed cells..

that's the main funcion of a bms: stopping discharge when a cell reaches minimum voltage, and stop charge when a cell reaches maximum voltage.
Battery Monitoring System is the minimum you need to prevent damage (and depending on chemistry: fire!)

balancing is next, to keep pack (top) balanced over time...
 

·
Registered
Joined
·
102 Posts
Discussion Starter #17
Yes you need both, or you'll destroy cells the same way many diy projects without these functions have destroyed cells..

that's the main funcion of a bms: stopping discharge when a cell reaches minimum voltage, and stop charge when a cell reaches maximum voltage.
Battery Monitoring System is the minimum you need to prevent damage (and depending on chemistry: fire!)

balancing is next, to keep pack (top) balanced over time...
Why not keep it balanced full time with charge only active balancing? The cost impact of alarming on both conditions turns out to be about $0.10, truely negligible. Now my questions are what " feel good" leds should I incorporate?
I've also added one "jumper" to select 4 cell or 16 cell power, this is needed to take advantage of automatic power down which will reduce the power to 5 milliwats per cell but an external switch to control power is much better it would reduce the parasitic load to pico watts of "leakage" current. The automatic power down would drain the cells to between 2.1 and 2.4 volts before powering down but it only costs about $0.02 to implement a bit of added operator error "grace"..
 

·
Registered
Joined
·
565 Posts
Just a little green led (not blinding) telling everything is ok...and a big red one for warning...followed by shutdown of the system.

'why not keep it balanced with charge only balancing'
well, I've made a nice drawing for you to explain why not...

But please trust me: bottom balancing is just not right, most of the time the battery will be 'full' (not driving, connected to the grid) and if necessary the bms has lots of time to balance.

balancing when battery is half-full doesn't work well, because the cells are on the 'flat' part of the curve, especially with LFP. bottom balancing can only be done with an empty battery....so why would you do that?
Then when charging, the weakest cell determines that your charger has to stop before the rest is full. and no, there is no way you can add enough current to the other cells to overcome this.

just top-balance, and throw out the cells that are bad. overall voltage will be lower, but usable capacity will increase.
 

·
Registered
Joined
·
102 Posts
Discussion Starter #19
Just a little green led (not blinding) telling everything is ok...and a big red one for warning...followed by shutdown of the system.

'why not keep it balanced with charge only balancing'
well, I've made a nice drawing for you to explain why not...

But please trust me: bottom balancing is just not right, most of the time the battery will be 'full' (not driving, connected to the grid) and if necessary the bms has lots of time to balance.

balancing when battery is half-full doesn't work well, because the cells are on the 'flat' part of the curve, especially with LFP. bottom balancing can only be done with an empty battery....so why would you do that?

But weaker cells show themselves by higher internal resistance so supplemental charging lowers their share of the load.

Then when charging, the weakest cell determines that your charger has to stop before the rest is full. and no, there is no way you can add enough current to the other cells to overcome this.

It depends on how fast you are charging, charge at 3 amps and this design will ALWAYS add enough charge to achieve top balance and all cells totally full at the same time if the cells have less than a 2 to 1 capacity ratio.

just top-balance, and throw out the cells that are bad. overall voltage will be lower, but usable capacity will increase.
I don't think you, quite yet, comprehend that what charge only full time active balancing does. As the pack is charged it is driven toward a "top balanced state" AND as it is discharged it is driven toward a " bottom balanced state"
 

·
Registered
Joined
·
565 Posts
I don't think you, quite yet, comprehend that what charge only full time active balancing does. As the pack is charged it is driven toward a "top balanced state" AND as it is discharged it is driven toward a " bottom balanced state"
Than please explain it to me, with a nice drawing like I made, you could use the drawing and write some numbers on it...

say the 'good' cells are 100 Ah, and the 'bad' one is 90Ah..
 
1 - 20 of 94 Posts
Top