Bottom balancing?

[dogman dan]

Nevertheless, it only matters that you know when the lowest capacity cell group is on the edge of the cliff.

Charging the rest of you pack to match that cell groups lower capacity will not put more capacity into the lowest capacity group. You will still only have a pack with the capacity of the weakest cell group.

But have fun anyway. It's not that what you are talking about is wrong, it's just not going to make the weak cell group stronger. Given the way you build your packs, I'd be inclined to just add one more cell to a weak group. That would add capacity to it, allowing you to draw the full capacity of the remaining cell groups.

True bottom balancing, where you discharge all cells way past the cliff, that's what we argue against.

 

[richardriver]

I have a pack of 36 cells CALB 130AH , which I BB at 2.70v and after about 300 cycles /12000+ miles the cells are behaving exactly the same. When I am charging, I have cut off charge 124.2v (3.45/cell ) and when stop driving when the voltage is below 105v ( which usually never let it go that low ). If the function of a BMS is to get every AH of the cells , instead on investing on $1000-$2000 buy more cells to extend the capacity. Yes, BB is inconvenient at the beginning + you wont take advantage of about 3-5% of extra capacity of top balancing, but In my case I don't see the necessity of a BMS. Now got tired of monitoring my cells every day for the firts months....now after 2 years batteries are behaving the same. Maybe my batteries are MAGIC!

 

[dogman dan]

Not magic. You are just using them in such a way that they don't get terribly out of balance in the first place I bet.

What's your typical dod? Is it 99%? If it's below 80% dod, only a defective cell would be likely to be getting out of balance at 300 cycles.

If my guess is correct, you just proved that less than 100% discharges work good, and you didn't get a defect cell. Nothing was proved about balancing whatsoever.

 

[999zip999]

I think it's unfair and going to call Jack Rickard and have him give some wind. Becareful what you wish for as it would be a very windy day.

 

[dogman dan]

I'm not a huge fan of fussy fuss top balancing either. Nor do I advocate blindly trusting any bms.

What I do preach more than anything else is the following. The short version is don't hammer your battery, and it stays better balanced and may last longer.



Have enough capacity to make your usual discharge rate much less than factory spec maximums, and have reserve capacity when you need it.

Keep dod moderate, so that .1v out of balance will not cause a disaster. If conventient, make most charges a slight undercharge as well.

Balance the pack at the top if it gets more than 1.v out of balance. Ideally, charge just the low cell once it gets that far out, or discharge if just one cell is high. BMS balancing can take forever.

Monitor at cell level often enough to know what is happening, and if possible, replace cells that cause a daily imbalance.

Don't store a battery long periods with anything whatsoever drawing current. If you must store with a bms or other device drawing current, recharge at a sufficient interval for the the bms to restore balance.

Bring on the wind, anybody wants to argue against this strategy. It kinda works. Works good enough to make balancing needed a lot less, and the conservative dod makes how you do it less of an issue. BMS or not matters less too. A bms can't murder the battery if it never gets much of a chance to try it.

 

[Hillhater]

+1 for Dogmans approach.
all i would add is that a "wisely" set LV alarm can be a useful addition.

 

[Doctorbass]

I'm not a huge fan of fussy fuss top balancing either. Nor do I advocate blindly trusting any bms.

What I do preach more than anything else is the following. The short version is don't hammer your battery, and it stays better balanced and may last longer.



Have enough capacity to make your usual discharge rate much less than factory spec maximums, and have reserve capacity when you need it.

Keep dod moderate, so that .1v out of balance will not cause a disaster. If conventient, make most charges a slight undercharge as well.

Balance the pack at the top if it gets more than 1.v out of balance. Ideally, charge just the low cell once it gets that far out, or discharge if just one cell is high. BMS balancing can take forever.

Monitor at cell level often enough to know what is happening, and if possible, replace cells that cause a daily imbalance.

Don't store a battery long periods with anything whatsoever drawing current. If you must store with a bms or other device drawing current, recharge at a sufficient interval for the the bms to restore balance.

Bring on the wind, anybody wants to argue against this strategy. It kinda works. Works good enough to make balancing needed a lot less, and the conservative dod makes how you do it less of an issue. BMS or not matters less too. A bms can't murder the battery if it never gets much of a chance to try it.

+1 for dogman i agree!

This i sthe aproach i've been using since the last 5 years now with the Konion and it work perfect!

This is the approach i would recomand to everyone using lithium cellas and also people that buy Konion cells.

People that ask me for konion often will calculate theyr battery size according to the total trip or energy they need and i have to remind them that discharging a pack fully is not a great solution for keeping balanced and in a good health their cells. Max of DOD whould be 80% or less

Also many will calculate the max current theyr controller can do and will calculate the battery they make by having the number of parallel cells that match this but they forget to have a margin! I recommand to have 150% more curent available form the battery at continuous discharge than what the controller can take.

Ex 1p konion 18650V cell can do 7A continuous and 15 burst until it reach 50 celsius. this mean that your controller should draw about 4.5A per cell i parallel. so if you have 50a controller, you should have about 75A max continuous discharge from your battery wich correspond to 75A / 7A = 10p cells. so for a 50A controller, i would recommand 10p of 18650V cells.

I also agree with Luke. By top balancing you use your cells in a way that you get more Wh energy per volt since you have a larger part of the dishcarge that occure at higher voltage than with bottom balancing The Ri is lower at higher voltage so the voltage sag is less so it translate that you get more Wh per volt of discharge since more Ah are at higher voltage.

I other words in top balancing, the first Ah you get is at higher voltage meaning higher Wh ( Ah x volt = Wh) than the first Ah with bottom balancing that have discharged at lower voltage due to the lower cell at the top that limit the total pack max voltage voltage.

The energy of each Ah depend on the voltage at wich it is charged or discharged.. and this is what make diference in this bottom or top balancing situation.

I wold thing you can get a little bit more energy( wh) by using top balancing than bottom balancing.

Doc

 

[headwind]

Holy necro-post ahead! (stumbled across this thread due to curiosity about bottom balancing

This lecture has lots of good stuff. He doesn't believe in BMS and top balancing

Fantastic vid, thanks for posting.

Hope this doesn't sound argumentative, but I took a different meaning from when he mentioned BMSes and balancing in passing. He said something like "BMS is the engineering solution, but for mass production it's too expensive". The context of the lecture was an electric car project, which has a huge number of cells. And since he recommended they monitor each cell, I think he was referring to automated systems which balance* the cells. His recommendation was to manually correct any balance issues.

* does BMS = Battery Monitoring System, or Battery Management System (ie something like a cellog to watch the cell voltages, or a devices which also actively balance cells)?
I think a lot of confusion is caused by not having a set definition of what the "M" stands for.

Cheers,
Simon

 

[dogman dan]

I always thought bms meant a monitoring, balancing, and individual cell lvc system. So high cell starts a discharger on that cell, and low cell shuts off the main discharge circuit for the whole pack.

It generally does not include a voltage display, but I often wish they did. This is what I mean by watch your bms. It's doing something, but even fresh out of the box is it doing what you wanted? Only by watching voltages as the bms works do you know what it's doing. The prime battery killing defect of a bms is when the balancing discharger fails in the mode that keeps discharging till the cell is at 0v.

The other classic bms problem is when you store a pack too long, and the bms discharges one or two cell groups with the tiny trickle of power that runs the bms. For some reason, that circuit does not have an lvc cutoff.

Sometimes other trickle shorts are present in an assembly, and the battery self discharges. Invariably, people blame the bms. Sometimes a cell grows dendrites and self discahrges. Invariably, people blame the bms.

Bms or not is a fairly cheap decision for a 20 amps continuous discharge EV. But when you get into cars, the amps involved greatly increases the cost. The cost of having a bms actually fail and ruin a pack gets pretty painful too. If I had a 20,000 car battery, I'd want to watch it like a hawk too.

For them, maybe bottom balancing makes this easier. Goes without saying, they are trying not to overdischarge too. I think many may misunderstand Jack, and think he means discharge that deep every cycle. That would be dumb. That deep discharge and bottom balance might be more like an annual thing. Jack and I may be very similar, I'm no electronic engineer. But some stuff I learned from folks who are I have converted into a common sense way to manage a small battery pack. Sum it up very simple, don't hammer your battery if you want it to last. Easy on it and it tends to stay balanced, and need less watching.

 

[999zip999]

I add sense wires to a bms lifepo4 pack so to check, monitor and help balance if needed. Just like to know...

 

[dogman dan]

I did that too, during the last months of my Ping lifepo4 packs life. I knew I had some bad cell groups, and adding a set of 8s jst plugs allowed me to watch what the cells and the bms were doing as it charged.

I could tell things were not right already, by the way the blinky lights on the bms were acting. The weak cells filled a very long time before the ones that still had capacity.

Adding the ability to use a cellog 8 on that pack made it easy for me to see the problem was the cells, not the bms malfunctioning. The bms had a hard time keeping up with the overcharge of the low capacity cells, so I started using a turn signal light bulb on the low capacity cells to discharge them faster as the rest of the pack charged. It helped them not overcharge so much each cycle. I did that for the last 10 cycles or so, before chopping up the pack into some still fairly good 12v sections. All the bad cells were in one section of the 48v pack.

 

[silviasol]

Anyone know where to find that bottom balancing tool that Jack Rickard of EVTV uses? The pack I built linked in my sig has cells from my old 10ah pack(with low cycles) inside it and stupid me but I left the pack at full charge over the winter, probably 4 months, so I have cells that are not charging to 4.2v completely and others that do so I wind up after a full charge even with bms from 4.05v to 4.25v each cell. Wanted to bottom balance to see if that helps. If you don't know what it is this tool will discharge a cell to 2.5v then shut off then power back on when the cell reaches 2.75v and goes again and again until it sits at 2.5v.



https://www.youtube.com/channel/UCtSO7VGybs6U__X-Y8l_QmQ

 

[John in CR]

Not someone to emulate. Bottom balancing would kill your weak cells faster than anything, because your lowest capacity cells will get overcharged every cycle. What kind of cells are you using that storing them at full charge once did that much damage?

 

[silviasol]

Not someone to emulate. Bottom balancing would kill your weak cells faster than anything, because your lowest capacity cells will get overcharged every cycle. What kind of cells are you using that storing them at full charge once did that much damage?

These are turnigy 5ah cells. I don't think it would do any damage or make any difference to discharge them all to 2.5v since the cells have only 10 to 20 cycles on them, despite them being stored over winter at full charge they are pretty much new cells. Storage is 3.6v to 3.8v correct? Over the 4 or so months I did not use the pack it was full charge 4.2v each cell.

I am currently using a lower amp charger to slowly charge them to 4.2v per cell. Just powering it on and off after it finishes to let the bms catch up. Almost to 4.2v per cell then take it for a run and see what happens.

I think this bms is not meant for anything larger then 10ah and I am using 20ah which may be my problem on the balancing side, it just can't keep up with the amount of amps. So I figure try to bottom balance them.

http://hobbyking.com/hobbyking/store/__18560__Turnigy_5000mAh_1S_20C_Lipoly_Single_Cell_.html

 

[silviasol]

The problem was entirely the PCM/bms. I had it stuffed in the back of te pan and the pressure from the deck was making it shut off at lower amps and must have drained certain cells somehow. Just went on a ride after adjusting them and all cells no more then .05v fifference. Yay! 8)

 

[bgt]

Where to find a BMS, for Lifepo4 batteries, which starts balancing when the battery switches off(empty state) so the batteries all start from the same low state f.i. 2.75V per cell? They mostly balance at the top of the charging curve.

 

[friendly1uk]

I think the 0z890 units can. They are 5-13s. I can only think of reasons not to do it though. Like having to run out of charge regularly. Or that the cells powering the bms will hit lvc first, then be expected to keep powering the bms while other cells are trimmed own. There was a real good reason not to do it, but my brains not engaging.

 

[pohjeremy]

Hi All,

I'd like to maximize the lifespans of the CAM72s without BMS and charging each cell with an isolated individual charger. Newbie but pretty keen. E-scooter replacement for Pb. 32s1p.

Charger Info:
- 32 Laptop style small 4A chargers.
- DC output reads between 3.51 and 3.56V. I cast the other 8 (out of this range) aside.
- Primitive individual cell charging style...yes, I have the doggone patience/sense to plug 32 bananas, then switch on the AC.
- I could buy a 32s charger if the above is stupid...
- Idea was...Since I use near 15Ah/day, I could charge once every 3 or 4 days. Occasionally I might use near 50Ah for longer journeys.

Monitoring Info:
- Again, primitive...32 near-identical mini-voltmeters (tested accurate to 0.005V within all 32), one on each cell.

BMS info:
No BMS; They are more trouble than is worth, but it's just my take.

Cell Info:
-36 CAM72s...(please see attachments)...I had hounded a CALB friend to provide such a selection. Hope at least on paper, they appear acceptable. I trust that CALB friend.
-4 spare cells; I might not use cell no 23 though.

The Qs:
1. Based on the specs, should I still bottom balance them at 2.75V..?
2. The individual chargers have a range...3.51 - 3.56V. How will this affect discharging patterns.
3. Should I get a 32s Charger..?
4. Should I place a mini-voltmeter on each cell?

Thanks.. You objectivity is appreciated!

Jeremy P.

 

[Ohbse]

32 individual chargers is madness and invites disaster. Many things to go wrong, many opportunities for human error to lead to failures. Time consuming, bulky and needlessly complex.

You should definitely get a suitable bulk charger and some sort of monitoring solution at cell level that isn't 32 volt meters.

What's the motor/power level you're running at?

 

[dnmun]

edited: i think i was sober too.

 

[Ohbse]

Thinking about this further and looking at the specifications of the CAM72's leads to more thoughts and more questions..

If you're only using 15ah a day, you're going to be much better off in every respect to run with a smaller, higher power pack and charge on a daily basis, especially if your charging solution is single plug/single switch and it's not a big deal. Do you do longer trips on a regular basis? Enough to justify paying for and carrying around 30+KG of cells you don't require?

For comparison

32s CAM72 72ah - 7.4KWh, 61KG, something like 150amps current continuous (spec is a bit hazy)
28s Lipo 30ah - 3.1KWh, 25KG, 300+ amps continuous, taking up only 40% of the space. Price ~$2k vs $4.5k sticker price on the CALB cells.

 

[pohjeremy]

Thanks for the replies.

Carrying the excess weight is fine as the motor and controller can handle it. There's a potential to go further (should I desire) and weight adds riding stability (up to a point). Zero Motorcycle's 7.5kWh is about the same weight or slightly heavier the last time I checked. I do long trips about once/month.

A 32s (115.2V) charger is readily available but if without a BMS, I'm not sure if any cell won't breach the 3.6V mark, especially near end stage charging. Using C11000 copper bus-bars and bolts....v low resistance. How would a 115V, 20A charger ensure no cell crosses 3.6V, whether I bottom balance or not?

Correct me if I'm wrong but I firmly believe (and I've tried) that the longer a LFP cell stays out of the 2.6 - 3.45V range, the more inherent damage it takes, the shorter it's lifespan. Trying to prevent that.

Cheap BMSes have a high chance of failure, especially if I throttle at 130A. They also tend to balance past the 3.6V mark.
In fact the last round, I emptied by pockets for this (see pics) and
1. It did not distribute at 3.5 - 3.6V.
2. It made some cells cross 3.8V (Lifespan shortened) , yet many other cells remained near 3.35V....they just won't go up!
3. It took up lots of space.

I'm not talking about a $60 BMS, but more like a $1500 one. No BMSes for me; They complicate and compound what can be a straightforward set-up. They get me stuck in the middle of a motorway too often.

The pictures are of a 30s1p CA60Fl set-up a year ago. The BMS was not a solution to the problem, it was the problem.

What human error will there be (besides the inconvenience I bear) if I've already measured all the voltages of the 32 chargers, marked them and assigned each to care for it's own battery? It will never overcharge past 3.55V.

As for the voltmeters, the same, just attach them and leave 'em there...so I can gauge the approximate capacity of any cell any point in time.

 

[pohjeremy]

Oh yes,

As for Lipo, they're all right...compact and light. Except that I've already bought CAM72 and..

Lipos might have a shorter DOD
Lipos might invite a fire more than an LFP might.

Hey, thanks for the replies..

 

[dmwahl]

Which BMS were you using that allowed that to happen? If it had access to each cell voltage as it obviously did, it should be able to shut off the charger well before the cells went that high.

 

[dnmun]

edit, negative comments reconsidered.