Bottom balancing?

Jeremy, i`ve also had experience with lipos drifting out of balance. Most new rc lipo packs now have balancing. but in the case of a larger capacity lifepo4 cell my experiences are similar to Jacks. i think high C or high capacity lifepo4 don`t need a lot of help. not the constant monitoring. once you have a stable pack, you are set.
 
I think the key part is as both of us have said, in your words "once you have a stable pack". The problems arise for those who don't start off with a stable pack, if they follow Jack Rickard's daft advice they may well end up ruining a costly investment. This is where I think he's being irresponsible - to him trashing a few thousand dollars worth of cells is no issue, for others it might be a year or so's worth of savings.

If you know what you're doing, start off with a well-matched set of cells, charge them properly and keep a check on how far you discharge the pack then I think you can pretty much get away without a BMS (I don't run a BMS on anything at the moment). If you start off with cells as they leave the factory, don't check their state of charge when you build the pack, don't take the trouble to balance them initially and check their capacity, then I think you'll run into big problems without a BMS.

I don't think there's a difference between LiPo and LiFePO4. My Headways are pretty good at staying in balance, but they did arrive in varying states of charge and there was a significant spread of capacities across the 64 cells I originally bought (less the handful that have now died). In many respects the Headways were a bit like the Turnigys I've just bought, pretty variable as they were delivered.

Jeremy
 
" Once you have a stable pack" sure is the key item. After that Knowing, not thinking or assuming the pack has all the cells charged.


I just had the exact same experience with a new set of zippy's from HK. One of the 8 new packs had a lazy cell on arrival. A few light cycles and it just wasn't perking up. I ended up putting a cell phone charger to it for a bit, and got it ahead of the rest of the pack dangit, and then it was woke up and was charging like crazy. Niether of my smart lipo chargers could tame it, so I had to put a light bulb on the one cell for a while.

Pain in the rump, but now it's behaving fine. One unruly cell just needed a bit more to break in than the others. How many ping packs or v power packs come the same way, who knows. But riding off on a 20 mile ride on brand new batteries trusting an untried bms is not wise, till you know how the cells are acting.

Another thing, for some reason my 4s turnigy's are stinkers. Exactly the same as my 5s turnigy's that are great. 3 of 3 4s packs puffed up. One my bad for sure, but the other two, one puffy on arrival, and the other puffy after about 20 cycles. Wierd. They still work, for now. Good ol HK quality. Good thing everything they sell is so inexpensively priced.
 
I used a 12v tailight bulb to discharge the single cell. A matter of what I had laying around. I used the balance plug to get at the cell, so I had to watch the heating of that wire.
 
A BMS that works is a fantastic and helpful thing that doesn't cause any problems at all.

Unfortunately, many are built to a low enough quality/design to be as much of a liability/hazard to the pack as any other danger the cells can encounter.


A properly working and well installed quality BMS is an asset to a pack, very much an asset for a pack being put in the hands of the general public.

That said, I've never used a BMS for any of my own packs, and 100% agree with Jeremy (on all aspects of his post), if you ensure they are balanced when finished charging, and are careful not to over-discharge them, you have no need for a BMS.
 
Those that are serious about designing a better BMS should acquaint themselves with MIL Handbook 217. It is available for a free download here http://www.quanterion.com/Publications/MIL-HDBK-217/index.asp

About the only thing with a zero failure mode is a Wicked Infinitesimal Resistance Element = WIRE!

Everything has a finite failure mode, some examples of failures rates per one million hours:
Fixed composition resistor 0.0017/1,000,000 hours
Film resistor 0.0037/1,000,000 hours
... those are at 25 deg C. Raise the temperature up to 100 deg C and multiply those failure coefficients by 4.8!

General purpose ceramic capacitor 0.0036/1,000,000 hours
Electrolytic Capacitor 0.029/1,000,000 hours
... these are also multiplied by a temperature, capacitance, and most important a VOLTAGE rating factor. For example for use voltage/rated voltage = Vs factor = 0.3 multiply by 1, for Vs = 0.9 multiply by 8.6, for Vs = 1.0 multiply by 14! So, see how part derating adds reliability?

FET 0.01/1,000,000 hours
Microprocessor 0.28/1,000,000 hours
Commercial Cylindrical Connector (mated pair) 0.002/1,000,000 hours per pin pair
Commercial Card Edge Connector (mated) 0.080/1,000,000 hours per pin pair

You get the idea. Acquainting yourself with real world failure data will guide you in what to avoid, and what to use in your design. When using a Mil HDBK 217 parts count approach, the number of parts does matter, as they add towards the composite failure rate. Simpler is better from a reliability standpoint... if the simple(r) design can do the job though!
 
Let's set aside the extra cost, complexity, and failure points. I question whether the top balancing approach typically used is even the correct thing to do. It seems to me that approach minimizes the pack's usable capacity as well as increases the cycle wear on the weakest cells. I completely understand that our packs need to be monitored because the cells have a safe voltage range, but I think these systems need to be smarter before they can manage the pack better than I can myself. Is anyone here at 2000 cycles on a pack yet without a BMS induced failure or problem? Make me a believer.
 
How would a top balancing approach minimize usable capacity? It should do the opposite by ensuring that all cells are at the same SOC at the "top", giving maximal capacity for all cells.
 
Oh.. again.. that Jack anti BMS thing :roll:

He might never use BMS... just like me with my lipo and konion ... but... to do that i must keep a close eye on my cells any time i charge and discharge them...

( it turn out that it's ME that ACT AS A BMS... when watching the cell voltages lol)

And... he admited itself that he monitor the cell voltages sometime..... I wonder why.... :p :roll:

A BMS is a great device that when working properly SHOULD AVOID YOU TO DO THAT! :wink:

Doc
 
There seems to be a lot of people here with stories about packs killed by wayward BMS systems. Also a lot of people taking about how they killed a pack in a manner that a BMS would have prevented. People that use BMS without problems never complain, so the real question is what are the ratios of killed packs to happy packs for each.

My opinions are: If you don't use a BMS, you will eventually kill your pack. All packs need to be kept balanced. Humans are terrible BMSes. Humans are great pack killers. Reliable BMS systems are possible. Beware of Chinese BMSes. If you have to plug and unplug connectors to charge your pack, you are doomed... the more connectors involved, the more epic the doom.
 
texaspyro said:
There seems to be a lot of people here with stories about packs killed by wayward BMS systems. Also a lot of people taking about how they killed a pack in a manner that a BMS would have prevented. People that use BMS without problems never complain, so the real question is what are the ratios of killed packs to happy packs for each.

My opinions are: If you don't use a BMS, you will eventually kill your pack. All packs need to be kept balanced. Humans are terrible BMSes. Humans are great pack killers. Reliable BMS systems are possible. Beware of Chinese BMSes. If you have to plug and unplug connectors to charge your pack, you are doomed... the more connectors involved, the more epic the doom.
Click to expand...


That's totally right Tp!. but some human are good BMS :wink:
 
the more connectors involved, the more epic the doom.
Click to expand...

If this is true, then none of us should be using Lipos. My 12s3p lipo pack needs 6 connections and i have to make sure the charger is in balance mode and have properly identified the number of cells.

I think we shouldn't be lumping all lithium batteries together. There is clearly a difference between Lipos and Lifepo4. The C rating and actual size of the battery also seems to be a factor.

You know, it's funny we never talk about the need for a BMS for a SLA but I've checked the balance on SLA cells before and sometimes it worries me enough to try to manually balance them with my power supply.

Another good question is how balanced is balanced? We all know that there is a "Knee" in the charge curve where the Lifepo4 cell voltage will start to take off. Cells can go from 3.5v to 4.2v in milliamps so we shouldn't necessarily use voltage as the sole indicator of state of charge (especially not hot off the charger).

As I indicated in my first post, the voltage at 100% discharge is not the best indicator of the health of the cell either.

I agree that If we're talking about a pack that can be sold to the public, they just need to plug it in and forget it. I believe Headways cells (or higher C/higher capacity Lifepo4 cells) can be made into this pack without a BMS.
 
I "top" balance my lithium. I understand all the stuff being said about voltage not being the end all indicator to cell health.. but the thing is I still don't see any better way to do it??

A bms smart enough to automatically switch out a low bank of paralleled cells on the fly might be worth some high dollars.. but until then I fear I'll be a member of the human bms club!

Well I guess they will become affordable eventually.
 
johnrobholmes said:
How would a top balancing approach minimize usable capacity? It should do the opposite by ensuring that all cells are at the same SOC at the "top", giving maximal capacity for all cells.
Click to expand...

The lowest capacity cells reach higher voltage first, so if you drain that extra charge off then they have less energy stored and will trip the cell low voltage cutoff sooner than it would have without balancing. It brings your pack down to what is stored in your worst cell group. I realize that's looking only at capacity, not IR, which would have the opposite effect by giving more of the charge input up to heat instead of energy stored. I view top balancing as a fairly crude approach, not an optimum. There must be some variability in cells and the voltage at full charge is not identical for all cells, so I don't agree that all will have max SOC at the same voltage. If it was in fact the maximum, then there wouldn't be anything to bleed off to help the others catch up.

For those running their packs at 90% DOD or more on a regular basis I don't doubt that an active BMS and balance charging is mandatory, but with conservative charge/discharge I'm not so sure. I look at running your tank low as something to avoid just like in a car, and until BMS's are closer to absolutely reliable and are a lot cheaper in relation to the batteries they're supposed to protect, then I think it makes more sense to spend that $ on more capacity given the nature of my use. I realize it's not for everyone, and maybe I'll be singing a different tune once I have cells nearing the end of their useful life and my loose approach isn't enough. In the meantime, my only dead lithium cells were killed by the systems intended to protect them.

John
 
A lower capacity cell will limit the pack no matter what, it can be both the first to reach high voltage and the first to reach low voltage. While I agree that top balancing may not get any beneficial runtime increases when dealing with a runt cell (it is the weak link on all accounts of capacity), without top balancing you either need to cut off the charge when one cell hits HVC or risk overcharging the "runt".


If we can agree that SOC is a %, then all cells should be at the same SOC at a certain voltage. Indeed a runt cell will fall faster than the rest in voltage (since capacity is lower) if all are brought up to the same voltage. If nothing is gained from top balancing besides the ability to work the runt in the upper range of voltage, we can at least have the knowledge of what cells may be problems.



Good discussion. Makes me think hard about what balancing actually does for a pack 8)
 
Agreed 100%, John!

A runt cell in a 6s pack is going to have less capacity than a robust cell. The runt cell is going to charge from empty to full in less time. Knocking its top off is only reducing its capacity that much faster.

Edit: Also, I charge my lipo to 4.1V, so I can allow the runts to go higher and closer match the robust cells capacity.

John in CR said:
johnrobholmes said:
How would a top balancing approach minimize usable capacity? It should do the opposite by ensuring that all cells are at the same SOC at the "top", giving maximal capacity for all cells.
Click to expand...

The lowest capacity cells reach higher voltage first, so if you drain that extra charge off then they have less energy stored and will trip the cell low voltage cutoff sooner than it would have without balancing. It brings your pack down to what is stored in your worst cell group. I realize that's looking only at capacity, not IR, which would have the opposite effect by giving more of the charge input up to heat instead of energy stored. I view top balancing as a fairly crude approach, not an optimum. There must be some variability in cells and the voltage at full charge is not identical for all cells, so I don't agree that all will have max SOC at the same voltage. If it was in fact the maximum, then there wouldn't be anything to bleed off to help the others catch up.

For those running their packs at 90% DOD or more on a regular basis I don't doubt that an active BMS and balance charging is mandatory, but with conservative charge/discharge I'm not so sure. I look at running your tank low as something to avoid just like in a car, and until BMS's are closer to absolutely reliable and are a lot cheaper in relation to the batteries they're supposed to protect, then I think it makes more sense to spend that $ on more capacity given the nature of my use. I realize it's not for everyone, and maybe I'll be singing a different tune once I have cells nearing the end of their useful life and my loose approach isn't enough. In the meantime, my only dead lithium cells were killed by the systems intended to protect them.

John
Click to expand...
 
I set out to collect 'real world' data to see if pack-level LVC and bulk charging was 'good enough' for a lithium pack.

Proof that controller-level LVC is NOT enough can be seen here:

http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=8675&start=30#p135632
file.php


I was able to get an additional eight months of daily use from this pack AFTER damaging a cell because I added cell-level LVC after my "stupid user trick."
http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=17035&start=45#p320679

http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=17035&start=45#p320679
file.php


http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=17035&start=60#p323518

There's no way I will charge the pack without some type of charge control and cell-level HVC, and I won't run without cell-level LVC that reduces the load on the pack - beepers or flashing lights are not enough.

Listen to Doc! Choose automatic or manual - but packs need to be managed one way or another.
 
Dmnun, I told ping that the # 2 fet from the p- con. on the bms had melded it's solder on discharge. At that piont Ping told me to replace it.
 
999zip999 said:
Dmnun, I told ping that the # 2 fet from the p- con. on the bms had melded it's solder on discharge. At that piont Ping told me to replace it.
Click to expand...


you should just order a new v1 signalab from ping. unsoldering the mosfets is really hard and then you need to replace all 4 at the same time since you don't know if the others will have the same problem.

you do not have a functional output as it is now? have a picture?

i have heard of the charging FET getting too hot and the v1 i got from greg had all the mosfets burned totally off along with the 14 pin comparator but he was using it in series with the lipo when he first went over to them.
 
Andy,

I think we can all agree that some management is necessary, and that blindly fully depleting a pack with regularity and only bulk charging would be a costly mistake. I hope that no one takes my posts as suggesting that is the way to go. My primary packs are LiMN Konion cells and I am intimately familiar with them, so much so that I can tell if a pack is out of balance just by the way voltage changes during bulk charging.

The pack on my bike for over a year and a half has been sealed up and only bulk charged for that entire period. Other than a handful of extreme weather days it has been used every day with an average of nearly two 30-50% DOD cycles each day, so I must be nearing 1000 cycles on cells that were used when I got them and are rated at a 500 cycle life in their higher demand design use. That's with no balance charging or even checking them for balance. I didn't even build the packs in what was considered the right way, since there is no parallel structure within the pack other than the 2p in each 20s string as they came from the factory in the Bosch and Makita packs.

For a year I ran the bike without even knowing the voltage while riding. That resulted in pushing it to 100% DOD on 3 occasions. I could tell that the pack was out of whack afterward because the voltage climbed rapidly from 80V to my 81V cutoff, indicating that some of the cells at higher voltage than others. Noticing that difference in the charge profile, and following it up with a couple of weeks using a timer to cut the charge off early to avoid charging the high cells to too high a voltage while the pack naturally returns to reasonable balance, is the sum total of my BMS. I take that back, I do feel the duct tape covering of the pack with some regularity to make sure there are no hot spots that would indicate a problem that needs immediate attention.

It's been suggested (even by myself) that Konion cells self balance to some extent. They must, since Bosch and Makita don't have any active cell balancing or cell level monitoring in their tool packs. While self balancing seems to occur, I believe it only takes place in the form that at top of charge the highest voltage cells give up some of that surface charge as heat after or during charging. Don't most cells behave in that same manner though? If so, why can't a similar approach work with Lipo and Lifepo4 by avoiding deep discharges and using a conservative cutoff voltage for bulk charging?

I understand how a pack gets out of balance with deep discharges, but with conservative DOD and charging, what would cause an out of balance? Judging by the voltage sag with Konions I'd be surprised IR differences are root of the cause, since I have to think my pack pack has a lot more IR variability due to the higher overall IR compared to other chemistries.
 
AndyH said:
I set out to collect 'real world' data to see if pack-level LVC and bulk charging was 'good enough' for a lithium pack.

Proof that controller-level LVC is NOT enough can be seen here:

http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=8675&start=30#p135632
file.php


I was able to get an additional eight months of daily use from this pack AFTER damaging a cell because I added cell-level LVC after my "stupid user trick."
http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=17035&start=45#p320679

http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=17035&start=45#p320679
file.php


http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=17035&start=60#p323518

There's no way I will charge the pack without some type of charge control and cell-level HVC, and I won't run without cell-level LVC that reduces the load on the pack - beepers or flashing lights are not enough.

Listen to Doc! Choose automatic or manual - but packs need to be managed one way or another.
Click to expand...

I agree that it would be crazy not to monitor/manage your batteries while (dis)charging. I prefer the celllog8-alarm active cut-off method Gary and Doc Bass have been advocating. It gives you all the necessary signals to cut-off/limit operation, with visual readout, plus the option of datalogging via usb. It is quickly interchangable between Lipo and LiFePo4.

There are other ways of coarse! Unless you like to try to through away 70 K, like Jack!
 
You must log in or register to reply here.

Similar threads

digitalbilly
Garmata EEB battery box says it can hold 400pcs 18650 cells. I think I can fit more.
Replies
7
Views
365
digitalbilly
digitalbilly
B
Parallel Charging into three different batteries with separate BMS's
Replies
10
Views
415
999zip999
999zip999
O
Lifepo4 cylinder 32650 cells simplified stationary design idea?
Replies
7
Views
268
jonescg
jonescg
E
Battery Pack Series Connection Layout Help
Replies
2
Views
327
Eruck
E
B
Battery pack shuts off after a while
Replies
2
Views
312
bogddy
B
Back
Top