Is this 2009 info still accurate?

**Mike90250** · 01-20-2019, 01:47 PM

Originally posted by zamboni

.... claiming they tolerate significant over-charging without degradation.......

I think I would lean a bit more conservatively and say, from what I understand, that LFP handles overcharge a bit more gracefully than overdischarge. But only a bit. Either will kill a cell, the overdischarge I think causes more fires when the cell is recharged.

And then you have temperature limits, below 40F you loose performance, and at 32F you essentially cannot recharge.

**Ampster** · 01-21-2019, 09:46 AM

I have been watchinting EVTV since 2010 and seen the evolution of Jack Rickards thinking. He has evolved from not believing a BMS was needed to focusing on selling interfaces to Tesla BMSs. I have been using a BMS to provide me information about the status of my packs since then. The debates go on and on about the benefits of bottom balancing To my knowledge none of the manufacturers of stationary Lithium packs bottom balance. They all use some form of BMS. In the end it gets down to how you want to manage the risk of damaging your pack.

**zamboni** · 01-21-2019, 02:42 PM

Ampster, perhaps my understanding is flawed; it seems to me cell balancing and use of BMS are not mutually exclusive things -- but they should be chosen to complement each other. You choose how to balance the cells when you build the pack, then choose to apply a BMS or not...whatever "BMS" means. I consider my system as having a "BMS" because I have an HVD and LVD relay triggered by both temperature (hi/lo) and string/pack voltage. It does no active balancing, however. This is the minimum amount of management I consider acceptable given the size and use of the pack (house battery for RV).

I am wary of claims that one way is always right, irrespective of the application. When both over-charging and over-discharging pose equal threats to cell life and safety, you choose the balancing that lines up best with your usage. My usage is similar to both an RE home and an EV because I do spend a lot of time "full" but also regularly go to "empty" (both of those being conservative setpoints, not extremes of the curve). So I vacillate between choosing top- and bottom- balancing. But if this 2009 info is correct about over-discharging being a more severe risk, the scales tip in favor of bottom-balancing (for me).

I also noted that Mr. Rickards did his "perfect top balance" at 4.000 VPC, which is higher than any of my cells have ever seen under charge -- even when the pack was very poorly balanced at the beginning of my learning curve. In contrast, he gives 2.85 VPC as a hard cutoff to prevent cell damage -- which is *really* close to 3.0 VPC. This makes me feel like I have a lot more safety margin at the top of the curve than the bottom.

**Ampster** · 01-21-2019, 08:51 PM

I am by no means the expert that others on this forum are but I can give you my practical experience. When I put a pack together I just parallel the cells at whatever SOC they are at to balance them. I think one well known member here called that mid balancing. If there is a significant difference between cells you need to be careful because of a potential for a large surge current.
What I do like about using a BMS is you can set a LVC and HVC for individual cells so even if your pack never hits the cutoff the errant cell will set off an alarm. And, yes, you are correct pack balancing someting that you will want to do whether you use a BMS or not. I just chose to do the simple mid balance rather than the more elaborate bottom balance that takes a lot more time.

**zamboni** · 01-22-2019, 09:30 PM

It sounds like we have the same ideas about what a BMS should be doing. I too like the idea of monitoring individual cells for over/under voltage as a precaution, regardless of the balancing strategy used in building the pack.

Thanks for your feedback specifically on EVTV / Mr. Rickard's "evolution." This makes for an interesting datapoint to consider for the future. I'm not sure that spending the time re-balancing my pack to the bottom is justified...but next time I have a reason to fool around with it, perhaps I will try.

**inetdog** · 01-29-2019, 04:12 AM

I would say that it is easy to make the obvious association that you should bottom balance if you expect, at any point in your routine operation, to be near the bottom (i.e. emergency power reserve). Conversely if you expect to spend significant time at the top (recharging close to 100% every day) then top balancing seems to make more sense.

**PNjunction** · 02-01-2019, 04:58 AM

Re "mid-balancing" by merely paralleling cells together. No such thing. Choose top or bottom. The reason is because of the very flat charge / discharge curve of LFP. Mid-balancing is a total oxymoron. One will find out soon enough when they try to fully charge, or fully discharge the pack.

Example: One individual cell can read 3.2v at rest when charged to 80%. Another individual cell can ALSO read 3.2v at rest when charged to only 20% capacity. Slap those two together in parallel to balance them. Since a voltage-differential is what drives current, and with no voltage differential between those cells, no charge is transferred to the other cell.

The flat charge / discharge curve is what makes the "stick analogy" of balancing a very general concept, but not good in the real world for LFP balance by voltage detection alone.

BOTTOM Balance recharge warning:
What a lot of EV'ers using LFP forget is that when you drop down into the discharge knee, that is below 3.2v at rest on a cell, then you MUST not apply any current higher than about 0.05C or LESS to the cell until it reaches the nominal usable 3.2v cell voltage just outside of the knee, and only then should you apply the rated current for recharge. You have to come out of it SLOWLY.

Why? When you recharge a cell that is well below the nominal voltage with too much current, it can't intercalate fast enough , er recharge efficiently. Too many ions and not enough holes to lodge back into. This shotgun blast of ions is also accompanied by active materials on the anode being forced off, and into the electrolyte, contaminating it. You may seem to bring the cells back by this high current charge, but you now have a cell that has been severely weakened compared to the others.

Typically the discharge knee is reached at the 20% capacity level (assuming normally calculated solar battery power applications designed with enough capacity to begin with, and not an EV accellerator pedal application drawing HUGE amounts of amperage compared to pack size)

In other words, this need for a gentle recharge when below 20% capacity is why I de-rated my cell capacity right off the bat. I didn't plan to go below that in the first place, AND MOST IMPORTANTLY, I have no BMS that is smart enough to do that. It will just hammer the battery if given the chance with my normal current setting. Not good when cells are below 3.2v at rest.

The other consideration is that in order to do that, that weak recharge current if you draw well down into the knee means you are eating into your solar charge time for normal currents very badly! Those low currents do take time to get back up to the minimal nominal voltage.

Prove it to yourself:
Tecmate / Optimate lithium chargers for motorcycle LFP batts have been doing this safely for years. So too these days are many of the hobbiest chargers, that even if you try to pump say 0.5C into them at first, they automatically drop it immediately to very little charge current until a nominal of 3.2v is reached, and then continue on with your original current setting.

Grab some solar flashlight LFP cells, or even seek out some better quality larger equivalents like 18500 or 18650 sized ones (NOT 3.7v nominals!). Bottom balance them. Hit them up with 0.5C and watch the voltage shoot up, but inside spin their tires and seem to never take a charge. Let them rest for an hour or two, hit them up with very very little current until 3.2v is reached, and then try again - TADA. Cell is back in business, although you did stress it out.

Or let it spin its tires when you hit up one that is well down into the knee with high current - at least no larger than the rating spec. Watch the voltage shoot up, and the current get pumped in, but reach a point where it doesn't seem to be reaching much closer to the CV point you set. It never seems to get there. Pull it off charge, and within 15 minutes, the cell voltage has dropped back down badly! It just spun it's wheels doing a burnout going nowhere internally.

General note - LFP abused badly seems to just go dead with no fire. The iron phosphate is greedy and holds on to oxygen. Thus you may swell, maybe stink the place up if it vents. 3.7V NON-LFP batts need the same recharge precaution - but here they tend to vent with flame. Big difference. Again, the reason you can put LFP motorcycle batteries under your butt, and not Tesla batteries 3 inches away from your cheeks. Know the difference.

An Opus hobby charger with the internal switch set for LFP will let you play / test like this. I heartily recommend those interested to do this on a small scale before investing in a big real-world battery system.

So I'm just saying - when bottom balancing, don't hammer your pack until a nominal 3.2v per cell has been reached with a gentle current of 0.05C or less if you want to get the most out of it.

Not sure if Rickard ever talked about that, but I've seen plenty of EV'ers hammer their low charge cells with normal charge current and wondered why they failed or needed constant rebalance babysitting afterwards.

**Ampster** · 02-04-2019, 05:46 PM

Originally posted by PNjunction

Re "mid-balancing" by merely paralleling cells together. No such thing. Choose top or bottom. The reason is because of the very flat charge / discharge curve of LFP. Mid-balancing is a total oxymoron. One will find out soon enough when they try to fully charge, or fully discharge the pack.

.............

Perhaps I should have said equalize the voltage. That is what I do before assembling a pack. The term middle balance came from another thread where Sunking was trying to describe what EV manufacturers do with their BMSs. He was trying to explain that one cannot truly call what the manufacturers do as "top balancing", because the cells are not actually at the top.

**zamboni** · 02-06-2019, 09:50 PM

My single personal experience with mid-voltage "balancing" lines up with PNJs comments above. Upon purchase, I wired my cells in parallel and let them "balance" for >24 hrs. Monitoring the resulting pack in actual operation, it was way out of balance at both my "full" and "empty" voltages.

This is the first Ive heard of being gentle with recharge current on LFP cells below 3.2VPC. It sounds a lot like what Ive seen for charging cells below freezing. Same underlying reason?

Lucky for my solar application, my array can put out maybe 0.1C at STC...so in many circumstances my system just cant provide excessive charge current. I will have to watch out when trying to refill the battery pack after a deep discharge, though. I will probably set my shore-power charger (which only really gets used in such circumstances, and rarely) current limit to 0.05C.

Is this 2009 info still accurate?

Is this 2009 info still accurate?

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment