Lithium battery individual cell voltage balance

[booster]

I think it might be interesting to have those users that can see the individual cell voltages in their lithium banks check them and post what kind of cell to cell variation and also how often and how their cells are balanced (if they are ever balanced). Our now a bit over two years in place 618ah SOK drop ins have probably 10-12 weeks of camping use with numerous day trips also over that time. Storage is inside the shop and temp never goes below about 42* out there as it does have heat. When I am working out there it is 65* so no lithium temp or heat issues. This last winter I was laid up with shoulder surgery for months so I put the system on the shore charger with full cutoff at about 85% SOC and rebulk at 40%. It went unattended for about 4 months and charged a couple of times as the batteries were not isolated and were running the monitor and detectors.

After our recently taken fall trip that we never plugged in and only charged by alternator a couple of times for about 20 minutes each while driving, but with the solar on to contribute I checked the cell balance after sitting overnight inside with no charging or solar. It was sunny and a part of the drive home so got home at about 65% SOC. The cell balance was within .003v total range for the 12 cells. These cells have not been balanced in over two years so that totally surprises me. It is contrary to what the trend seems to be in the manufacturer's recommendations for charging and balancing, which most seem to be at 14.4v-14.6v full charge and auto balance on every recharge. Those recommendations are probably in response to early failures or complaints of the batteries not being able to get to full charge voltage without tripping out the high voltage safety disconnect. The voltage is any cell high so bad balance will cause that trip. Our SOK batteries use that same 14.4-14.6v and auto balance every charge setup, but we midrange charge so never get anywhere near 14.4v, so no balancing is done.

I had even initially bought 3 autobalancer boards to add to ours because they could be set for balancing voltage, but never needed to install them. It is interesting that the spec on them is that they won't balance until they .020-.030v inbalance and we have never gotten there. One thing I do see if I watch the balance during big discharges or charges (100+ amps) is that the cells will show up to about .030 inbalance, but the overall battery voltage between the 3 batteries is constant. The balance very quickly evens out once the high current is off them. If they had active, always available, balancers on them they would probably balance under charge and discharge high current situations, but I think that would make the resting balance worse most likely.

For those that can see their cell balance, what are you seeing over time with them? Doesn't matter if it is a low cost drop in setup or a megabuck system, all data is good to have for us and others to be able to reference in the future.

 

[FredA-ClassB]

There is a common misconception that Li cell balance can be measured at partial SOC. The cell voltage is mostly dependent on the chemistry of the battery and the charge or discharge current. When the cells are being charged or discharged at low C rate (1C being the current that will fully charge the cell from 0 SOC in one hour) the cell voltage is not a measure of the state of charge. So, the voltage does not indicate the balance.

Li cells have a very flat charge/discharge voltage that only rises when the cell is almost charged and only falls when the cell is nearing bottom of charge.

The graph shows cell voltage during discharge after being charged at various voltages. You can see that as long as the charge voltage is above 3.4 volts for this cell the cells all read alike until they are almost fully discharged.

The balance circuits only start to work when the cells are above a voltage that would indicate near top of charge. So, the only time one can check cell balance is during the absorption cycle of the charger. Before someone tells me that Li cell doesn’t have an absorption cycle the terminology of the states of charging that were used for Lead Acid cells is still relevant. Most, if not all, chargers have a constant current stage called bulk charge. This is followed by a constant voltage/aborption stage during which the balance circuits in the BMS can balance the cells. The balance period is usually a set period that is established when the charger is programed. After the balance period times out the charger will drop the voltage to a float voltage that is set below the absorption voltage.

It is only during the end of the bulk and during the absorption cycle that cell balance will be meaningful.

 

[booster]

What you are describing is top balancing and is what those that build there own batteries would do on the individual cells before putting the battery. Charge cells to full voltage for a single cell and near zero current and then let them rest. Reduce the charge in cells that go higher. It is also what the majority of purchased 12 batteries do in use.

But balancing can be done at most any voltage, even at 0% charge, called bottom balancing.

I don't know if the still do it, but a few years ago I saw active balancers that were always evening out the cells when resting, charging or discharging.

For quite a while the claim was you top balance and then never need it again, but that proved to not be true in real world.

We do midrange charging to 80% so never hit the balancing point in use. About once a year I take the batteries up to just short of balancing voltage and let them rest and they are always in balance the same as at 80%. I have to do this to reset 100% on the Victron battery monitor do to charge efficiency mess up accuracy over time.

Ours are 2.5 years old and 3x12v batteries for 618ah and have never needed rebalancing. The cell to cell voltage is .003 millivolts max over the 12 cells. Millivolt readings are exactly how all balancers know if the cells need balancing. It is true that voltage is not great for checking SOC because of the flat curve but that doesn't really matter for balancing IMO because you are measuring variation between them.

I have documented this elsewhere in discussions, but what I found in setup was the with parallel batteries like we have, the charge/discharge amperage matching is critical. I used equal length cables, but the resistance variations, though very small, still messed up charge amps. Even charging which cable is over another on a post makes a difference. I finally added .015" stainless washers as needed to the battery connections to balance the current. It took anywhere from zero the two washers to get the current to even out. That may be why our batteries have stayed in balance so well.

 

[@Michael]

My battery has an active balancer, so the cells are always really close. Right now, they're within .01 volt.

 

[booster]

My battery has an active balancer, so the cells are always really close. Right now, they're within .01 volt.

Well within the .020v that is commonly stated as needing balancing. Hold old is the system and what voltage are you charging to?

Is that checked rested or being charger or discharged? Some digging for info on that seems to indicate no clearcut answer to that question.

 

[@Michael]

The battery is three years old. Right now it is fully charged, floating at 13.41 volts. The Bluetooth app only reports two decimal places, so there likely is a rounding error somewhere.

I left it between 60 & 80% for a few months, then fully charged it a few days ago. If there was any drift it was patched up by the balancer.

 

[booster]

The battery is three years old. Right now it is fully charged, floating at 13.41 volts. The Bluetooth app only reports two decimal places, so there likely is a rounding error somewhere.

I left it between 60 & 80% for a few months, then fully charged it a few days ago. If there was any drift it was patched up by the balancer.

Is that balance with batteries floating or resting? If floating can you see tha charge amps accurately?

If you are referring to drift in accuracy of an addon battery monitor do to charge efficiency, which always happens to some degree even with lithium especially if there is an always active even when asleep bluetooth BMS in them (ours are that way), the problem is anything that happens within the battery like balancing or parasitic within the cells plus basic charge efficiency aren't seen by the monitor as losses. If you are using the SOC from the BMS, which ours does provide, it may or may not see the balancing effects but still won't see pure charge efficiency losses. It is a bummer that you don't have the third decimal voltage measurement, not only from the rounding standpoint, but good 5 digit meters that actually read to .001v also have greatly improved accuracy and repeatability. Be aware of the current influx of low cost 5 digit meters that don't have improved accuracy and the last digit is only .5 or zero as they aren't very good. Downside of good meters is cost and the fact that they are slower to display the reading as they have to sample more times, I think.

Calibrating the monitor isn't all that difficult as long as it has manual reset to 100% or can reliably determine full with lithium (or any other repeatable state of charge). You just charge to full and reset the monitor to 100%. Will be doing that this year as I need to find out the voltage at the 80% point so I can get the new shore charger to charge to that point. I think it is currently set at 13.77v and 3 minute hold but I don't know if that is totally correct as I had to go by the uncalibrated monitor as the Magnum was failed to find that point and time was short this spring. We normally do 35-80% in use and want it all automatically done if possible. The new charger sets in .02v increments and 1 minute absorption to shutoff increments so should be much easier than the failed Magnum that was .1v increments and 5 minute absorption time intervals. The Victron monitor allows me to set the SOC to anything I want so works well for calibrating. All I need to do is a shore charger charge to the setpoints that give 80% and then set the monitor the same. With the Magnum I found the drift was about 5*-10% SOC per year. I always also check the cell balance with the batteries resting after the shore charge and setting of the monitor, but they have always been within .003v. I have ability to remove the battery covers which I did when doing intitial top balancing and at that time I also checked the accuracy of the cell voltages from the BMS against my couple of year old Bryman 5 digit meter. I found the BMS readings to be quite good and with .002v of the Bryman and consistent between the batteries themselves. I was pretty surprised they were that close, actually.

When we travel we almost never plug in so the second alternator does most of the charging then. The alternator is controlled for on/off by the Victron monitor so having it calibrated is important for us. We also use the 300 watts of solar and turn it on when we hit 45-50% SOC. It does not have full cutoff charging and we don't use floating on any source, but it takes at least 10 days for it to gain enough in good sun to get to 80%. We just shut it off if we get close. The solar charges at 13.4v so is OK to leave all the time if we wanted to but like to stay totally midrange.

I think there are still a very lot of things we need to learn about lithium batteries as the "recommended) (but marketing influenced) specs from the manufacturers continue to change at a pretty rapid rate. What they say now is a very large amount more conservative than they recommend 5 years ago. Who knows where they will be next year.

 

[@Michael]

Is that balance with batteries floating or resting? If floating can you see tha charge amps accurately?

If you are referring to drift in accuracy of an addon battery monitor do to charge efficiency,

Floating.

By drift, I meant the difference in individual cell voltages. I'm assuming that with an active balancer, the cells will always be in balance?

 

[booster]

Floating.

By drift, I meant the difference in individual cell voltages. I'm assuming that with an active balancer, the cells will always be in balance?

Interesting idea and may be true if the balancers are set for always on including when charging and discharging, which is, I think, pretty common for active balancers. On float, if they are on I would expect that would be the case, probably within the .020v range the balancers normally allow. I am no sure what the active ones will do once you disconnect from shore power, and that would be interesting to know, I think. Do they continue to monitor and balance at that point of do they shut off completely until in use again.

What I have found with our setup, which are passive and only were used at initial setup to test what they did. I found that if they balanced at the 14.4-14.6v balancing voltage they would be not as well balanced when resting afterward with not input or output current. I also found that they would go out of balance by some amount based on charge or discharge current, usually around .030v max and that imbalance went away as soon as they were resting. Since our batteries rest probably 95% of their life when stored, I am OK with that. There are a lot of stuff I think we don't yet fully know about lithium systems, but it is interesting for sure. The same was true in the learning curve for AGM in past so not surprising.

 

[booster]

I am currently doing a monitor calibrating charge on our 618ah bank. l do this as needed to be able to get the Victron monitor so it is accurate at 80%. This time is a bit unusual as I need to take them near full (our considered 100% and probably about 600ah) so I can discharge and get to 80% by taking out 120ah with a load. I will then set the monitor to 80% with the Victron app, shut down everything, and read the resting voltage after a couple of hours. I will then run them down further and recharge to see what the current settings in the new charger give us for SOC. The goal is to get the charger settings so that it takes us to 80% directly so monitor calibrations are easier.

I just looked at the cell to cell balance on the 3 batteries and none of them are running more than .010v total range variation. They have not been balanced in over 2 years and this calibrating charge stops well short of balancing voltage.

The shore charger is almost only used for the calibration charging as we never plug in unless we need AC, so it is set at only 40a (out of 130a max). I do this for voltage accuracy as it does not have remote voltage sense.

We do use the inverter a lot, though, and it does a much better job than the Magnum ever did.

 

[booster]

I just finished the calibration tests I needed to determine the right settings for the new inverter/charger shore charging. A mentioned, I am at very low current because of no remote voltage sensing and also because we never really need shore charging to be faster. I needed to know what voltage and absorption time (no float) to use to put us at 80% SOC. The charge rate is at 40 amps, .06C, which is really low and I started at 50% SOC that I determined by taking the batteries up to about 14v and holding a bit to get 100% SOC. I then discharged the batteries by 50%, 309ah to get the starting point. I settled on using 13.72v (the batteries showed less than that by .14v consistently), 1 minute absorption time after a bit of testing that way, and the charger went off at 80.3%. Charge efficiency was set at 97% on both the charger and monitor which if anything would have given slightly high actual SOC in this test. I chose that efficiency knowing it would not be the best choice for this charging cycle because the batteries had sat off for only 15 hours so the BMS parasitic losses that don't show on the monitor would be very low. In our real world the batteries would sit much longer than than that, probably 30 days or more in storage so the non seen by the monitor would be much larger. The 97% is a guess based on our past use and watching drift vs actual over the last 3.5 years.

I did monitor the cell balance during the charging period, but should have done the 75 amp discharge also but didn't. At that low charge rate the balance between all 12 cells was within .010 volts and it went back to .002v as soon as the charger stopped. Of even more interest to me was that just off charge the batteries read 13.37v and only dropped to 13.35v overnight. I think the indicates that there was little to no excess ions on the cathode to cause lithium plating, at least I hope it does.

It is a bit of hurry up and wait type work with long charge and discharge times but it will be a very large benefit going forward because we can recalibrate the monitor to 80% whenever we want just by running a shore charge cycle (with no loads on if possible as the charger is not shunt based and it wouldn't know charging current from load current). I expect to just run a charge before we leave on a trip and be good to go for SOC accuracy. The engine charging controls off the monitor so it will always shut off at the 80% it reads. Solar is set to only 13.4v and we don't turn it on until we are in the 50-60% SOC range so it has little chance to go above our 80% target. We have done that with the solar for the last year and it worked well, and never plugged in on trips except one night on the way home from Custer at Lake Mitchell campground in South Dakota. We had over 300 miles of freeway driving at over 100* to get there and it was still 99* when we stopped with lots of heat stored in the van, so we did plug in there.

 

[Winston-ClassB]

Booster, we've been following your comments with interest in this thread and elsewhere on 'balancing' and wonder if we're talking about the same thing. Your concerns seem to focus on whether your 3 SOK batteries are at the same charge level. Indeed, if we understand your comments, you've gone so far as to add "shims" to assure that each of your three 'banks' are receiving identical charge currents (and presumably are equally splitting their respective discharge cycles/currents). But why is this important? Do we really care that two of your 3 SOKs are being charged at, say, 33 amps, while the third is receiving a disproportionate 34 amps of charging? By "shorting" your three SOK banks in parallel, you are assuring that these banks will be equally charged and balanced. Yes, maybe for those instants when you've just completed your charge cycle, Bank 3 will have slightly more charge than the other two (since we sent more amp-hours into Bank 3), but when the dust settles - - which means when all current between your three SOK banks has stopped flowing - - and it will, it must - - the three SOK banks will be resting at precisely the same voltage. We content that by definition, this marks perfect balance.

For us, balancing has nothing to do with the differing states of charge of parallel connected (shorted) battery banks; rather, it refers only to the balance between each of the four series cells comprising, individually, each of those parallel banks - - in your case, your 3 SOK batteries.

Interested to hear your thoughts.

 

[booster]

Hi Winston, I was hoping we would here from you as you have been messing with this stuff way longer than I have.

When I was doing the initial setup even though I had as close as I could measure and get on cable lengths I had some big differences in charging current to the batteries and also the ones that got the higher amount of current were generating larger within battery cell to cell voltage variance. I assume this all was because that although the cables were well matched the batteries themselves varied in impedance.

It did take a while for the 3 batteries to settle into the same voltage at rest but not more than a few hours, but this is with a 5 digit meter so pretty picky.

What surprised me was that the cell to cell balance in the batteries stayed larger in the batteries that had been taking the most current and did not age back to even within those batteries when looking at cell to cell balance in each battery. It got better but was always higher in the the higher current batteries. I had carefully top balanced all the cells to the same voltage and current with a regulated power supply before installation. IIRC the worst cells were about .03v variation within the batteries that charged the fastest and had the highest voltage right off of charge and presumably higher SOC.

At that point I messed with the cables a bit cleaning and tightening stuff and looked for higher temps with an infrared and found almost no issues so I pulled the batteries and top balanced them again.

They showed the same way after putting them back in so I started messing with the shims to even out the the current to the batteries as well as I could with a clamp on meter and very poor ammeters in the BMSs in the batteries. The amp split got much smaller and varied with charge rate, but not all that bad at any speed up to the .2C max charge rate I use. The cell to cell balance also got better within the batteries at closer to .015v when charging and within under .008v or so after resting. Well within the .020v that would trigger the active balancers that I had on hand but not installed, so I didn't put them on.

Since the batteries were new and unknown, plus it was pandemic time, I ran a capacity test at that time. To do that I had tot take them to the 14.4v and 15 minute hold time the manufacturer told me defined 100%. Of course that also triggered their internal passive balancers in the BMSs, they got balanced and cut cell to cell within the battery unbalance to under .005v.

I ran the capacity test and recharged seeing the same charge amp variances between batteries and similar on charge within battery cell to cell imbalance. I again ran them to the 14.4v to try to determine what the charge efficiency should be set at and again they balanced and got closer to .003v within battery and between batteries.

At that point I discharged them to 80% SOC per the Victron monitor amp hour counter and shut off all loads and chargers to get the resting voltage, which I used to set the shore charger so I could use it to go to true 80% as our top charge reference in use. The Magnum only allowed .1v increments if charge voltage and 5 minute increments in absorption time to off so could not hit it right on as it very picky in midrange like this.

I connected the internal Victron monitor switch to control the .2C alternator charging between 40 and 80%.

We then started using them in the real world and found the within battery was actually getting better with more charges and under .003v within battery and between batteries at rest and light loads. At 120 amps we see some higher cell to cell but usually under .015v within battery max and tapers as we approach 80%. It returns to under .003v almost immediately after charge is removed.

It has now been 2.5 years and they have never been balanced since the first setup and did OK even with Magnum messing up voltage control when it was failing. The are still super close both within battery and between battery.

I don't know why they are staying so well balanced, but I will take it! The reason this thread was started was to see if what I was seeing was abnormal when compared to what others have seen over time with their systems and other charging methods. There doesn't seem to be much information out there related to the topic.

As to the premise of the vaiations in amps between batteries not mattering much, I think this probably mostly true as they will self balance between themselves when taken off charge. From what we have heard and makes sense is the being more out of balance within battery can make the battery trip out on high cell voltage limit when they are taken to full charge 14.4-14.6v like the manufacturers recommend. Ours have the high limit set at 13.7v so only .100 variation when charging could do that. Most would then run a balance cycle and restart. If the balance does get good enough they will do it all over and over until they do balance. At least that is how I think they are doing it based on reports of batteries in parallel tripping out at various times and coming back on then another battery might do the trip out as it gets more full but hasn't yet balanced. By then they are accepting pretty low current so it could take battery long time to catch up and the charger may have shut off by then if it is running to shutoff or float on time. That way the slow battery would never get balanced. A 20 amp slower battery over a 4 hour charge time could be 80ah of charge short and not balanced at shutoff depending on settings, I think. It may be a factor in those that have one battery in a parallel fail early, but may not be and I have no idea if the faster or slower battery would be most likely to fail.

I think a lot of the manufacturers of the drop ins pretty much agree these days that having good balance if very important to capacity maintaining and battery life. Most have gone to much low charger rates than in the early days, big difference between 2-3C and .2-.4C although some are still saying .5-1C are OK. Marketing has probably influenced the recommended charge voltage recommendations as they want to not loose the capacity/cost thing they need to sell them. At least it lets them do a top balance every cycle so that failure point is reduced.

I am sure our system will continue to spring surprises on me over time, but that is also a major reason I didn't just replace with AGMs when our old Lifelines were nearing age out. The learning and trying to understand the whole thing has been very interesting to me, and they are performing better than I anticipated so all good.

Although they are different chemistry, I was surprised when we got out Samsung S24 phones that they specifically recommended using a low charge rate charger if practical and even have a built in automatic setting to end charging at 80%. They claim potential doubling of battery life if both are used. Big change from past high rate and full charging recommendations in most past phones. Electric vehicle builders also struggle with charge rate and capacity recommendations for obvious reasons. If they don't have big capacity and charge fast, they won't sell. Interesting that they do seem to be using LiFePo4 batteries more now, probably for extended life and safety.

A double major Chemical and Electrical engineering degree would sure come in handy sometime, though.

 

[Winston-ClassB]

Booster, we're traveling now (western Oklahoma) heading to Albuquerque (balloons) so a bit limited timewise for a full response. We're assuming when you did your top balances that you opened up the SOK's and individually charged each cell? Also, what voltage did you finally 'discover' represented 80% SoC? You may remember in the early days of our pack, we determined that 13.36 volts represented 90% SoC but on a couple of measurements - - that same voltage also represented 80% SoC. It surprised us that the lithium discharge curve (for new batteries) in the region of 80-90% is so flat that we couldn't detect a difference on our 2 decimal (10mV) resolution Fluke.

We have a slightly different philosophy on using ‘battery monitors’ to regulate or control charging. We have a Victron 712 in addition to the monitor built into our originally installed Elite Power Systems BMS. Both can terminate ‘excessive’ charging and shut-down the system if excessive discharge is detected. But we view these as “fail-safe” devices that should never be used, that is, should never trigger. We chose, instead, to program and use our three charging sources (shore power, solar, and 2nd alternator) to control the charging function and only if one of these sources fails - - “runs away” - - will the battery monitor(s) intervene.

But this is easier for us as we have yet to understand how placing, for example, a 13.36 constant voltage power supply (charger) continuously across a lithium pack charged to an SoC corresponding to that same 13.36 volt level can harm the lithium pack. Yes, we remember that you have a “white paper” that argues that it does ‘hurt’ the battery but our engineering and intuition suggest otherwise. Consider your system where you charge to 80% then pull the charger. Every load now has to be powered by the battery. As you commented, you are discharging you battery down to 40% before the charger is again enabled. In our system, we add a load - - it is, effectively, the power supply (charger) that supplies the current. Our battery is not constantly, and we contend unnecessarily, being cycled. So, is it really better to repeatedly cycle a lithium battery through 40% of its range, or to let the battery sit there, essentially idle, at 80% SoC? But, that said, we are not unaware that our (9 year old) batteries are showing serious age issues while your, albeit much newer, pack is performing very well.

 

[booster]

Thanks Winston.

Yes, the SOK batteries have removable covers so easy to do. I used a 10 amp regulated power supply set at 3.600v and 7 amps and stopped the charge when the got to 1 amp.

At 80% ours were 13.37v when right off the charge at 13.72v and 40 amps with 1 minute absorption time. They only fell back to 13.35v overnight, so very similar to what you saw and hopefully very repeatable with Samsung which show very stable outputs of voltage and amps.

It is an excellent question and I think that so far that there is no real well tested proof of what effects you might see from continuous floating at your chosen SOC point. I think that white paper may have mentioned plating issues, but not certain that is where I saw that claim made and I think they floated higher than terminal resting voltage. I also think they may have been held at totally full at that test and that is bad even without float. The more deep into the plating issue papers I have seen pretty much all of it excess ion buildup on the anode causing a voltage difference across the anode to media bridge. My question then and now is if there is still a voltage potential across that bridge that could be enough to cause problems with the float and resting at the same voltage. I have no clue and have no evidence either way. One thing the complicates it for my understanding is the battery is charged to 13.35v and 80% and I put 13.35v on it all time after that, my understanding is the might continue increasing capacity. I have found some opinions and no facts about that would 13.4v or 13.5v like many of the systems float at charge them a lot further or cause damage because to are above resting voltage? This also relates to question of why the recommendations to store them at 50% SOC and not higher? More wear and tear if higher? I have hedged us at this point with the solar which is only 300 watts. It normally will keep up with daily use or gain an amount each day, but also will lose some if very cloudy raining like we had for long periods on our last trip. I don't turn on the panels until we are at about 50% SOC and it can take weeks to get up to 80% so is not holding the 13.4v almost all the time and only runs during the day so maybe better than on shore float, and maybe not.

I hear you on the standalone for all 3 sources. Ours are also. The Victron only controls the alternator charging and it is because the Wakespeed can't control accurately enough to it accurately. The Wakespeed does very well on voltage and amperage max, but not good for ending points. We normally don't get near 40% and I have an override switch on the dash to turn it on manually if I want to just bump it up a bit.

The solar and shore power are also complete standalones and control themselves.

It will be interesting to see what your capacity is this year. I don't recall if ARV called theirs bad if lower the 70 or 80% and if they always replaced all the cells which they probably should at their prices.

I am hoping the resting voltage after the now much more accurate 80% charge will give me early warning on capacity loss, but that is hopefully and long ways away based on what you have seen on yours.

 

[Winston-ClassB]

As an aside, please comment a bit more on your new Samsung. Is it both a charger and inverter with all of the same auto-switching (and neutral bonding/unbonding features) of the Magnum? We ask, of course, because - - not only have you given us 5 digit voltmeter envy - - but you've planted the seed of worry as we have a Magnum 2812 and are wondering if our days are numbered?!

Now, returning to lithium . . . we continue to argue that the terms of Bulk, Absorb, and Float are arcane, unnecessary and meaningless in the world of lithium. And while we understand that others on this Forum do not concur, please accept that we find these terms to be ambiguous and unhelpful and wish to utilize, instead, “engineering” terms that we do understand, for example, voltage, current, state of charge etc, to define a set of rules for charging, operating and maintaining lithium battery systems.

We were surprised and delighted by this thread as it seemed we were both on the same track . . . to at least start the process by defining what the resting voltage of a lithium pack is at a ‘meaningful’ percentage of charge. You selected 80%. That’s arbitrary, but we agree that, in view of the currently accepted wisdom that maintaining SoC at higher levels can stress lithium cells causing damage and some loss of cell longevity, an SoC well below 100% is a sound choice. Indeed, the experts are telling us that we should store our cells at 50% or less of capacity to minimize cell deterioration. But you didn’t pick that low percentage because such a number would be giving-up too much operating capacity simply to extend the life of our cells an uncertain and speculative duration. So you compromised and selected, we think, a very reasonable 80% SoC target.

But then you lost us. You started talking about the voltage “right off the charge”. You said something about 40 amps and 1 minute absorption time . . . huh? There’s that **** term, absorption. What did Booster mean? Booster, we started together, we both agreed to fully charge our lithium packs (assuming we know when a pack is, in fact, fully charged). We agreed . . . discharge the pack a known 20% amount. STOP there. Wait 10 hours with no further discharging or charging. At that point you can be assured that the pack has come to a “rest”. It’s that voltage we think we both want to know. That voltage is, per your “overnight” resting, 13.35 volts.

We agree that we want our packs to be at 80% SoC or 13.35 volts. So how do we get them there? For us the answer is simple. We take a power supply or charger (they’re really the same thing in this context) and set it to a Constant Voltage of 13.35 volts and connect it to our lithium pack.

Unfortunately, there really isn’t such a thing as a true Constant Voltage source because, were we really to connect a 13.35 voltage source to a discharged lithium pack, we’d have an explosion in amperes . . . maybe thousands of amps. Fortunately modern power supplies simply roll-back that thousand amp expectation to a lower number - - a number corresponding to whatever the power supply can provide and the power supply outputs this lower current until the terminal voltage of the lithium pack nears the target 13.35 voltage. So, in reality, our so-called Constant Voltage is really a Constant (Maximum) Current-then to-Constant Voltage profile. For us it’s just one step, connect a power supply of voltage 13.35 to the battery. Is this Bulk? Or a combination of Bulk and Absorb? Or none of the above?

Now things get dicey. In Winston’s world, we just leave things alone. The 13.35 volt power supply stays connected to the system forever. When low power loads are connected, it is the power supply that is essentially supplying the current to the load. The battery is not being discharged. If we light-up our induction stove on “high” - - exceeding the capabilities of our power supply (charger), yes, the battery steps in to assist . . . however, when the load is removed, whatever charge was supplied by the battery will be returned by the power supply to the battery in short order. Most of the high current demand was supplied by the power supply, the battery providing the necessary auxiliary boost.

But in Booster’s world? You want to turn-off the power supply (charger) at 80% SoC and then use your lithium pack to power all your loads? Why? Because some paper suggested that you’re getting anode “plating” and moving “ions”. Again, we cannot see how a lithium battery, truly resting at 13.35 volts will even know that there’s a power supply (charger) connected if that charger is set to the same identical potential. There will be no potential difference, there will be no current, there will be no moving electrons . . . indeed, we’re uncertain about your “ions”, but what’s going to cause them to move, where’s the force to cause them to move? But we think even you are uncertain - - maybe, as you opined, these “white papers” are really addressing the consequences of maintaining too high an SoC not whether one can connect a power supply in parallel with their battery, continuously.

In any event, as you suggested, to answer these questions may require expertise in both Electrical and Chemical disciplines - - the latter we lack. But for the present, we’ll stick to our simple Lithium Charge Profile - - turn-on the power supply (charger) set to the voltage that corresponds to the desired SoC and leave it there.

 

[booster]

As an aside, please comment a bit more on your new Samsung. Is it both a charger and inverter with all of the same auto-switching (and neutral bonding/unbonding features) of the Magnum? We ask, of course, because - - not only have you given us 5 digit voltmeter envy - - but you've planted the seed of worry as we have a Magnum 2812 and are wondering if our days are numbered?!

Now, returning to lithium . . . we continue to argue that the terms of Bulk, Absorb, and Float are arcane, unnecessary and meaningless in the world of lithium. And while we understand that others on this Forum do not concur, please accept that we find these terms to be ambiguous and unhelpful and wish to utilize, instead, “engineering” terms that we do understand, for example, voltage, current, state of charge etc, to define a set of rules for charging, operating and maintaining lithium battery systems.

We were surprised and delighted by this thread as it seemed we were both on the same track . . . to at least start the process by defining what the resting voltage of a lithium pack is at a ‘meaningful’ percentage of charge. You selected 80%. That’s arbitrary, but we agree that, in view of the currently accepted wisdom that maintaining SoC at higher levels can stress lithium cells causing damage and some loss of cell longevity, an SoC well below 100% is a sound choice. Indeed, the experts are telling us that we should store our cells at 50% or less of capacity to minimize cell deterioration. But you didn’t pick that low percentage because such a number would be giving-up too much operating capacity simply to extend the life of our cells an uncertain and speculative duration. So you compromised and selected, we think, a very reasonable 80% SoC target.

But then you lost us. You started talking about the voltage “right off the charge”. You said something about 40 amps and 1 minute absorption time . . . huh? There’s that **** term, absorption. What did Booster mean? Booster, we started together, we both agreed to fully charge our lithium packs (assuming we know when a pack is, in fact, fully charged). We agreed . . . discharge the pack a known 20% amount. STOP there. Wait 10 hours with no further discharging or charging. At that point you can be assured that the pack has come to a “rest”. It’s that voltage we think we both want to know. That voltage is, per your “overnight” resting, 13.35 volts.

We agree that we want our packs to be at 80% SoC or 13.35 volts. So how do we get them there? For us the answer is simple. We take a power supply or charger (they’re really the same thing in this context) and set it to a Constant Voltage of 13.35 volts and connect it to our lithium pack.

Unfortunately, there really isn’t such a thing as a true Constant Voltage source because, were we really to connect a 13.35 voltage source to a discharged lithium pack, we’d have an explosion in amperes . . . maybe thousands of amps. Fortunately modern power supplies simply roll-back that thousand amp expectation to a lower number - - a number corresponding to whatever the power supply can provide and the power supply outputs this lower current until the terminal voltage of the lithium pack nears the target 13.35 voltage. So, in reality, our so-called Constant Voltage is really a Constant (Maximum) Current-then to-Constant Voltage profile. For us it’s just one step, connect a power supply of voltage 13.35 to the battery. Is this Bulk? Or a combination of Bulk and Absorb? Or none of the above?

Now things get dicey. In Winston’s world, we just leave things alone. The 13.35 volt power supply stays connected to the system forever. When low power loads are connected, it is the power supply that is essentially supplying the current to the load. The battery is not being discharged. If we light-up our induction stove on “high” - - exceeding the capabilities of our power supply (charger), yes, the battery steps in to assist . . . however, when the load is removed, whatever charge was supplied by the battery will be returned by the power supply to the battery in short order. Most of the high current demand was supplied by the power supply, the battery providing the necessary auxiliary boost.

But in Booster’s world? You want to turn-off the power supply (charger) at 80% SoC and then use your lithium pack to power all your loads? Why? Because some paper suggested that you’re getting anode “plating” and moving “ions”. Again, we cannot see how a lithium battery, truly resting at 13.35 volts will even know that there’s a power supply (charger) connected if that charger is set to the same identical potential. There will be no potential difference, there will be no current, there will be no moving electrons . . . indeed, we’re uncertain about your “ions”, but what’s going to cause them to move, where’s the force to cause them to move? But we think even you are uncertain - - maybe, as you opined, these “white papers” are really addressing the consequences of maintaining too high an SoC not whether one can connect a power supply in parallel with their battery, continuously.

In any event, as you suggested, to answer these questions may require expertise in both Electrical and Chemical disciplines - - the latter we lack. But for the present, we’ll stick to our simple Lithium Charge Profile - - turn-on the power supply (charger) set to the voltage that corresponds to the desired SoC and leave it there.in quote

The Samsung is basically a clone of the Magnum, it appears, but lacking two important features the Magnum had. First is no shunt based charging as no shunt on the internal measurements which aren't great. Second is no remote voltage sense which is pretty important for lithium, I think. It does have much better voltage control although not exactly the same as the batteries which are about 6' away. Very steady with no bouncing around like the Magnum always did. It will do tail amp charging, but not accurately because of the internal ammeter that also sees any loads. I think that only Outback may still have a shunt based system and of course a multipiece and control Victron system. We got a 3K inverter, 130 amp charger unit to replace the 2K and 100 amp Magnum. The Samsung certainly looks to be better engineering and quality and the inverter is much more efficient it appears.

I struggle with the calling of "stages" also, especially on the forum. Technically in the industry of chargers and batteries they call our method two stage charging. Constant current and constant voltage. Bulk and off or in standby to rebulk as needed as the constant voltage can be zero. Again, technically speaking we don't follow that description exactly, because I have a one minute hold at the bulk max voltage. That in lead acid would the absorption time but I would call it more a "confirmation" of truly having max bulk voltage. In the Samsung I program what is called bulk and the one minute is put into the absorption timer in the charger. Another setting takes care of the actual no float/stop with rebulk by calling out "lithium" battery in type. There are something like 4 different ways that setting allows you to chose. I found most on forum have no idea what I am talking about if I don't relate it to common stages they are used to seeing, even on chargers called "lithium" chargers or the lithium settings in other chargers that all use the old terms. Even the constant current/constant voltage is not understood. Very, very confusing to be sure.

The difference in leaving it on 13.35 all the time vs shutoff and use is one main question that I have as mentioned previously. The ions they speak of appear to be lithium ions with probably and extra electron or such (no chemist here). They are supposed to transfer something, maybe an electron to the electrolyte but it that electrolyte is depleted near the anode so they get trapped. The whole process is claimed to be having too high a charging current vs dispersion speed in the electrolyte under the given conditions. It is very geeky and I could be totally wrong, but that is what the descriptions of it sound to me. At a steady 13.35v on the charger you would always have as many electrons available as are needed but none can go into the solution do to the existing SOC. The big question to me has always been "what happens then". Is there a chance of plating? That is what they seem to be saying although they temper it with saying is much less at lower voltages than high, which makes sense as all plating works the way. If you charge to 13.35v and then put the batteries into discharge, even tiny, there will never be a surplus of electrons to do any "ion making" for lack of a better term, so no real amount of plating possible in theory. It is easy to see how if at high voltage plating would probably be more likely. I think it would very interesting to check a cell at constant 13.35v no charge and with a 13.35v charge on to see if there is any current flowing that is measurable to that cell. I probably should be referring to cell voltage though if on single cell so 3.3375v.

At this point, without a way to quantify which is actually better, you have a nine year life on your side saying it is not a bad way, if you have done this all along. Your have lasted longer than most others I have heard of. We never use the shore charger at all anyway so kind of mute for us and nice to have it off with the inverter sitting in very low power use standby, which is much lower drain the using either the search setting or on setting.

We off today on our annual SE Mn apple trip which is down the Wi side of the Mississippi to Winona and back on the Minnesota side with a stop at our favorite orchard. Fun and comfortable drive in the 96 Buick Roadmaster wagon with seats better than our recliners at home.

 

[Winston-ClassB]

By coincidence we had the opportunity on two occasions this summer to spend time along your portion of the Mississippi getting as far north as Merrick State Park (Wisconsin) - - not too far north of Winona - - down to one of our favorite campsites, Wyalusing State Park (again, on the Wisconsin side) overlooking Praire du Chien.

Our Magnum has similar capabilities to your Samsung: 2800 watt inverter and 125 ampere DC charger. We have never used the Magnum in any mode other than CC/CV (constant current/constant voltage). There are no settings other than to set the CV target voltage. Importantly, there are no ‘time’ settings as time is irrelevant when selecting the single CC/CV profile (2 stage CC and CV if you prefer that view).

We are uncertain what you mean by “shunt based charging”. We have shunts for both our Elite Power Systems and Victron monitors which, of course, record the number of “coulombs” being dumped by the Magnum into the batteries. But this ‘accounting function’ is completely independent of the Magnum so we don’t understand how the Magnum might use “shunt information”. In any event, as we do not use a shunt in connection with our Magnum, this Samsung limitation does not represent a problem.

Concerning your second point, no remote sensing for the Samsung . . . we’ve never used remote sensing on the Magnum as, in our case, the Magnum is quite close to the lithium pack somewhat negating any advantage for remote sensing.

By coincidence (lots of coincidences today), we were trouble-shooting an intermittancy of our 2nd alternator and discovered that our remote sense wire located near the rear-mounted batteries was, and has been for 9 years, shorted to the 2nd alternator B+ output. When we removed this “short” - - thereby enabling remote sensing for the first time, our alternator output jumped from 50 to 200+ amperes! Apparently when we designed our 2nd alternator regulator, we designed it to provide proper charge rates for local sensing. To solve this problem short-term (without redesigning our regulator), we simply added a switch allowing us to return the “short” (i.e. local sensing). The point being, in the final analysis . . . as charging nears completion with the corresponding charging current dropping to low value . . . the importance of remote sensing disappears as the “local” and “remote” sense voltages ultimately converge to the same value.

We noted your comment that you rarely connect to shore power. That changes everything as, without shore power, there is no option (particularly at night when parked) other than to let your batteries power your loads. By way of comparison, we keep our van’s electrical system permanently enabled . . . it’s operational 24 hours a day, every day of the year. Much of that time is when the van is parked at home. It is always plugged in. It is probably accurate to say that our Magnum charger is connected to our lithium pack 80% of the time.

 

[booster]

By coincidence we had the opportunity on two occasions this summer to spend time along your portion of the Mississippi getting as far north as Merrick State Park (Wisconsin) - - not too far north of Winona - - down to one of our favorite campsites, Wyalusing State Park (again, on the Wisconsin side) overlooking Praire du Chien.

Our Magnum has similar capabilities to your Samsung: 2800 watt inverter and 125 ampere DC charger. We have never used the Magnum in any mode other than CC/CV (constant current/constant voltage). There are no settings other than to set the CV target voltage. Importantly, there are no ‘time’ settings as time is irrelevant when selecting the single CC/CV profile (2 stage CC and CV if you prefer that view).

We are uncertain what you mean by “shunt based charging”. We have shunts for both our Elite Power Systems and Victron monitors which, of course, record the number of “coulombs” being dumped by the Magnum into the batteries. But this ‘accounting function’ is completely independent of the Magnum so we don’t understand how the Magnum might use “shunt information”. In any event, as we do not use a shunt in connection with our Magnum, this Samsung limitation does not represent a problem.

Concerning your second point, no remote sensing for the Samsung . . . we’ve never used remote sensing on the Magnum as, in our case, the Magnum is quite close to the lithium pack somewhat negating any advantage for remote sensing.

By coincidence (lots of coincidences today), we were trouble-shooting an intermittancy of our 2nd alternator and discovered that our remote sense wire located near the rear-mounted batteries was, and has been for 9 years, shorted to the 2nd alternator B+ output. When we removed this “short” - - thereby enabling remote sensing for the first time, our alternator output jumped from 50 to 200+ amperes! Apparently when we designed our 2nd alternator regulator, we designed it to provide proper charge rates for local sensing. To solve this problem short-term (without redesigning our regulator), we simply added a switch allowing us to return the “short” (i.e. local sensing). The point being, in the final analysis . . . as charging nears completion with the corresponding charging current dropping to low value . . . the importance of remote sensing disappears as the “local” and “remote” sense voltages ultimately converge to the same value.

We noted your comment that you rarely connect to shore power. That changes everything as, without shore power, there is no option (particularly at night when parked) other than to let your batteries power your loads. By way of comparison, we keep our van’s electrical system permanently enabled . . . it’s operational 24 hours a day, every day of the year. Much of that time is when the van is parked at home. It is always plugged in. It is probably accurate to say that our Magnum charger is connected to our lithium pack 80% of the time.

Interesting as always, Winston.

More coincidence is that early in our camping life, like 15 years ago, we stayed a couple of nights Wyalusing, and disliked it a lot for a couple of reasons.

The biggest was noise as there seemed to by parties going on outside until near midnight with loud music, shouting and bright floodlights. Not a ranger in site/

Second was the traffic. Maybe we hit a bad time of year, and not to stereotype, but there was constant traffic of very noisy diesel pickups going by, full of drunk, "big", men heading to the bathroom in groups that was half a block away. Very odd and nothing we have run into elsewhere in our travels.

It sounds like it must be different now to be a favorite place, I certainly hope so.

Our favorite in the area is Perot State Park in Tremeleau just north of LaCrosse. Great hiking there and that is our favorite activity when traveling.

I think a lot of the CC/CV chargers are like yours. Interesting is that they would not be able to balance on our SOK batteries as they want 15 minutes at 14.4+v to do the balancing.

If you have a shunt controlled charger you get different options to consider. It is really needed for lead acid so you can do accurate tail amp charging profiles, which we did for years with AGMs. If you are doing CC/CV charging you get a more precise voltage and amperage control than the internal controls usually give. That said the old design of the MS2000 lost a lot of the benefit do to overly large voltage, amp, and time settings. I had that control with the Samsung, I would be CC/CV and then stopping the charging as soon as the amps dropped off max amp setting. That drop should happen very soon after hitting the CV voltage if you have chosen the voltage correctly, I think. Having it shunt based also allows you control the daily charging be SOC if you prefer, like I do with our Victron monitor and engine charging. It can be very handy depending on personal preference.

The convergence of the battery and charger voltages has always been a saving grace on lead acid systems that like to go full capacity and very low amps at the end. (We used 2 amps at end for our 440ah AGM bank). We are not able to get that advantage with the midrange charging with any reasonable rate of charge. Our shore charger is set low at 40 amps, and when it disconnects at the voltage endpoint that gives up 80% SOC it is still pulling nearly 40 amps so we still have voltage drop. Remote sensing would help us a lot, I think, but it would take someone with high end skills to tap in and move the sensing off the internal boards to a remote site. Well beyond my skill and knowledge level. and assumes you can actually get a full board schematic for it, which is probably unlikely these days.

Good find on the regulator reference short. Amazing how many small things can mess up so many big things.

Yep, not plugging in does change a lot things. We knew that there would be some convenience benefits with never being plugged in, but in real use found they are even better than we anticipated. In and out of campground at will without a cord to disconnect, roll up, and store is very nice and we find ourselves going off to do stuff more often now. This has been a big help on those cold, rainy mornings trying to get on road early when I would be out a 5am in the dark and rain trying to wrangle a stiff muddy cable into a big garbage bag for instance. That would have happened on our just completed fall trip this year also. 43*, windy and raining and would have been out there at 5:30am. We literally got out bed, switched the interior things to travel mode and drove away without even leaving the van.

We can, and have, stored are van either way. Fully powered with charger on CC/CV shutoff and auto rebulk while plugged in lets everything be running full time. Our we can shut it all off and shut off the battery disconnect switch and pull the wired solar controller fuse to get very low use. Both ways work fine for us and what we do depends on what I have planned over the winter tweaking months.

 

[Winston-ClassB]

Your comments on Wyalusing are interesting as we’ve probably camped there at least a half-dozen times and never found such annoying conduct. Of course our infatuation with that campground is its ‘perch’ high overlooking the confluence of the Wisconsin and Mississippi Rivers . . . so one of those ‘front row sites’ along the ridge is a necessity. Otherwise, it’s just another campground. Thanks for the hint on Perot State Park - - always looking for good recommendations.

Not sure why a CC/CV charger would create any difficulty in balancing your lithium. In our case we set the Magnum to 14.2 volts (14.4 volts in your case) and leave it there until balance is achieved.

And with your Victron controlled system, we don’t see the relevance of “time” or “current” to the equation. Your Victron measures when 80% SoC has been achieved. Bingo, that’s it. At that point - - and regardless of whether you were charging at 2 amps or 50 amps - - your Victron terminates further charging. If you have no loads connected, your pack voltage will come to rest at 13.35 volts. That’s the advantage of using SoC rather than pack terminal voltage and/or ‘tail currents’ and/or ‘time’ as the ‘rule’ for discontinuing charging. You know exactly where your battery is - - and it’s where you want it. With voltage, tail currents, and time . . . you may get close, but you’re still guessing.

A short story on extension cords. We were tent campers and knew nothing of RV’s when we decided to build our DIY Promaster campervan. So what did we order? A 50 amp (220 volt) receptacle for the van with a corresponding 50 amp 50' long power cable. Ok, so we’re wired for 220 VAC and 50 amps . . . of course, in a small Class B . . . there’s no occasion to ever use such a connection. When the 50' 50amp cable arrived, we practically needed a folk-lift to move it about. It’s huge, inflexible, heavy, cumbersome, irrational . . . needless to say, we’ve never used it. Instead, we purchased a 50-to-15 amp adapter and a small 25' (16 gauge) 120 volt extension cord. Small, lightweight, easy to store . . . works just fine. Yes, if it’s raining, we do get wet. But overall - - as a hassle - - it really isn’t.

If you know anyone looking for 50' and 50 amps - - this power cord is brand-new . . . we’ll make ‘em a real deal.