91cavgt
New Member
I already have the EPEver set to charge up to 14.6 volts. It charges to 14.6, then drops to 14.1 float every day.
Huge lithium degradation
- Thread starter 91cavgt
- Start date
I already have the EPEver set to charge up to 14.6 volts. It charges to 14.6, then drops to 14.1 float every day.
>They are xiaoxiang model# JBD-TP04S006
I have used JDB BMS’s and they have done a good job. Do you have Bluetooth on those BMS’s?
If unsure there is an Overkill Solar app (for iPhones - don’t know about other phones), that will read the BMS. Mine came with the Bluetooth module that I just plugged into the board - maybe you could find/buy three of those modules?
I have used JDB BMS’s and they have done a good job. Do you have Bluetooth on those BMS’s?
If unsure there is an Overkill Solar app (for iPhones - don’t know about other phones), that will read the BMS. Mine came with the Bluetooth module that I just plugged into the board - maybe you could find/buy three of those modules?
Vigo
Solar Wizard
Well, unless you want to spend weeks on this im going to recommend you spend about $35-60 getting another charge source. A 20-amp adjustable voltage power supply can be had for ~$35, and a 50-amp for about $55.
Example: 50a for ~$55
I personally would not mess with a 1-5a charge source just because you potentially have 300ah worth of charging to do, so you'd be spending something from 60-300 hours on it if you were limited to that small of a charge source, and thats the MINIMUM because realistically the charge current is going to slow tremendously near the top. This is why i would hate for the 'bulk' part of the charge to be so slow, because you're going to be stuck with the last part being slow regardless. A higher charge rate would also help you better identify any failing cells via 'noticeable' heat buildup, and a 25p pack is going to take a fairly large amount of current to make that heat build up easily detectable without a thermal camera. Since you have no other way to compare cell IR without breaking down all the spot welds (nightmare, not worth it), just trying to get the cells hot becomes the easiest practical way to look for outliers.
Upside is, its basically set and forget. You can set a power supply to something in the 3.45-3.65v range, parallel all 12 of your 25p packs to it at once if you want to, and then just periodically check on it until the current drops almost to nothing. Once that happens, reassemble all the batteries as they were before and go 3 more years (is the theory). In fact, if you didnt do this sort of thing when you first built your batteries, it may go MORE than 3 years this time. If you can improve your BMS balancing function situation, it may go ' basically forever'.
I think you have to minimize the opportunity cost of this project by charging with a higher current source and making most of that process 'hands off'. Doing ANYTHING on a single cell level with 300 cells would take so long that a person with a minimum wage job would still be better off going to work and buying 3x new 100ah batteries.
Last edited:
Agreed, he may just not have the Bluetooth dongle.
OP can you share a picture of your BMS and ports?
Other than the JBD weak balancing current they are solid units and once configured to “balance when charging TO OFF”, they can bring a pack back to balance, especially when they are trickle charged 99% of the time.
If you have good cells, they are entirely unnecessary from a functional standpoint. This is why evs and commercially available ESS do not have them.
They create an unnecessary fault point for failure.
If you have bad cells and want to balance them, sure. That's the only application I see them working for someone. I still think it's better to charge the cells individually when building a battery though. That way you know they're all 100% SOC.
I put 60Ah and 40Ah cells in series five years ago and even at extremely low rates, I would still pull 40Ah. So even if you have an obscene cell voltage differential, the active balancer will not help anything. I think the most I could get was half an amp hour more. But that was with a really high amperage active balancer.
I was exploring this benefit when I was cycling grade b cells in large packs. The active balancer really didn't help anything.
I'm also making a video right now about how some of these BMSs cycle not with voltage but with state of charge determined by the shunt on the BMS.
@BarracudaBob
Last edited:
robbob2112
Doing more research, mosty harmless
I will look forward to the video
One other thing I have seen repeatedly suggested is to use the BMS balancer to do the top balance (i.e. skip doing it with a PSU) then turn off the balancer in software so you do know if a cell is having issues verse covering it up
91cavgt
New Member
I should have taken a picture of the BMS units while I had the pack apart. They do not have any ports on them. All that they’ve got is two large tabs that the negative leds are bolted to, and the little plug that has the wires that go to the cells. That’s it.
91cavgt
New Member
So if these BMS units are typically good, is it possible that the settings in the EPEver charge controller don’t ever put the BMS into a position to balance the cells?
Drunkin Solar
Solar Enthusiast
Think it's tough to tell what's causing the issue overall without being able to monitor/adjust the BMS. That's why the suggestion to replace them with new BMS. Generally the answer is yes they are good BMS but the limitaions of them in your setup are not producing the desired results.
Vigo
Solar Wizard
Sort of. You said this in your first post:
13.4v equates to 3.35v per 25p pack *if all 4 packs were actually equal* which may never have been the case if you never 'top balanced' the 25p packs to each other originally. At that total voltage, the individual packs may still be far apart in SOC. You need to get them charged enough to get into the 'upper knee' of the voltage/SOC curve to be able to reliably identify which 25p packs need balancing. If the batteries never 'rested' above 13.4v, you never got them charged enough to get into that region. This is my understanding.
If you look at something like this pic and then look at your variations between voltages of 25p packs that you posted previously, you can imagine how, with 13.4v total resting voltage and some amount of variation between how much each 25p pack is contributing to that 13.4v total, there could be a wide range of SOCs between the 4 packs of cells inside each battery. You would need to charge them enough to get them 'resting' preferably somewhere 13.8+ to have a good indication that the cell-level voltage was high enough to trigger the BMS to start balancing. This is why having BMSs with bluetooth so you can see the cell voltages is nice, because otherwise you have to see the 'total' resting voltage get pretty high to have any reliable indication that the cells are reasonably balanced.
Hopefully if my understanding of that is off, someone will point it out.
Last edited:
If you want to manually balance your cells without an active balancer, especially if you have a bad imbalance with a budget battery, charge the whole battery to 100%, then use a power supply and set it to 3.65. then charge up every single cell. They won't be on there for very long because the battery is already mostly charged. That's how I do it. I got tired of using the active balancers and I burned a few of them out.
But again if you have good cells, just cycle the battery. If there's an imbalance it will go away over time. If you're pulling full capacity then I wouldn't worry about it.
But again if you have good cells, just cycle the battery. If there's an imbalance it will go away over time. If you're pulling full capacity then I wouldn't worry about it.
Koyaanisqatsi
Electron addict
When the cell OVP kicks in at 3.65-3.7V, you can't feasibly charge a pack to 14.6V without causing an over-voltage in at least one cell, probably more. Back off your charge voltage to 14.45, so that you are not stressing cells by exceeding 3.65V. You might need to go even lower. Something down as far as 13.8V, just to get the BMS into cell balancing range, but not trip any OVP alerts. Observe cell voltages, then step up the charge voltage a small amount and hold for balancing. Rinse and repeat. But you should not be charging to the full 14.6V until you are certain the cells are FULLY balanced. And since that's never guaranteed, you should never charge to 14.6V. I believe this is what caused the capacity degradation. At 14.6V charge voltage, you are cooking one or two cells that have higher voltage than the rest of the pack.
91cavgt
New Member
Since these cells quality is questionable, and if I plan on doing ANY kind of other projects with cells in the future(and I do), then it sounds like an adjustable power supply is what I need.
Charge the entire battery up fully, disassemble the entire pack, charge each cell up to 3.65 volts, then reassemble the pack. At that point I can redo the original test with the 250 watt heater and can compare the results.
Vigo
Solar Wizard
If the OVP works at all then it becomes impossible to over-voltage any cell. Then, the voltage at which the bms 'disconnects' to protect cells, becomes an indicator of total balance. If for example the battery disconnects at 13.8v, you know one cell hit OVP while the others were still majorly behind. The closer you can get your 'disconnect point' to 14.6, the better the cells are balanced. This may require limiting charge current to a very low level so that the BMS's balancing function can drain the high cells into the low cells faster than the high cell's voltage can 'run away' over 3.65 and trigger a total disconnect.
Again, this is using indirect indicators to make up for not having a BMS that can just report cell voltages to you.
Yes!! You are now on the path!
Last edited:
Drunkin Solar
Solar Enthusiast
And after you go through all that replace those BMS's or you will wind up back in the same boat
This^ It works. It costs very little.
91cavgt
New Member
While I’ve got the whole thing disassembled, I’m going to add spring loaded speaker terminals to the external side of the large wooden battery housing. Each cell will have its own set of terminals. This way, I will be able to test voltage of each cell, and also charge each cell, without having to disassemble the entire pack again.
Eventually this setup will just be used for all of my 12 volt needs in the shed(lights, fans, stereo, ect) and I’ll put together a 48 volt system to use as my emergency power source and for all of my 120/240 volt needs.
42OhmsPA
Who likes Electrical Gremlins?
Something like this might be a better option.
(Not an affiliate link, copied from app).
91cavgt
New Member
That would probably work good! I picked up a bunch of those speaker terminals from Parts Express when they had them on clearance last year so it is something that I’ve already got on hand.
wpns
Solar Joules are catch and release
Remember, you don’t need two for each cell, you need one for each cell plus one additional for the battery.
justoneguy
New Member
- Joined
- Jan 30, 2021
- Messages
- 87
I did not think to add ferrules to mine, I'm glad I took the time to look at your link
sunshine_eggo
Victron's little biatch
No it doesn't. What's happening is that the BMS engages in protection mode at some lower voltage while the EPEVER output jumps to 14.6V with nothing going to the battery.
14.1V is way too high for float. REduce to 13.6V until the balance issue is sorted. Once sorted, drop to 13.5V.
91cavgt
New Member
91cavgt
New Member
Once I get all of the cells fully charged, I’m going to be asking for guidance on the settings to hopefully prevent this from happening again.
What BMS are you thinking of replacing it with?
I have two JK 4-8s 200a that I have zero complaints about, they go for around $80 if you order via AliExpress.
Similar threads
