diy solar

diy solar

Active balancer advice required

gillettadam

New Member
Joined
Sep 6, 2021
Messages
3
Good afternoon all. I have a Diy solar setup which I installed in December 2021. I am using a Diy lifepo4 pack which is 16s 2p. 280ah grade b cells. The setup is 2 parallel strings each with a daly 16s 200a smart bms. I know this isn't the best way to do this but originally I was going to have 1 string but decided 2 would be better slightly down the road. I top balanced both packs before installing but it looks like I have a potentially large issue in that one battery is operating fine but the other battery is drifting out of balance. This means that the bms is switching off the charge mosfet when the worst cell hits 3.6v this happens at around 54.8v ish. I then have the issue that the pack is no longer being charged but the other pack continues to charge which means the voltage of the 2 packs starts to diverge. As the voltage drops on the high cell the bms allows charge again and then I get a big spike current as string 1 dumps into string 2 which then spikes the worst cell again. Currently the system is performing well and if I hadn't monitored it I wouldn't know there was a problem, but would like to rectify it before it becomes an issue which I'm sure it will.

Would an active balancer help me or do I need to get some new cells and replace the ones that are weak? It does seem 2 of my cells are down on capacity (tested at 265 and 272ah respectively) compared to the rest which tested very close to 280ah. But that's the risk I took with grade b cells so can't really moan about that.

Obviously an active balancer is much cheaper than buying replacement cells so I'd rather do that if I can.

Many thanks in anticipation

Adam G
 
Can you post an actual photo, your description is confusing. Below is from my guide related to 48V Packs. The bottom one shows how to setup a 48V Battery with 32 cells set into a 2P parallel config. If you built 2x 16S and are attempting to "share" a BMS this will not work, not for long in any case. If this is indeed the case which is what your description seems to indicate, you are now seeing the downsides. The cells will require top balancing and another BMS will be required for the second pack.

There are several types of Active Balancers, some better than others and those which can be configured & those which cannot be configured. A Configurable Active Balancer with a MINIMUM of 2A Handling will be required for cells above 200AH. I have a fleet (6 packs) with QNBBM Active Balancers and Utility Packs with Heltec Capacitance Active Balancers (all not congurable) and they do work BUT can create some issues if not careful... I got NAILED by those issues, particularly on the "Low Side" (one pack has a channel fail, the result was 7 of 8 cells dropped to 1.50V whicle the dead channel remained at 3.320. No Warning or Indicators and no Low Volt Stop. As a direct rtesult of that Fiasco, everything is now being changed over to JK-BMS w/ 2A Active Balancer (which is fully configurable).
Link to Hankzor - JKBMS Vendor with excellent descriptions, pages, manuals etc.

Your charging profile is also likely too Aggressive, especially for B/Bulk Cells !
The max charge (Bulk/Absorb) should really be no higher than 3.425V per cell or 54.8V for a "48V" pack. Then set your SCC to FLOAT @ 54.6V (3.4125Vpc) which will allow for variable amps and to saturate the cells. ** REMEMBER to Correct for Voltage Drop between the SCC, Inverter & Battery Packs and there is ALWAYS some so you the Multimeter and check the voltages at the "Device Posts".

Hope it Helps, Good Luck.



1653230466455.png
 
Hi, thankyou for the reply. There are 2 seperate 16s strings each one with its own bms and fuse. They are joined together at the shunt and at a 2 way isolation switch which allows me to isolate either battery bank, or both from the inverter chargers. I have currently isolated the suspect pack from the system. The charge profile I have set is aggressive yes. 56.0v I can tone it down as you suggest. What cell voltage would you recommend I set for the bms(s) balancing to occur at this is currently at 3.42v. Do you not recommend I stick with the daly bms? It seems to work well, other than this issue, so I guess perhaps they don't work that well!!



Many thanks



Adam
 
IF that BMS is a Passive Balancer model, then you can set safely to 3.390 to 3.400. Do note that the moment charging stops the cells will immediately start to settle and they more that likely will have a fair bit of deviation initially. B/Bulk cells can deviate by 1mv per AH of capacity, especially when at the edges of the "Working Voltage Curve" (3.000V-3.400).

I was using Chargery BMS's with Solid State Contactors *Passive Balancing Capable* but had to add Active Balancers due to the various cells I have (Packs built over timer & a some used for thrash tests). Let me put it this way, $ I am replacing #3600 USD worth of BMS, Balancers & related for JK-BMS with 2A Active Balancing, with full comms etc. I will also put it plainly, I have never been a fan of Daly, they have a hell of a history.

Imbalances can also occur as a result o loose or imperfect connections, grease/waxes between terminals & busbars and other things, even a sense lead loosening up (that's nasty). If this is MOBILE, best to pop the tops and check everything is clean & tight.

Download this resource for charging info, curves etc... Luyuan Tech Basic Lifepo4 Assembly Guide
 
Hi Steve,

Yes the BMS(s) offer passive balancing. I believe it offers 0.2a of balancing only off of the highest voltage cell. I have reset my float and bulk charge voltages as you have suggested. I have also set the balance start voltage to 3.39 as you said. Hopefully this will help. The system is static, and once the packs have discharged down to 53v the cells are all within 0.01v so I think the balance/monitoring leads have good connection.

I will have a look at that bms. I have 16kw of inverter attached to the battery so hopefully they have a 400 amp version which would let me consolidate my pack into 1 system. At least with a DIY battery I can replace the individual cells if I need to. Next weekend I will switch to the suspect pack and monitor it to see if your suggested settings are enough to settle it down or if I need to take further action.
 
The Daly likely does not provide much passive resistor dump balancing current. Two things that degrade balancing faster is poor cell matching and high current discharge or charging.

I don't recommend the capacitor based active balancers due to reliability. They are okay if you only connect them periodically when you want to balance near top of charge. Leaving them connected all the time will do more to imbalance cells then balance them.

The Neey active balancer is DC-DC converter based and looks to be the best cheap active balancer.

Don't balance below 3.40v cell voltage, especially if you have marginally matched cells.
 
There are several types of Active Balancers, some better than others and those which can be configured & those which cannot be configured. A Configurable Active Balancer with a MINIMUM of 2A Handling will be required for cells above 200AH. I have a fleet (6 packs) with QNBBM Active Balancers and Utility Packs with Heltec Capacitance Active Balancers (all not congurable) and they do work BUT can create some issues if not careful... I got NAILED by those issues, particularly on the "Low Side" (one pack has a channel fail, the result was 7 of 8 cells dropped to 1.50V ...

Your charging profile is also likely too Aggressive, especially for B/Bulk Cells !
The max charge (Bulk/Absorb) should really be no higher than 3.425V per cell or 54.8V for a "48V" pack. Then set your SCC to FLOAT @ 54.6V (3.4125Vpc) which will allow for variable amps and to saturate the cells.

View attachment 95437
Hi Steve, ... Love your lessons and this option. I translated this for my 24v 280Ah battery banks to: "Bulk Charge: 27.4 v 3.425V per cell / then FLOAT 27.3v (3.4125Vpc) " which (you say) will allow for variable amps and to saturate the cells.

... Why I am seriously considering this option: I am pretty sure my Chargery BMS8T with its' 1.2A passive balancing (configured for >3.4v On Charge Only) ... would keep my cells all nicely balanced at this lower top charge voltage setting. When I originally obtained my B quality 280 Ah cells, my 24v battery bank could take a full charge to 29.0v while showing balanced cell v profiles at the top charge v. I see age has its affects on top of my few <2.5 v cell cycles (my learning curve mistakes). ... Learning from you and other folks, ... I appreciate the Heads Up Perspective: ... Active balancers that are ON all the time can unbalance your LiFePO4 cells at the lower cell voltage (that show @ high cell volts), ... plus can fail w miserable consequence to your batteries. I would only test an Active Balance I can somehow adjust for ON at >3.4 volts ... Thanks Steve, ... with my two cents of opinion :+)
 
Download this resource for charging info, curves etc... Luyuan Tech Basic Lifepo4 Assembly Guide
I would like to read that "Luyuan Tech Basic Lifepo4 Assembly Guide" for the lesson value .../ but that link only mentions it without any option to download (a few minutes ago) ... did notice 30 plus thumbs up likes. Wounder how to review the info ???
 
I would like to read that "Luyuan Tech Basic Lifepo4 Assembly Guide" for the lesson value .../ but that link only mentions it without any option to download (a few minutes ago) ... did notice 30 plus thumbs up likes. Wounder how to review the info ???
Top Right of the screen there is a download button, when you go to the resource.
 
I just stared down this road of building my own banks. I read up a lot on different BMS's models and such watched a fair mount of youtube videos. The ones I watched basically said .. "you dont need an active balancer .. the charger will deal with it over time". Well, I'm not sure how much I buy into that now that I have started top balancing cells and hooking them up to a BMS.

I got the Overkill BMS (basically the JBD BMS). It has a "passive" balancer, so I thought, "that will be better than nothing and can give me that little extra help". Not so much. To "balance" it just bleeds off the power from the cell via a resistor to match the others.

Now I am thinking I need to get a separate active balancer for each bank. Here is screen shot of my cells after I top balanced EACH one individually to 3.65v down to 0.2 amp charge. They all sat for about 3 days and then hooked them up into a 16s bank.

When I hooked up a 48v charger to the BMS, it charged until one cell hit 3.65v then shut off the charger .. But non of the other cells will ever get fully charged. Now it will just use the passive balancer to drain the higher cells faster, but the others will never get full.

Mu understanding is that an "active" balancer will take the power from the higher cells and use that to charge the lower cells (so they will all eventually get to to the same state of charge).

Screen Shot 2022-06-03 at 12.11.03 AM.jpg
 
Last edited:
Yours is a common problem with parallel packs. You can either apply a bandaid solution of active balancing and carry the associated reliability risk, or periodically manually rebalance your cells (will need to be done more frequently over time)
 
When an active balancer is properly implemented (only on when needed), what reliability risks are there?
 
When an active balancer is properly implemented (only on when needed), what reliability risks are there?

You can turn off the balancer, but i’ve seen a few active balancers fail in a state that continues to discharge a cell. Turning off the balancer won’t stop this issue.

The biggest issue is that the cell discharges through the active balancer, so the BMS can’t disconnect the cell when it reaches its LVD.

If you have the active balancer set up so that all the balancing leads are disconnected when not required or when cell level voltages are out of range then you remove this risk. I haven’t seen anyone implement this strategy.
 
A simple Response...
I spent an OBSCENE amount of Cash on Chargery BMS with Passive Balancing and then adding QNBBM-8S Active Balancers (@ 150USD ea) to every pack, it worked but not great, I've written a LOT about that whole "trip". Obscene = (>$3500 USD)

Everything is being replaced with JKBMS with 2A Active Balancers.
Two Packs I have are Used EV LFP and they have been always "quirky" and as such the "Finickiest" of the packs in my Bank. To get the cells above 3.350 and balanced was ALWAYS a challenge and usually, the best achieved 30mv differential. NOW with the JKBMS starting Active Balance @ 3.37V (I charge to Bulk/Absorb to 3.4375Vpc and Float at 3.425Vpc) and when the packs reach 100% the cells are ALL 3.425 +/- 0.005 Millivolts within 30 minutes as the deviation is usually 0.040 between cells @ 100% point.

The Active 2A bursts from the High Cells can be monitored on the BMS App or with a Clamp Meter and it is a beautiful thing to watch them balance up nicely with no stress.

JKBMS's unlike many others allow fully programmable settings for the Active Balancer while others do not, like the QNBBMs which never stop and CAN total a Battery quickly if they fail.

Case in point, I had one LFP Pack collapse because of a Failed QNBBM-8S. No App/No Control/No Alerts-Warnings !
Took the cells down to this:
1.163V 0.37 ir
2.737V 0.40 ir
1.180V 0.38 ir
1.184V 0.39 ir
1.181V 0.37 ir
1.183V 0.35 ir
1.178V 0.42 ir
1.184V 0.36 ir.
I did manage to Recover these Cells and Rebuild the pack, it took two weeks of intense effort and I "Got Lucky" and capacity loss is minimal.

Take this info and consider it. There is an astronomical pile of Twaddle around balancing systems and some still forward the wivetales and debunked prattle on their UTube & other outlets.

PS:
1 Chargery BMS8T-300 & 300A DCC = $350 USD +S & H & Duties/taxes
1 QNBBM-8S Active Balancer = $144.00 USD + S&H etc.
1 JK-B2A24S20P (2A Active Balancer w/ 200A capacity) = $175.00 USD +S&H etc. With App, Programmable & Communications capabilities !

so $495 wich can poop in your yard like it did to me, OR vs <$200 and happiness.
 
Most common issue with out of balance cells when attempting a top off, full charge, is a BMS shutting down for a cell overvoltage condition. It only takes a misbalance of about 1% state of charge capacity for this to be an issue.

Second issue is being significantly out of balance for cells, reducing the extractable battery pack capacity to the lowest state of charge cell capacity when charging was terminated. It also causes cells to wear at different rates reducing battery longevity.

This is a common problem with novice users of self contained 12v LFP batteries. The internal BMS has passive resistor dump balancer that does not dump current for balancing until a cell gets above 3.4v during charging. Their BMS balancing dump current is usually only 50-100 mA's.

When LFP battery is repeatedly only recharged to a voltage of less than 3.4v x number of series cell in a battery pack, the misbalance will grow over time and you will end up with significant amount of capacity state of charge difference between cells. It will take a very long absorb voltage time (>week) at higher charger voltage to get the cells back in balance. You will likely get multiple BMS overvoltage shutdowns along the way before cells get balanced again.

Running with cells out of balance for a long time will cause cells to wear at different rate, resulting is worsening cell matching. When cells are mis-matched they will have different internal impedance so will have more or less internal cell loss during discharge. Mis-matched cells are more difficult to keep in balance.

High discharge and charge rates accelerates balance degradation. It amplifies cell mismatching effects.

The greater the balancing curent, above 3.4v per cell, the better. There is a resistor heat dissipation limitation that can realistically be done with passive resistor dump balancers. 200 mA current dump at 3.6v cell voltage is about the limit for commonly used BMS PCB mounted chip resistors.

Chargery is one of the few BMS's that use MOSFET's with heat sink to do cell balancing current dumps. It creates a lot of BMS heating. Much better to have an energy efficient active transfer balancer.

16T top PCB.pngCharger internal heat sink.png
 
You can turn off the balancer, but i’ve seen a few active balancers fail in a state that continues to discharge a cell. Turning off the balancer won’t stop this issue.

The biggest issue is that the cell discharges through the active balancer, so the BMS can’t disconnect the cell when it reaches its LVD.

If you have the active balancer set up so that all the balancing leads are disconnected when not required or when cell level voltages are out of range then you remove this risk. I haven’t seen anyone implement this strategy.
WOW ... to: " seen a few active balancers fail in a state that continues to discharge a cell. AND... Turning off the balancer won’t stop this issue." ... Re: My future experimenting with the less expensive Heltec Active Balancer for one of my 24v LiFePO4 battery banks: ... Here is an inexpensive off/on switch that would fully disconnect that Active Balancer, ... that could be controlled by a voltage monitor. I was gona lean on just that on / off ... solder wire to a switch Heltec Option, ... but not after reading your story, plus Steve S's story about battery drain incidence from an active balancer malfunction. ... IMO: That newer JK BMS w 2A active balancer looks like the best option so far, from what I have reviewed. I also wonder when Chargery will put out their version A - BMSs ???, ... which will have an active balancer (info. from email dialog w Jason). Wondering how good they will be/ or not.
 
Last edited:
May want to check THIS posting - ponder it.
JK's hard at work, not really. It's mentioned elsewhere but the "174-1" Pack is the Pack that Failed due to a QNBBM-8S Failure.

More detail on the Recovered 174-1, with the Cell Voltage & IR's when it was Drained.
May open a sleepy eye somewhat.
 
You can turn off the balancer, but i’ve seen a few active balancers fail in a state that continues to discharge a cell. Turning off the balancer won’t stop this issue.

The biggest issue is that the cell discharges through the active balancer, so the BMS can’t disconnect the cell when it reaches its LVD.

If you have the active balancer set up so that all the balancing leads are disconnected when not required or when cell level voltages are out of range then you remove this risk. I haven’t seen anyone implement this strategy.

Can you tell me where you have seen this? Where you using the flying capacitor style active balancer. This has me concerned, I understand these can fail and stop working but I have not heard of them failing and then deciding to intentionally start killing cells. The voltage range on the active cell balancer I'm using is between 2.90-4.2 volts it turns off if any cell is lower than 2.9
 
Back
Top