diy solar

diy solar

Catastrophic failure 200ah 12v battery

Temp outside was 107*, inside the shed was less than 120*, so I can not see that being the problem.
I had two days prior added "Heltec Whole Group Balancer " to the mix, but can not imagine that overriding the BMS to allow runaway voltage.
The BMS, from aliexpress, was only one week new replacing a Daly which allowed up to 3.7v -Daly stated was OK but I could not get comfortable with.
Also, I am more comfortable topping out at 3.4V/13.6v to extend life of cells (I thought...)
Cell 1 expanded almost uncontrolled. Currently sits at 3.3v SOC after 4 days.
Cells 2,3,4 expanded but much less so. They currently sit at 3.50/51V.
So, a couple of questions:
-I assume the BMS is the cuplrit -any thoughts? I certainly do not want to attach it to my good bank to experiment.
-Is it possible that it was simply a faulty cell? I have had them for just over 1 month, but have certainly less than 10 full cycles of use.
-How could I test the remaining 3 to verify fitness for service, without pulling another cell from my good bank?
-Or is this simply an horrible idea, and just walk away from the $525 worth of cells?
I purchased cells from an amazon seller with no feedback, suspect, but do not want to pursue chargeback without a high degree of certainty.
Thank you in advance for any assistanceView attachment 55938View attachment 55939

That is a JBD BMS, what are the settings? Any alerts (like over voltage)?
Certainly if not connected correctly, an active balancer can cause problems.

Do you have a close up picture of the active balancer? I ask because there are different settings for it fo different chemistry cells.
Did you need the active balancer, or just decide to add it?

Perhaps screen shots of the BMS settings, and a close up of the active balancer can assist with diagnosing the problem.
My cells in the garage in Arizona regularly hit high temperatures, however I try to avoid charge and discharge when too high, a swamp cooler can help cool things off by up to 20 degrees if you have low humidity.

One of the reviews on Amazon even shows what I'm talking about for the active balancer:
balancer.jpg
 
Does Heltek know not to charge a cell to 4.5V, even if it thinks (incorrectly, due to connection issues) that another cell is that high?
Perhaps it got confused into overcharging one cell.

So it takes power from multiple nominal-voltage cells to charge individual low cells? Or takes voltage from the series string, some 40 to 60V, and uses that?

I haven't studied circuit diagrams for the gizmos, but sounds like BMS can't stop active balancer.
 
I apologize for not adding anything to your specific question but seeing this, I am curious what your and anyone else opinion is on how much the box failure contribued. Had the box remained intact, would the batteries have survived?
The cells would have blown the pressure relief valve on the top under the plastic cover (that is its purpose).
 
This is the second prismatic cell LiFePO4 battery with Daly BMS that is destroyed in a month. ZERO FIRE !
The BMS main function is to protect the cells. In the other case 90V+ was going to the 24V BMS because of a faulty MPPT. I agree that is 4 times higher, could not handle it.

Here I see other issues:
- PWM charge controller for LiFePO4 ? That's bad. PWM is only a switch that allows the 19V solar to 12V battery, then stops, then starts ... in a 1-200 Hz frequency. Lithium batteries hate it. Even lead acid batteries hate it in a long run (in short de-sulfates them with this hammering).
- BMS will not recognize the V spikes. Visible only on an oscilloscope. So BMS can not protect against it. Can even damage BMS (BMS has MOSFETs and they have a max allowed V).
- Charge to 3,65V? Not needed, too much. You need to use CC(-CV) charger. If you reach 3,5V then battery already 95-98% full. No need even for absorption (see @Off-Grid-Garage videos in this matter)
The pictures I see show a JBD BMS, where is the Daly part coming from?
Specifically a 120 amp 4S JBD BMS, I have one. I also have a Daly.

A PWM controller should indeed not be used with LiFePO4 cells. It pulse voltage in, and likely fast enough the over voltage protection of the BMS doesn't catch it, the BMS disconnects, waits a specified interval, then reconnects (so you can hopefully DISCHARGE).
 
Last edited:
Probably not. I expect that the vent would have blown, spilling electrolyte everywhere.

The cells would have blown the pressure relief valve on the top under the plastic cover (that is its purpose).

So here is my next question then. What is causing the expansion? Heat or current? And, is this a gas that is being created in the cell or is it liquid that is expanding?
 
The pictures I see show a JBD BMS, where is the Daly part coming from?
Specifically a 120 amp 4S JBD BMS, I have one. I also have a Daly.
Misread in first post and remembered wrong ... Heltec has a naked BMS like it. Low Amp Daly is red too. Sorry.


A PWM controller should indeed not be used with LiFePO4 cells. It pulse voltage in, and likely fast enough the over voltage protection of the BMS doesn't catch it, the BMS disconnects, waits a specified interval, then reconnects (so you can hopefully DISCHARGE).

The BMS will not catch the high volt pulses from a PWM SCC.
Like you have 28V solar, 12V battery, want to charge at 14V.
The PWM is 100Hz so in a sec 50 times circuit is closed but between them 50 times open. So 1-0-1-0 ...
The average is 14V.
But 50 times in a sec for 1/100 sec you give 28V to the battery.
 
So here is my next question then. What is causing the expansion? Heat or current? And, is this a gas that is being created in the cell or is it liquid that is expanding?
The most common cause is the electrolyte boiling internally. That can happen from over charging, and also from over discharge.
 
So here is my next question then. What is causing the expansion? Heat or current? And, is this a gas that is being created in the cell or is it liquid that is expanding?
Is there a reason why you don't give any useful details, like BMS settings?
 
The BMS will not catch the high volt pulses from a PWM SCC.
Like you have 28V solar, 12V battery, want to charge at 14V.
The PWM is 100Hz so in a sec 50 times circuit is closed but between them 50 times open. So 1-0-1-0 ...
The average is 14V.
But 50 times in a sec for 1/100 sec you give 28V to the battery.
You are missing something important here, inductance and capacitance of the system. Those PWM peaks are not really damaging as long as average is still under the max allowed. Nothing is wrong with using PWM on Lithium as long as it's correctly setup and not faulty.
PWM will go to zero when battery reaches preset limit. If it doesn't, then it's faulty or limit is set too high.
 
A PWM controller should indeed not be used with LiFePO4 cells. It pulse voltage in, and likely fast enough the over voltage protection of the BMS doesn't catch it, the BMS disconnects, waits a specified interval, then reconnects (so you can hopefully DISCHARGE).
Where does this FUD come from? Nothing is wrong with PWM, it is smoothed well enough by inductance and capacitance by the time it reaches the cells.
 
Where does this FUD come from? Nothing is wrong with PWM, it is smoothed well enough by inductance and capacitance by the time it reaches the cells.

A PWM does not have it. You need high frequency (like kHz) for that and that is only done in MPPTs (based on buck converter).
 
A PWM does not have it. You need high frequency (like kHz) for that and that is only done in MPPTs (based on buck converter).
A PWM is a term describing a principle of operation and all modern power conversion devices work by this principle, so to say that PWM is bad is just nonsense. What people call a particular product and efficacy of that product can be analyzed and debated, but a generic statement is nonsense.
Every system has inductance and capacitance, it comes from all cables and components. LFP cells alone have very high capacitance and inductance.
PWM cannot hurt a battery unless we are talking some ridiculous low frequency like 1 Hz. Only average current/voltage matters, peaks don't matter.
 
The working voltage of the Heltec is rated 2.7V-4.5V. This is so they can be used with different lithium chemistries. The active balancer moves current from the highest cell to the lower cell to balance the voltage. It has no external connection to a power source.

The high voltage disconnect is regulated by the charging source and if a cells voltage goes too high the BMS should disconnect the charging source. This is my understanding of active balancers and I don't understand how a defective active balancer would cause a cell to go over the BMS voltage settings. I could be missing something and I am always open to being educated :)
Unless I'm missing something, the balancer is connected directly to cells, so it's bypassing the BMS, in which case it will happily destroy all cells while trying to save one. Active balancers are menace, why would anyone use them? It's a rhetorical question, no need to explain....
 
Is there a reason why you don't give any useful details, like BMS settings?
Because I am asking a question about what is happening inside a cell when it is bloating. I dont care about the BMS settings.

The cells would have blown the pressure relief valve on the top under the plastic cover (that is its purpose).
So, boiling is a condition that pressure can suppress. if it was expansion of a liquid, then yes, confining the cell would have resulted in the pressure relief valve releasing pressure.

The fact that the box failed, the pressure was allowed to act on the cells container and deform it. We have no idea if what the internal pressure of the cell was, only that it was more than the container could withstand and less than the pressure relief valve. The fact is that if the box had not failed, the pressure relief valve may have very well relieved the pressure. It is also very possible that the pressure relief may have remained intact and the cell not deformed.

Think about it like a tire with a tube. The tube by itself will expand at very low pressure. Stick it in a tire and now it is constrained by the tire and can experience much higher pressures. Unlike a tire the cell has a pressure relief valve but again, we have no idea what the internal pressure of the cell was or what the pressure relief valve will unseat at. Those cells may have very well survived this incident in a lot better condition if the box had not failed. That cell may have relieved itself and been bricked anyway.

Story short. Build better boxes. ;)
 
Temp outside was 107*, inside the shed was less than 120*, so I can not see that being the problem.
I had two days prior added "Heltec Whole Group Balancer " to the mix, but can not imagine that overriding the BMS to allow runaway voltage.
The BMS, from aliexpress, was only one week new replacing a Daly which allowed up to 3.7v -Daly stated was OK but I could not get comfortable with.
Also, I am more comfortable topping out at 3.4V/13.6v to extend life of cells (I thought...)
Cell 1 expanded almost uncontrolled. Currently sits at 3.3v SOC after 4 days.
Cells 2,3,4 expanded but much less so. They currently sit at 3.50/51V.
So, a couple of questions:
-I assume the BMS is the cuplrit -any thoughts? I certainly do not want to attach it to my good bank to experiment.
-Is it possible that it was simply a faulty cell? I have had them for just over 1 month, but have certainly less than 10 full cycles of use.
-How could I test the remaining 3 to verify fitness for service, without pulling another cell from my good bank?
-Or is this simply an horrible idea, and just walk away from the $525 worth of cells?
I purchased cells from an amazon seller with no feedback, suspect, but do not want to pursue chargeback without a high degree of certainty.
Thank you in advance for any assistanceView attachment 55938View attachment 55939
I think the same happened to me!
i attached two 12v/100 amp batteries in pararel and thet swell in 4 days.

what could be wrong?

thanks
 
Certainly if not connected correctly, an active balancer can cause problems.
The balancer is able to balance LTO, LI-ION, and LFE cells. I wonder what the default jumper setting is and if there is a need to change it for LFE cells?
 
Unless I'm missing something, the balancer is connected directly to cells, so it's bypassing the BMS, in which case it will happily destroy all cells while trying to save one. Active balancers are menace, why would anyone use them? It's a rhetorical question, no need to explain....
Yes you are correct it is connected directly to the cells. But I am still having a difficult time comprehending how it could cause a cell to overcharge. I think we need to know what the default setting is for this particular Heltec active balancer. We know it's not for LTO. So it has to be either Li-ion of LFE or the same setting works for both.

I will explain why I don't use them. Simple...I keep my cells between the knees which keeps the delta low enough to satisfy me :)
 
Last edited:
Yes you are correct it is connected directly to the cells. But I am still having a difficult time comprehending how it could cause a cell to overcharge. I think we need to know what the default setting is for this particular Heltec active balancer. We know it's not for LTO. So it has to be either Li-ion of LFE or the same setting works for both.

I will explain why I don't use them. Simple...I keep my cells between the knees which keeps the delta low enough to satisfy me :)
It probably didn't cause overcharge, but possibly made it worse by spreading excess energy into adjacent cells. Cause was likely misconfigured or misbehaving solar controller. Who knows what if anything the BMS was doing ?‍♂️
 
You are missing something important here, inductance and capacitance of the system. Those PWM peaks are not really damaging as long as average is still under the max allowed. Nothing is wrong with using PWM on Lithium as long as it's correctly setup and not faulty.
PWM will go to zero when battery reaches preset limit. If it doesn't, then it's faulty or limit is set too high.
PWM pulses in 18v, cell gets over 3.65v, BMS disconnects, active balancer eventually gets ALL cells to 3.65v, BMS disconnects. BMS waits specified interval, reconnects, disconnects after first 18v pulse. Repeat every minute for 3 days.

PWM is not a good idea with an active balancer. MPPT on the other hand charges to a set voltage and either stops (due to equal voltage) or floats, hopefully someone intelligently sets float below 3.65v per cell. Someone got everything to above 14.6v, for days on end.
 
Back
Top