diy solar

diy solar

BMS Getting confused

ianmorris1960

New Member
Joined
Nov 26, 2020
Messages
27
Hi,

OK, my set up, 300W solar, epever 30a controller, 100Ah lithium, with JBD BMS.

I have set the BMS to shut off charge at about 90%, and discharge at 20%, and mirrored these settings in the solar controller.

The loads are connected directly to the battery NOT the epever 'load' terminal

The problem, the BMS keeps shutting off both the charge and discharge switches.

I have been trying to work out when this occurs as the logging is not perfect on the BMS.

I have narrowed it down to the situation where the sun is shining, battery is charged (no solar output), a load starts to take power, firstly from the battery, then as the Epever algorith kicks in from solar. when the load is switched off the solar keeps supplying the power until it sees the battery is full, then it stops, however before this transition happens in the epever, the BMS has cut boh charge and discharge FETs. could this be what the 'load' port on the epever is to prevent?

I hope this makes a bit of sense, If someone can suggest useful avenues for further investigation, I am all ears.

Thanks in advance

Ian
 
Hi,

OK, my set up, 300W solar, epever 30a controller, 100Ah lithium, with JBD BMS.

I have set the BMS to shut off charge at about 90%, and discharge at 20%, and mirrored these settings in the solar controller.

The loads are connected directly to the battery NOT the epever 'load' terminal

The problem, the BMS keeps shutting off both the charge and discharge switches.

I have been trying to work out when this occurs as the logging is not perfect on the BMS.

I have narrowed it down to the situation where the sun is shining, battery is charged (no solar output), a load starts to take power, firstly from the battery, then as the Epever algorith kicks in from solar. when the load is switched off the solar keeps supplying the power until it sees the battery is full, then it stops, however before this transition happens in the epever, the BMS has cut boh charge and discharge FETs. could this be what the 'load' port on the epever is to prevent?

I hope this makes a bit of sense, If someone can suggest useful avenues for further investigation, I am all ears.

Thanks in advance

Ian

A commodity fet based bms should not be used as a primary means of disconnect for loads or charge sources.

If the bms disconnects the battery from the system on under-voltage this will turn off the solar charge controller.
If the solar charge controller is turned off it can't recharge the battery.
 
Thanks for you comment,

agree, the arrangement is far from ideal,

the scenario I have at the moment is not arising from low battery, in fact the battery charge level has not fallen below 60% - whatever that means as it is all based on voltages anyway! not coulomb counting

I am thinking I should use the load output of the epever, and give that more responsibility. I understand the BMS FETs should be the last line of defence, and by puttting the load through the epever load port, and only connecting the battery to the 'battery' port of the epever it seems like the layout the designers of the pever had in mind anayway (good job i'm not running a massive power inverter!
 
Thanks for you comment,

agree, the arrangement is far from ideal,

the scenario I have at the moment is not arising from low battery, in fact the battery charge level has not fallen below 60% - whatever that means as it is all based on voltages anyway! not coulomb counting

I am thinking I should use the load output of the epever, and give that more responsibility. I understand the BMS FETs should be the last line of defence, and by puttting the load through the epever load port, and only connecting the battery to the 'battery' port of the epever it seems like the layout the designers of the pever had in mind anayway (good job i'm not running a massive power inverter!
The load port on the epever is very likely a "checklist" item for people who don't know any better.
Its probably rated for 30 amps or less and I would not rely on them for half of that.

Please consult the manual and see if they give a current rating for the load port.
Also look at the largest wire the mechanical lugs can support.
Those 2 datums will likely be revealing.
 
Let's say you set the BMS to stop charging at 14.6v. The solar charge controller should have a setting below that, say 14.4v. That's how mine is setup. This way, the BMS should never have to step in and if it does, something is really wrong.

Cell over voltage is a different matter. What are your cell voltages when the BMS stops charging?
 
The load port on the epever is very likely a "checklist" item for people who don't know any better.
Its probably rated for 30 amps or less and I would not rely on them for half of that.

Please consult the manual and see if they give a current rating for the load port.
Also look at the largest wire the mechanical lugs can support.
Those 2 datums will likely be revealing.
the output rating of the port... I cannot find a spec anywhere, but the connectors are the same size as the solar and battery ports. I guess I could get the cover off and work out what they are! - I don't see why it would not be the same as the overall rating. Yep, I am a chinese specification skeptic, I have seen power electronics where the claims are way in excess of the laws of physics, and then way beyond the electronics inside.... lack of heatsinks is always a clue!

The design of the epever is for lighting I believe, there are all sorts of programming options to shedule the load on/off times, so the battery is soley for the purpose of shifting the power from day to night... the algorith seems to provide power to the load port from solar, and if there is more solar available than required then the battery is sent charge within parameters, if there is insufficient charge available from solar for the load, then the battery kicks in, so it is quite logical.
 
Yeah you should never be using the BMS to control charging. Ideally it's the last resort.

If it cuts off then it'll also shut off your power (if you've got a common port model that is), as you've likely noticed. All the charge control should be done via the charger and the BMS should only shut off if an excess voltage condition that can cause damage is present.
 
Let's say you set the BMS to stop charging at 14.6v. The solar charge controller should have a setting below that, say 14.4v. That's how mine is setup. This way, the BMS should never have to step in and if it does, something is really wrong.

Cell over voltage is a different matter. What are your cell voltages when the BMS stops charging?
I have forgotten the exact settings, but yes, the Solar should be cutting charging before the BMS gets upset, however this is not perfect, one of my cells charges faster than the others, so the BMS is tripping on cell over voltage, before the Solar cuts out on 'pack' voltage as it cannot see the individual cells of course.

I guess putting the PC on both the BMS and the solar would make sense, and just let it log until the event occurs would make sense, and I guess that would tell me the pack voltage at the point the BMS trips for charge, then work the settings back from there, which would work the one cell harder than the other three....

Still don't understand the discharge, again maybe the more detailed logging may shed some light

Ian
 
If the BMS cuts off on high cell voltage you need to top balance the pack so that they all reach the high voltage together.
 
If the BMS cuts off on high cell voltage you need to top balance the pack so that they all reach the high voltage together.
I agree. Top balance all your packs. If that cell still acting up, time to toss it and get a new cell. Power cut off with bms is last resort. One huge issue is when your bms auto reconnects while the panels are charging, you can easily fry your solar charger. Ive done it before.
 
If the BMS cuts off on high cell voltage you need to top balance the pack so that they all reach the high voltage together.
Top balancing was the first step I took when I got the cells, then made them up into a battery and hooked up the BMS to the PC and logged the first discharge and subsequent charge cycle, that is how I know there is one mismatched cell, three track each other but one is different.

So finding this I decided I had not top balanced correctly, so did it again, and left them paralled up for a week, but still the battery was the same. I guess that is what you risk buying cells off Alibaba ha!
 
Just hooking them in parallel will not top balance them. Did you use an actual power supply to do so?
 
I agree. Top balance all your packs. If that cell still acting up, time to toss it and get a new cell. Power cut off with bms is last resort. One huge issue is when your bms auto reconnects while the panels are charging, you can easily fry your solar charger. Ive done it before.
what sort of controller was it that fried? I have read about this, and put a panel to controller cutoff. this has now happenned a few times, and I still have an epever, oddly after a minute or so, assuming the sun is shining, it self powers, albeit it shows a dead battery!
 
Just hooking them in parallel will not top balance them. Did you use an actual power supply to do so?
Yes, I have a lab power supply that I use for the electronic development work I do, so I used that to fill each up to the recommended voltage, then let them rest and hit them again individually until the current at the max voltage (from the data sheet) had fallen to zero, before parelling I had done this to each cell.

Trust that is the correct process
 
I have a 48v system so I was using an epever 48v 100amp, it was about $500 bucks with all the accessories. I didn't know better back then and I damaged my solar charger when it got auto connect while the panels were charging in full sun. Dealing with epever was extremely frustrating. I posted my experience with them here. Long story short, I got a cheap, no name, mppt and it's been great and i also wire in such a way that the battery only disconnected from the inverter when low, not the solar charger.
 
Yes, I have a lab power supply that I use for the electronic development work I do, so I used that to fill each up to the recommended voltage, then let them rest and hit them again individually until the current at the max voltage (from the data sheet) had fallen to zero, before parelling I had done this to each cell.

Trust that is the correct process
Doing them individually will disrupt your balance.

Sticking them in parallel without a charger will not correct this in any amount of time.

Did you charge the whole pack to 3.65v after the imdividual charge while in parallel? Your comments are kind of ambiguous on this.

The first full charge of a top balance should be done in parallel rather than on the individual cells.

If you did do so, then it's safe to say there's a janky cell and it might be a good application for an active balancer.
 
Thank you for those who contributed to this, it turns out the BMS had a hardware fault, the seemingly random problem became more frequent, and a simple restart of the BMS no longer fixed the issue.

The BMS is a JBD from LLT, a 4s 80a. when I got a reason in the log as to why it had switched off temperature seemed to be the occasional culprit.

I finally extracted the battery from my van and dismantled it. I found that the connector for the external sensor was not soldered correctly, and well hanging off.

Not proof final, I have now soldered this back on, the BMS has remained working all day in the warm of my house, and of course without the vibration in the van.

I will be testing this over the next few days to see how it goes... 300 mile road trip this weekend, that should shake out any issues!

Thanks Ian
 
Back
Top