diy solar

diy solar

What is the advantage of choosing for a BMS that can talk to your inverter? (Seplos)

drbytes

New Member
Joined
Oct 21, 2021
Messages
113
Hi

I'm going to be adding a 48v 280AH battery to my existing 48v 304AH battery storage. Currently the 304AH battery pack has a JBD bms which is doing it's job, keeping the cells safe. It does not have any communication to my inverters (2P SPF5000ES), the inverters know when to disconnect the battery based on voltage they see on the conductors going to the battery.

I've seen some posts here on the forum of guys and gals using a Seplos BMS so that the battery pack can communicate with the inverter to report SOC, while I think that's pretty neat .. what does it really add ? The inverters know based upon voltage when to charge or discharge.

I'll have a 304 nd a 280AH, unmatched capacity, . does two Seplos BMS's make sense here because they will be able to talk to each other ?
And what's the use of that screen they sell with the BMS? I see that as extra wiring and complexity to mount the screen which will never be seen since the packs are in a crawlspace.

Would really appreciate any input.
 
IMHO: Coms between the battery and the inverter add very little value for all the reasons you have already mentioned.

Where Coms with the BMS can add value is if you have a centralized system/display that can show information about all of your different components.....and even then it only becomes important if the system is not working as expected.

The primary bit of data that I find useful to have easy access to is the State-of-charge, and that is primarily on systems where there is no easy back-up to the batteries (Like in an RV). For a stationary system that has grid back-up, it is less important.

Having said that: DIYers also tend to be tinkerers that like to play with the system and constantly tweak the system. If you are that type, then having a centralized data display can be fun and useful.

In one of the @Will Prowse videos, he showed how to hook up the coms to the BMS and then promptly said he does not think it is worth it.

Here is a video from another channel that talks about this question:

Edited:
Finally, Does the Inverter have the ability to show SOC on multiple batteries?
 
Last edited:
When you have multiple batteries in the system, which BMS does the inverter know to listen to? If one BMS says, "I'm full, stop charging," but the other BMS has yet to achieve a full state of charge, how do you referee that?
 
When you have multiple batteries in the system, which BMS does the inverter know to listen to? If one BMS says, "I'm full, stop charging," but the other BMS has yet to achieve a full state of charge, how do you referee that?
To tell the truth, I am not sure what commands and communications are passed back and forth and it is not very well documented. If the BMSs are telling the inverter to stop charging, you have a point but..... If both BMSs are set with the same parameters, when the first BMS says to stop charging, the other one will be pretty close to full. (The parallel batteries should be at pretty much the same voltage).

However, if the BMS is deciding when to stop charging and not the inverter, then you are down to only one level of protection for the cells. I like to set things up so the Charges and loads cut out before the BMS. That way, if the charger or load fails to do its job, the BMS is there as a backup.
 
However, if the BMS is deciding when to stop charging and not the inverter, then you are down to only one level of protection for the cells. I like to set things up so the Charges and loads cut out before the BMS. That way, if the charger or load fails to do its job, the BMS is there as a backup.

No disagreement there.

It's been a while since I had to do programming for serial devices. A long time ago with Honeywell hardware. Those devices identified themselves within the communication packet. If the BMS doesn't have that ability then the inverter (or whatever upstream device is receiving) would have no idea what to do with conflicting information coming from different BMS. BMS 1 says stop charging, I'm full. So the charger stops. BMS 2 says I'm not full, so the charger, not realizing that there are two BMS on the line, resumes charging.
 
No disagreement there.

It's been a while since I had to do programming for serial devices. A long time ago with Honeywell hardware. Those devices identified themselves within the communication packet. If the BMS doesn't have that ability then the inverter (or whatever upstream device is receiving) would have no idea what to do with conflicting information coming from different BMS. BMS 1 says stop charging, I'm full. So the charger stops. BMS 2 says I'm not full, so the charger, not realizing that there are two BMS on the line, resumes charging.
Yes.... it could get weird. Furthermore, trying to get details on what happens with the Growatt and MPP coms is nearly impossible.
I like to know how the system works, so that is just another reason to not use the coms.
 
Having interconnected bms allow you to get a general and detail overview of you battery systems. Your bms has a better view of the soc, and know better when to charge or not. In particular the bms need to recalibrate from time to time, to maintain soc accuracy the way to do it is to instruct the inverter to fully start to charge. The bms can precisely tells the inverter at what rate it could charge or discharge. That being said, not all inverter can use this information wisely.
 
In that case, I mean, reading all this, there doesn't seem a clear answer. Two people that are quite into it even made videos about it stating that there is no real benefit.

This system I have installed and plan to expand will need to perform and will need to be robust. It sounds like adding another level of complexity (bms <-> inverter comms) is a potential additional failure that could bring my system down with no real benefit.

However, what about pack to pack communication? Is there, for instance with the Seplos BMS, a benefit of having the 304 pack talking to the 280 pack?
 
The one bms connected to the inverter have the knowledge of all individual bms, it can report total capacity and soc. If your concern is to have more simple system don't use it. This kind of bms is beneficial as long as your inverter can exploit all the information report by the bms.


Some inverter (like voltronic) just use it to setup the 4 parameters for the battery setting. There is no real value. It is up to you and your inverter.
 
The one bms connected to the inverter have the knowledge of all individual bms, it can report total capacity and soc. If your concern is to have more simple system don't use it. This kind of bms is beneficial as long as your inverter can exploit all the information report by the bms.


Some inverter (like voltronic) just use it to setup the 4 parameters for the battery setting. There is no real value. It is up to you and your inverter.
well, the newer voltronics also use it to get a proper soc and ramp down amps when almost full
 
I use an SMA system, having the BMS control the inverter and charger has the following obvious benefits:

- ability to ramp down current while in cell balancing phase.

- controlling current with respect to SOC and cell temperature.

- being able to turn off non-critical loads at low SOC setpoint, and turn on excess loads (eg hot water) at high SOC setpoint.

- being able to monitor all chargers / inverters from a central interface.

The BMS is the best placed component in the system to control inverter / charger functions as it has cell level temperature and voltage, as well as SOC hysteresis data. I wouldn’t set up a whole house power system any other way.

Key functions that improve cell life:
- low balance current
- low current as high/low SOC is approached
- low current at high or low temps.

All these things are very difficult to achieve using functions of individual components.
 
The biggest reason why I went with the Seplos is because of communication. Seplos has it programmed so that as the pack reaches full charge, it will ramp the charge rate down, so that way it gives the pack time to balance before it hits its float charge voltage.

I had a Daly BMS that would only balance the cells while charging, which the time it takes from 3.4v to 3.55v its very little, making the balance window very small. Seplos on the other hand can drag that time out, and give the balancing circuit time to do its job. Better balanced pack = more capacity.

Its also another level of defense. First level being the settings in the inverter, then the BMS telling the inverter to stop or start charging and at what rate, and the last is the MOSFET's disconnecting in the BMS. Can't have too many.
 
A few things I don't see here that deserve being mentioned:
1. In my opinion, a BMS is valuable for LiFePo batteries because of how they read SOC. It isn't a straight line for a charging/discharging curve. The BMS reads a number of parameters to determine where 100% is, whereas without that communication the inverter has to go off voltage. With voltage, .5v could be a difference of 30% SOC or 5% SOC depending on a number of other factors. So, it helps keep batteries healthy by taking other factors into account on charging/discharging.
2. BMS comms between batteries are similar - voltages can vary slightly between the batteries so when 1 fills up, it turns off that battery for charging/discharging and you still might have a little juice in another one. Typically your system will assign 1 battery as the master and the rest as slaves for the actual comms to the inverter. The inverter isn't listening to 10 batteries, it is only listening to the master (typically).

Just a little input from me.
 
A few things I don't see here that deserve being mentioned:
1. In my opinion, a BMS is valuable for LiFePo batteries because of how they read SOC. It isn't a straight line for a charging/discharging curve. The BMS reads a number of parameters to determine where 100% is, whereas without that communication the inverter has to go off voltage. With voltage, .5v could be a difference of 30% SOC or 5% SOC depending on a number of other factors. So, it helps keep batteries healthy by taking other factors into account on charging/discharging.
2. BMS comms between batteries are similar - voltages can vary slightly between the batteries so when 1 fills up, it turns off that battery for charging/discharging and you still might have a little juice in another one. Typically your system will assign 1 battery as the master and the rest as slaves for the actual comms to the inverter. The inverter isn't listening to 10 batteries, it is only listening to the master (typically).

Just a little input from me.
its a big input, you said it clear and better than I could :-explain )
 
Thanks for your replies.

So there are some advantages but it all kind of depends on the charge controller, or inverter, that will talk to the BMS; how does it use the data it is getting from the BMS to manage it's charging.
Unfortunately, there isn't a lot to be found on the growatt website abiut this. I had another look at the Seplos doc and it says it uses the Goodwe protocol, I'll have to see if I can find more info on that online.

As it stands I honestly don't think it justifies the additional cost. For instance my inverter has Bulk/Float settings for me the set thresholds, the BMS is aware when to start equalizing and knows it's own SOC. Add to that that I've been pulling at minimum full capacity from my battery bank since it has been installed and none of the cells is out of balance despite being hammered hard.

I would've been interested if the interpack comms would help keeping the packs in balance but I heard from Seplos that they do not recommend having packs of different capacity in parallel. If they were the same, one pack would charge the other pack to keep balance.

That makes me wonder once again, if I shouldn't just pair a 280 and a 304 and create a 16S of 584AH cells.
 
That makes me wonder once again, if I shouldn't just pair a 280 and a 304 and create a 16S of 584AH cells.
That is an interesting proposition. I have not seen anyone do something like that but it may be the better way to parallel cells of different capacities (assuming the BMS can handle the capacity of the 'mega cell'). The problem you are likely to run into is that the cells will be of different physical dimensions and getting the Busbars to all line up might be a challenge.
 
Last edited:
you need to check the max cells capacity of you BMS. you can't do that with seplos ( 500Ah max for each cell). you need to build one 16s BAT with the 280Ah cells and BMS, and an other with the 304Ah cells and BMS. then put the 2 BAT in parallel.
 
you need to check the max cells capacity of you BMS. you can't do that with seplos ( 500Ah max for each cell).
I wonder what the limit is from. It may be primarily a limit of it's internal SW. I also wonder if there are similar limits on other BMSs.

So..... I will modify my previous post to include the BMS limit proviso, but I still think parallel first may be a better way to parallel cells of different capacities from the point of view of keeping the cell loads the same.
 
I wonder what the limit is from. It may be primarily a limit of it's internal SW. I also wonder if there are similar limits on other BMSs.

So..... I will modify my previous post to include the BMS limit proviso, but I still think parallel first may be a better way to parallel cells of different capacities from the point of view of keeping the cell loads the same.
Yep It is in the seplos bms documentatio
 
Back
Top