diy solar

diy solar

Mixing Batteries from different Manufacturers?

Kenny_

Solar Wizard
Joined
Aug 6, 2023
Messages
1,643
Location
Denver, Co
I searched around a bit but not sure I got a definitive answer. The question is, 'Is it okay to mix different manufacturer's rack mount batteries in parallel?'

Or is that just not allowed. I suspect it's a no-no so once you commit to one brand you have to stick to it if you expand?

I also suspect there's all kinds of monitoring and management issues that make different brands incompatible if used together.

Thanks!
 
If you use LFP, you can mix them.

Be sure they all have the same nominal voltage. e.g some are 15s and while most are 16s. Only mix the 16s with 16s and 15s with 15s.
you can't mix 15s with 16s.
Makes sense!
 
If you are not using closed loop BMS communication different brands can be mixed if they are the same chemistry and number of cells.
Makes sense. But that means your inverter can't really monitor them and know the status etc right? Or are you talking something else?
I'm new to much of this terminology and stuff.
 
Makes sense. But that means your inverter can't really monitor them and know the status etc right? Or are you talking something else?
I'm new to much of this terminology and stuff.
Some people find direct communications with the inverter very important. I don't. My solar charge controllers are configured to charge to a certain voltage (in my case 55.8 volts). If 100% is reached the current tempers automatically, since the batteries can't absorb anymore. Communications is not neccesary for that.

Indeed without communications, the inverter does not 'know' the exact state of charge of the batteries. I connected solar assistant to my batteries and that provides me with a reasonable indication of the SOC. Enough for me.

In case my batteries are running too low, my inverter will cut-off/switch to the grid and that is based on a voltage setting as well in the inverter. The inverter measures the voltage of on the busbar and 'knows' when to switch to AC power.

What I meant with the 15S vs 16s.
Most server rack batteries (48v) are 16s. Meaning that there are 16cells in the batteries. Pylontech delivers batteries that are 15s.
 
Correct, the inverter would not be able to communicate with the BMS so State of Charge information would not be available. You would have to select a USER settings battery charging profile based on Voltage.
The industry is starting to adopt brand crossover communication compatibility but at this point its difficult enough just to get multiple batteries of the same brand as the invert to work correctly. The chances of having mixed brands all work together is doubtful.
 
Some people find direct communications with the inverter very important. I don't. My solar charge controllers are configured to charge to a certain voltage (in my case 55.8 volts). If 100% is reached the current tempers automatically, since the batteries can't absorb anymore. Communications is not neccesary for that.

Indeed without communications, the inverter does not 'know' the exact state of charge of the batteries. I connected solar assistant to my batteries and that provides me with a reasonable indication of the SOC. Enough for me.

In case my batteries are running too low, my inverter will cut-off/switch to the grid and that is based on a voltage setting as well in the inverter. The inverter measures the voltage of on the busbar and 'knows' when to switch to AC power.

What I meant with the 15S vs 16s.
Most server rack batteries (48v) are 16s. Meaning that there are 16cells in the batteries. Pylontech delivers batteries that are 15s.
Interesting....I'm learning fast... :) I see in this thread/post that the user was instructed to set the communication to PLY (pylontech) but to 16 Cells
I guess PLY is just type of battery communication but not necessarily the number of cells. Right.
(I just ordered this same setup and will be setting it up with the sungold batteries as well)
 
Interesting....I'm learning fast... :) I see in this thread/post that the user was instructed to set the communication to PLY (pylontech) but to 16 Cells
I guess PLY is just type of battery communication but not necessarily the number of cells. Right.
(I just ordered this same setup and will be setting it up with the sungold batteries as well)
that is correct
 
here a video from Will.

I'd like to see the same demonstration of charging with the buss bar setup. I'm a bit skeptical because the buss bar has resistance as well.
Wonder at the difference if the wires/cables were connected in the same manner as his buss bar diagram - i.e. diagonal: Positive at upper left, Negative at lower right. Hmmm.
 
I concur that different batteries with different manufacturers but same chemistry should be fine. In addition to the comments above realize that different BMS manufacturers have different sequences of operation (logic, programming) which will result in some batteries taking a larger share of charge/discharge cycles and may wear out sooner for reasons other than age, capacity, or equal length conductors etc. I may have batteries that reach some form of cycle-based degradation before others and have to be re-matched into different banks.

For example between two Trophy batteries in my bank they have different BMS and one will stop charging (disconnect charge mosfet) at 100% SOC regardless of the DC bus voltage and coast down to 96%SOC before re-charging. When the batteries go for days without a lot of use this battery experiences less charge/discharge cycles than another battery (same "package" manufacturer) that has a different sequence and will always charge when applied with voltage greater than pack voltage.
 
Last edited:
Some people find direct communications with the inverter very important. I don't. My solar charge controllers are configured to charge to a certain voltage (in my case 55.8 volts). If 100% is reached the current tempers automatically, since the batteries can't absorb anymore. Communications is not neccesary for that.

Indeed without communications, the inverter does not 'know' the exact state of charge of the batteries. I connected solar assistant to my batteries and that provides me with a reasonable indication of the SOC. Enough for me.

In case my batteries are running too low, my inverter will cut-off/switch to the grid and that is based on a voltage setting as well in the inverter. The inverter measures the voltage of on the busbar and 'knows' when to switch to AC power.

What I meant with the 15S vs 16s.
Most server rack batteries (48v) are 16s. Meaning that there are 16cells in the batteries. Pylontech delivers batteries that are 15s.
Hi , I would like to know the following, i have 2 x luxpower inverters in parallel with 2 x Freedom Won (51.2 100a ) batteries connected to the slave inverter (with BMS) running perfect, can I add another brand battery (1 x 51.2 100a green lantern) to the master inverter and set it up as lead acid until i correct bms connection for the green lantern battery?
 
I do have a mix of 3 SunGoldPower SG48100P batteries in one rack and 6 EG4 LifePower4 batteries in another all connected in parallel. Closed loop to the 6 LifePower4 to an SGP 10kw inverter via an EG4 Communications hub. Charging and discharging are pretty much balanced. Obviously SOC being reports to the inverter is only based on the 6 LifePower4 batteries.

I am creating a software BMS that can read/monitor all batteries and provide closed loop communications to the inverter. Software will replace the Hub. It would be nice if EG4 would allow protocol select on the two "Bat" inputs to the hub so different protocols could be accommodated.

I am also toying with software that extends the LifePower4 battery comms to the SunGoldPower batteries so the hub sees 9 batteries instead of 6.

Bottom line, the mix so far has not presented any issues
 
Hi , I would like to know the following, i have 2 x luxpower inverters in parallel with 2 x Freedom Won (51.2 100a ) batteries connected to the slave inverter (with BMS) running perfect, can I add another brand battery (1 x 51.2 100a green lantern) to the master inverter and set it up as lead acid until i correct bms connection for the green lantern battery?
If you run two inverters in parralel, you can share a battery bank and connect both inverters to that battery, using a busbar.

Adding another battery to that system is, in my opinion, fine as long as it has the same chemistry.

I personally should not use different battery settings (lead acid and lfp) per inverter as suggested.

If all batteries are in parallel, it should not be an issue if one has no communication. Charger looks at the voltage of the total bank and after one or two cycles they should charge up evenly.

Maybe the best way forward is to keep your current configuration, since that works fine for you, and just add the new battery in parralel to that bank.
 
Some people find direct communications with the inverter very important. I don't. My solar charge controllers are configured to charge to a certain voltage (in my case 55.8 volts). If 100% is reached the current tempers automatically, since the batteries can't absorb anymore. Communications is not neccesary for that.

Indeed without communications, the inverter does not 'know' the exact state of charge of the batteries. I connected solar assistant to my batteries and that provides me with a reasonable indication of the SOC. Enough for me.

In case my batteries are running too low, my inverter will cut-off/switch to the grid and that is based on a voltage setting as well in the inverter. The inverter measures the voltage of on the busbar and 'knows' when to switch to AC power.

What I meant with the 15S vs 16s.
Most server rack batteries (48v) are 16s. Meaning that there are 16cells in the batteries. Pylontech delivers batteries that are 15s.
It is hard though to get Solar Assistant to monitor batteries from different manufacturers as the battery to battery communications between the different batteries in the stack is not possible unless both manufacturers batteries have the same internal BMS). If only Solar Assistant had multiple inputs for batteries and shunts.
 
I am toying with a software solution that will allow me to add three SunGoldPower batteries to the end of my six EG4 LifePower4 batteries. I just need to convert and forward the "Master" query to the SunGoldPower batteries and convert the response back to the "Master"

Got the pieces and parts in source code (C# and Python), just need to glue it together and test.

Problem is; not enough coffee in the morning and not enough scotch in the evening.
 
I am toying with a software solution that will allow me to add three SunGoldPower batteries to the end of my six EG4 LifePower4 batteries. I just need to convert and forward the "Master" query to the SunGoldPower batteries and convert the response back to the "Master"

Got the pieces and parts in source code (C# and Python), just need to glue it together and test.

Problem is; not enough coffee in the morning and not enough scotch in the evening.
Mr. Coffee! That's the answer!

P.S. was wondering where you were on that so thanks for the update!
 
I wish the battery manufacturers would provide BMS documentation in the battery user manual.

For example; the SOC and SunGoldPower batteries use the same monitoring software, SOKTools or PBMS_Tools. Do both batteries have the same BMS with the same firmware? Is it might be possible to mix them and the battery to battery communications would/might work.

Just need better documentation and hopefully not in Chinese (it is PITA to convert)
 
I wish the battery manufacturers would provide BMS documentation in the battery user manual.

For example; the SOC and SunGoldPower batteries use the same monitoring software, SOKTools or PBMS_Tools. Do both batteries have the same BMS with the same firmware? Is it might be possible to mix them and the battery to battery communications would/might work.

Just need better documentation and hopefully not in Chinese (it is PITA to convert)
Yeah, my understanding is that they both use the PACE BMS but version? release? type?
 
Not sure if it was mentioned.

When batteries work, life is good.

When batteries don’t work, you may wish that you bought batteries that you can view individual cell voltages … usually through communications or Bluetooth.


When there is no visibility to individual cell voltages, best you can do is take the misbehaving battery out and charge it at a lower top end voltage (like 3.4v per cell) and let it absorb for a couple of hours. Through trial annd error, you will need to adjust the aggregate voltage so the individual cell voltage does not trigger the common 3.65v cut off.

All pre-built batteries have welded terminals that I am aware of except for SOK. It’s not practical to dismantle, grind away and some how re-attach and replace a misbehaving cell.
 
Not sure if it was mentioned.

When batteries work, life is good.

When batteries don’t work, you may wish that you bought batteries that you can view individual cell voltages … usually through communications or Bluetooth.


When there is no visibility to individual cell voltages, best you can do is take the misbehaving battery out and charge it at a lower top end voltage (like 3.4v per cell) and let it absorb for a couple of hours. Through trial annd error, you will need to adjust the aggregate voltage so the individual cell voltage does not trigger the common 3.65v cut off.

All pre-built batteries have welded terminals that I am aware of except for SOK. It’s not practical to dismantle, grind away and some how re-attach and replace a misbehaving cell.
and that's why you need a good warranty and warranty support. (and that ain't likely Sungold)....
 
I am in a bit of a tiff with SGP over the delivery of six solar panels where two were damaged by a forklift. They offered $150.00 "compensation" for $592.00 worth of panel. I filed a complaint with SGP and also filed a complaint with FEDEX. FEDEX sent me the picture of the pallet and panels as it was just prior to loading on the FEDEX truck. (FEDEX has video cameras that point down to the lift gate area). No damage. I sent the picture to SGP and SGP's response; "Could you please let us know what compensation you need from us?" (not from Cindy but from Alisia). I will wait until FEDEX has completed whatever they do. I do hope that they (FEDEX) sends the $592.00 to SGP (which they are obligated to do) that SGP does not stiff me. The "compensation" may be a good deal on three more SGP batteries and a rack...

I may have my fingers crossed for so long it will take a pry bar to uncross them....

Once all is resolve, good or bad I will share.
 
A bit off topic …

I’m still bitter for paying 0.50/watt for solar panels through Amazon but I do appreciate their no hassle return.

Had to return over 15 packages, it took some effort to tape them and drive them to UPS, but the refund was seamless.
 
As for mixed batteries. I do have software (in development/debug/trial stage) that acts as a "Software BMS". The software collects data from each EG4 LifePower4 battery as well as each SunGoldPower battery (and a Victron Shunt). The software listens for and responds to the SGP Inverter BMS queries (there are three). The responses to the inverter represent one big battery which reflects the SOC (average) and voltage (average) of each battery in the stack. The software has the ability to fudge the data going back to the inverter to better manage battery charging.

So far trials are looking good.

One issue I am having is getting documentation for the communications between the EG4 Communications Hub (or Master EG4 LifePower4 Battery) and the EG4 LifePower4 Batteries as well as the response from the batteries. I can eventually parse/break it down but documentation ensures my assumptions are correct.

I had sent an email to Signature Solar (I purchased the Hub from then) requesting the documentation and their response; "Contact the manufacturer", I have tried that also with no response, go figure. @EG4_Jared, @SignatureSolarJess
 

diy solar

diy solar
Back
Top