EG4 Lifepower4 V1 SOC doesn’t track Victron Smartshunt

[DVD]

I’m running a 6 pack of LiFePower4 V1 into an EG4 8k using SA along with a Victron SmartShunt as referee.

With each initially reporting 100%, the SOC readings begin to diverge significantly. As the EG4 (and idiot lights on Lifepower4) readings drop to single digits the SmartShunt is still reporting over 20%. Battery voltage readings support the SmartShunt readings. I only noticed this recently as my normal SOC range has been >50% nd I wasn’t concerned. Poor solar in the last few weeks has periodically dropped my SOC into the alarm range on the batteries.

Does anyone have a similar experience/solution? I haven’t connected BMS Test yet. I’ll try contacting SigSolar later if there’s no clues from the forum.

 

[chrisski]

I have a different brand Inverter and despite three built in shunts, I have wildly inaccurate SOC readings on my Outback Radian. I have the Victron Shunt on a different built, and I find it spot on.

On my Outback Radian, I would trust the Victron Shunt for average SOC if it was attached way more than Outback Shunt for average chargebuilt into the system.

Does the EG4 have anything like charge efficiency that can be updated? Also, not all batteries will be same SOC, is it possible that the EG4 gives you the lowest battery?

For my individual batteries, I have a display on each battery I check.

 

[DVD]

I’ll have to connect the BMS Test. I’m a little suspicious it is what you said, EG4 reporting the lowest battery SOC. Next week…..

 

[marionw]

I’ll have to connect the BMS Test. I’m a little suspicious it is what you said, EG4 reporting the lowest battery SOC. Next week…..

The "Master" EG4 LifePower4 V1 reports the average of all batteries SOC in the stack to the Inverter. (I have six EG4 LifePower4 V1 and a Communications Hub).
The Master BMS also reports total capacity of the battery bank/stack for all batteries the Master BMS can communicate with and whether the batteries Discharge MOSFETS are closed/on. If a battery is no longer available due to the Discharge MOSFETS being turned off (Low SOC, low battery Voltage or some other fault) the Master BMS will only report total capacity of the others and will only average SOC for the batteries that can still provide power. The Master BMS also reports any alarm conditions for all batteries. What the inverter does with all the info the Master BMS provides is another undocumented mystery.

The EG4 LifePower4 V1 RJ45 ports are wired: Pins 1&2 are an RS485 bus running at 9600 baud and used for software monitoring (BMS_Test or Solar Assistant) or BMS to Inverter Closed loop communications, can only do one at a time. Pins 7&8 are a second RS485 bus that is used for Master BMS to other battery communications and runs at 19200 baud. Pins 7&8 can be used for the BMS_Test software but only if you do not have a "Master BMS" (address 0). Not sure what inverter you are using but you might find the addition of an EG4 Communications Hub to be useful. The Hub acts as a MASTER BMS (address 0) with all batteries starting from address 1. The Hud has a selectable BMS to Inverter communications protocol which can be set to match your inverter (Both RS485 and CAN). Using the Hub will free up the Pins 1&2 RS485 bus for use by the BMS_Test software of Solar Assistant as the Hub now provides the "Closed Loop" communications. Just check with Signature Solar as I do not know what inverter you have.

You need to get the BMS_Test software hooked up so you can monitor all the batteries. This is an excellent tool that does not have to be connected full time but only periodically when you need to check the batteries. The same connection between the BMS and your computer will work with Solar Assistant which will display each battery in your stack.

Like most server rack batteries they need to be fully charged before being placed into service. LifePo4 batteries are not plug-n-play, they do need some special attention from time time. How often is dependent on your charge and discharge cycles.

Use the BMS_Test software to monitor the batteries during charging to ensure no battery has an "runner" cells, one (or more) that charge faster than the others and will eventually reach the Cell Over Voltage Protection setpoint (3.60 volts) which will cause the BMS to shut down further charging. As cells reach 3.45 volts the BMS will start balancing and will reset SOC to 100%, this most likely won't happen if a battery has any "runner" cells.

Also keep in mind the Victron shut being a great tool also needs to have it's SOC reference periodically reset to be accurate.
In addition the Victron shut can only measure current (in both directions). If you pull 10kw from the batteries then it will take more than 10kw charging to replace what you took out. The Victron shunt can measure this BUT once 10kw has been pushed back into the batteries they are not at the same SOC as when the discharge started (conversion and other inefficiencies). This is exacerbated if a BMS shuts down the charge for a battery and the remaining batteries charge to the Victron shunts "Charged voltage" setpoint. Victron shunt synchronization is an automatic process and will occur when the battery has been fully charged. The battery monitor will look at a few parameters to ascertain that the battery has been fully charged. It will consider the battery to be fully charged when the voltage has reached a certain value and the current has dropped below a certain value for a certain amount of time from 5.7.1. Automatic synchronization of the user manual. But if one or more batteries charge was stopped due to a Cell OVP then the battery bank/stack is not fully charged and the Victron shunt may synchronization SOC with incorrect data.

Check these two threads if you want to make your own communications cables rather that order them:

M

Post in thread 'Help with Pace BMS Comms'

Dec 29, 2023

First, do you have the pinout of each RS485 port.
The EG4 Lifepower4 batteries for example use pins 1&2 as an RS485 bus as well as pins 7&8 as a second RS485 bus in the same RJ45 jack.
I purchased a RS485 to USB Converter:
RS485-USB-Converter

Using a spare cat5 cable, just cut off the RJ45 connector on one end, remove about 3 inches of the outer insulation then you can connect the correct pair of wires to the RS485 to USB Converter and plug the RJ45 connector on the other end to the correct port on the battery.
Look at the RJ45 connector on the cable (With my tired old eye I use a...

• marionw

• 

M

Post in thread 'query on SRNE reset times.'

Feb 6, 2024

I use a spare ethernet patch cable, cut the RJ45 connector off one end, strip two to three inches of the outer insulation and select the pair of wires I need depending on which pins of the Inverter or BMS I need to use.

For a normal cable pins 1 & 2 are the orange/white and orange. Pins 7 & 8 are the brown/white and brown. I even at times use several of the wire pairs in the same cable.

Holding the RJ45 connector with the pins up and facing you (the latch/lock is facing away from you) pin 1 will be on the left and pin 8 on the right.

I use these as you can never tell which two pins in a...

• marionw

• 

 

[EPicTony]

The smart shunt is as correct as you can get.

 

[Lardino]

you can turn your Smart Shunt into a bms with this project.

GitHub - sijones/DiyBatteryBMS: Integration of Victron to Pylontech supported inverters with DIY Batteries

Integration of Victron to Pylontech supported inverters with DIY Batteries - sijones/DiyBatteryBMS

github.com

I don't have EG4 but have pace bms and know that the SOC can get way out of whack if the batteries aren't fully charged for a few days.
Also my bms doesn't count small currents of about half and amp or less. So that can add up .
I am just now getting my Smart Shunt running the above project ready to use as a bms .

 

[Cmiller]

Solid info in the comments already, but just as an FYI, I worked on one system that had batteries with (I believe) the same BMS as the EG4s, and we ran into issues with the BMSs reporting high/low SOC% to the inverters. Essentially whenever the batteries were discharging, the lowest SOC% was reported, and when charging, it was the highest SOC%. It showed up because of using Time-of use settings, which caused massive power swings when reaching a target SOC%, as the inverters would swing between charge/discharge, trying to hit the target, but constantly getting higher/lower feedback... It was just a back and forth cycle at times.

Again, those were not EG4 batteries! But I believe the same BMS. We did get a firmware update later that I believe would have fixed it, but we ended up swapping out the battery bank for a different model, and larger overall bank (3,000 AH, 48V), therefore we never did get to test out the newer firmware.

 

[marionw]

The BMS used in the EG4 V1 is certainly a bit different than most BMS.
Many BMS such as a PACE or JBD have a "Max Charge Voltage" or "Full Charge Voltage" parameter that when reached will cause the BMS to reset SOC to 100% (I assume there might be other factors as well). The Narada BMS in the EG4 V1 has no such parameter so what parameters does the BMS use to determine when to reset SOC to 100%.

In my experience (I have six V1 in a rack) the BMS seems to reset SOC to 100% when a number of cells reach 3.45 volts (default Balance Start voltage) and the Max Differential Voltage (default is 20mv) exceeds 20mv and the BMS Balance circuit is activated. Even at this point in the charge the batteries are not fully charged so a 100% SOC is really bogus. In cases where the Max Differential Voltage does not even get to 20mv but cells do charge to at least 3.45 (BMS Balance circuit will not get activated) and the charge continues the BMS eventually resets SOC to 100%.

From time to time I use my ChargeVerter to get as many cells as possible charged to at least 3.5 volts without triggering a Cell Over Voltage Protection. The BMS resets SOC to 100% and at this point for me the SOC is as accurate as you will ever get with the V1. When you get the batteries to an accurate 100% SOC then you need to stop the charge, let the battery rest for a bit then reset/resync the Victron Smart Shunts 100% SOC ("Synchronize SOC to 100%" button). Resetting/Resyncing the SmartShunts 100% SOC may be required more than you would expect depending on your charge/discharge cycles. Some folks do this as often as once a week.

If you are at 100% SOC and pull 5000 watts from the battery you will need to push more than 5000 watts back into the battery to get the battery back to where you started. Therefore the shunt will read more charging current flowing into the battery than what it measured as discharge current flowing out of the battery to accurately calculate SOC. It is not an amp for amp situation.

The shunt needs to be properly set up. In particular the "Battery capacity" needs to be set to reflect your battery bank/stack and "Charge efficiency factor" which defaults to 95% might have to be tweaked. Also keep in mind, if you get the battery charged well above to "high or 80% knee" and "Synchronize SOC to 100%" rather than let the battery settle for a bit before "Synchronize SOC to 100%" then the shunt SOC will be off.

 

[DVD]

The BMS used in the EG4 V1 is certainly a bit different than most BMS.
Many BMS such as a PACE or JBD have a "Max Charge Voltage" or "Full Charge Voltage" parameter that when reached will cause the BMS to reset SOC to 100% (I assume there might be other factors as well). The Narada BMS in the EG4 V1 has no such parameter so what parameters does the BMS use to determine when to reset SOC to 100%.

In my experience (I have six V1 in a rack) the BMS seems to reset SOC to 100% when a number of cells reach 3.45 volts (default Balance Start voltage) and the Max Differential Voltage (default is 20mv) exceeds 20mv and the BMS Balance circuit is activated. Even at this point in the charge the batteries are not fully charged so a 100% SOC is really bogus. In cases where the Max Differential Voltage does not even get to 20mv but cells do charge to at least 3.45 (BMS Balance circuit will not get activated) and the charge continues the BMS eventually resets SOC to 100%.

From time to time I use my ChargeVerter to get as many cells as possible charged to at least 3.5 volts without triggering a Cell Over Voltage Protection. The BMS resets SOC to 100% and at this point for me the SOC is as accurate as you will ever get with the V1. When you get the batteries to an accurate 100% SOC then you need to stop the charge, let the battery rest for a bit then reset/resync the Victron Smart Shunts 100% SOC ("Synchronize SOC to 100%" button). Resetting/Resyncing the SmartShunts 100% SOC may be required more than you would expect depending on your charge/discharge cycles. Some folks do this as often as once a week.

If you are at 100% SOC and pull 5000 watts from the battery you will need to push more than 5000 watts back into the battery to get the battery back to where you started. Therefore the shunt will read more charging current flowing into the battery than what it measured as discharge current flowing out of the battery to accurately calculate SOC. It is not an amp for amp situation.

The shunt needs to be properly set up. In particular the "Battery capacity" needs to be set to reflect your battery bank/stack and "Charge efficiency factor" which defaults to 95% might have to be tweaked. Also keep in mind, if you get the battery charged well above to "high or 80% knee" and "Synchronize SOC to 100%" rather than let the battery settle for a bit before "Synchronize SOC to 100%" then the shunt SOC will be off.

Good info. I bought a 16S active balancer to get my two recently purchased Eco-Worthy rack batteries top balanced.
It sounds from above that if I top balance the LFP4 v1’s I should see a better SOC reading compared to the SmartShunt. It turns out that I’m relocating the v1’s to another rack anyway, so inserting a top balancing step during the process won’t be that disruptive.

??? Any thoughts about this?

 

[Lardino]

Solid info in the comments already, but just as an FYI, I worked on one system that had batteries with (I believe) the same BMS as the EG4s, and we ran into issues with the BMSs reporting high/low SOC% to the inverters. Essentially whenever the batteries were discharging, the lowest SOC% was reported, and when charging, it was the highest SOC%. It showed up because of using Time-of use settings, which caused massive power swings when reaching a target SOC%, as the inverters would swing between charge/discharge, trying to hit the target, but constantly getting higher/lower feedback... It was just a back and forth cycle at times.

Again, those were not EG4 batteries! But I believe the same BMS. We did get a firmware update later that I believe would have fixed it, but we ended up swapping out the battery bank for a different model, and larger overall bank (3,000 AH, 48V), therefore we never did get to test out the newer firmware.

My EG46000xp inverter using Pace bms does the same thing with repeated charging cycles. I have been trying to figure out why it does that . You have given me some ideas as to why. I don't understand why the EG46000XP doesn't have a clock or some way to only charge to setpoint once a day instead of going up and down like it often does in closed loop.