diy solar

diy solar

SMA Sunny Island 6048 and Lifepo4 batteries (open loop)

marcvl

New Member
Joined
Jul 25, 2022
Messages
10
I'm struggling to get 2 48V Ampere Time Lifepo4 batteries to work with my Sunny Island 6048 and hoping to get some advice on how to program the Sunny Island correctly. This is what I tried:

1. Configure the batteries as Lithium. This doesn't work because it requires communication between the SI and the battery BMS. The Ampere Time batteries are a black box. No BMS communication possible.

2. Configure the batteries as VRLA with battery voltage set at 48V and battery capacity at 200Ah. The inverter starts correctly and everything seems to work fine, except that the SOC readings are off. I did a test with a 2.5 kW load on a fully charged battery (10kWh) and the inverter shut down after 75 minutes when the SOC reached 0%. The batteries were far from drained. The voltage on the batteries was around 53V, which is about 70% charged. That's in line with what one would expect given the load.

3. Configure the batteries as 'Other', with following parameters:
- BattCapNom: 200Ah
- Batt Charge Voltage Manual: 55.2V
- Batt Charge Current Max: 150A
- Batt Di Voltage: 43.2V
- Batt Di Voltage St: 49.6 V
- Batt Di Ch Current Max: 150 A
When starting the system, I'm getting an error on the SMA: F141 Malfunction InvVolLow

Any recommendations on how to fix this? It seems to me that the best solution would be to configure the batteries as 'other', so I can set the disconnect voltage correctly. That setting is not available when you configure the batteries as VRLA. However, I haven't figured out how to avoid the F141 error...
 
Best solution is to get a compatible BMS and use the SI in lithium mode and retro fit that to the cells in your battery.

You can adjust the VRLA settings for SOC disconnect etc to match the actual performance of your battery, just make sure the battery will disconnect from the Sunny Island if a cell goes out if voltage range.

Adjusting the VRLA settings is easier than setting up the “other” settings.

If you plan on keeping this power system for the life of the inverter (10+ years), a compatible BMS is a wise investment. There are many compatible BMS’s.
 
Best solution is to get a compatible BMS and use the SI in lithium mode and retro fit that to the cells in your battery.

You can adjust the VRLA settings for SOC disconnect etc to match the actual performance of your battery, just make sure the battery will disconnect from the Sunny Island if a cell goes out if voltage range.

Adjusting the VRLA settings is easier than setting up the “other” settings.

If you plan on keeping this power system for the life of the inverter (10+ years), a compatible BMS is a wise investment. There are many compatible BMS’s.
I was trying to adjust the VRLA settings, but there is actually not much you can adjust, other than nominal voltage and battery capacity. Those are set in accordance with the battery specs, but don't result in reliable SOC measurement as mentioned above. Is there anything else I can tweak? Or should I trick the system into believing that the battery capacity is higher than 200Ah? That feels iffy to me.
 
I found only a list of compatible batteries, not compatible BMS's.
Do you have/know such a list published somewhere?

I have personally used:
REC
Batrium
Zeva
I have helped others with:
Seplos
Emus

I’m sure there are others but i don’t have first hand experience with them.
 
I was trying to adjust the VRLA settings, but there is actually not much you can adjust, other than nominal voltage and battery capacity. Those are set in accordance with the battery specs, but don't result in reliable SOC measurement as mentioned above. Is there anything else I can tweak? Or should I trick the system into believing that the battery capacity is higher than 200Ah? That feels iffy to me.

I’m not sure if the US version is different to the 8.0H version i use, if i go to expert mode i can change all the VRLA settings.
 
I have personally used:
REC
Batrium
Zeva
I have helped others with:
Seplos
Emus

I’m sure there are others but i don’t have first hand experience with them.
I intend to use my SI6048-US-10 with lithium batteries.
They are on the boat from china.
I know they have a SEPLOS BMS which can "speak" many protocols including pylontech.
But when you look at this list:

There seems to be a big difference between US islands & non-US islands.
Page 5 for US model SI
Pylontech listed in table for European model 8.0H, not in table for 6048US.
I have some time before they arrive, I hope I am able to let them talk to each other.
Would make life a lot easier.
 
I intend to use my SI6048-US-10 with lithium batteries.
They are on the boat from china.
I know they have a SEPLOS BMS which can "speak" many protocols including pylontech.
But when you look at this list:

There seems to be a big difference between US islands & non-US islands.
Page 5 for US model SI
Pylontech listed in table for European model 8.0H, not in table for 6048US.
I have some time before they arrive, I hope I am able to let them talk to each other.
Would make life a lot easier.

I’m sorry i haven’t any first hand experience with the 6048US. It surprises me that the comms are so different.
 
Does anyone have experience configuring lithium batteries using the “other” battery type on the SI? That seems to me to be the best way to set all parameters manually, but I can’t get this to work. The SI inverter refuses to start with error F141 - Malfunction inverter voltage low. I dont understand this error. Battery voltage is 55V, so definitely not low.
 
Has anyone tried reaching out to SMA's tech support regarding these issues? Lately they've been totally MIA but they might have a tech note in their library.
 
I suggest trying precharge through a resistor.
"voltage low" must mean voltage on input makes an excursion low, since 55V is in the middle.
Maybe BMS chokes on the inrush or start-up current.

Try isolating SI (+) input from battery. Turn on battery. Connect some kind of load so battery believes it should be working (e.g. 120V 100W incandescent lamp, or 120V space heater (thermostat on high, switches on). That should cause battery to power its output.

Connect SI (+) input with a resistor. SI draws 4W at idle, 72 mA from 55V, 750 ohms. So use a resistor 100x smaller, 7 ohms. Precharge resistor will dissipate 40 mW powering SI, but starts with a peak of 400W before capacitor is charged. I suggest a 120V space heater (later, smaller resistor sized to withstand precharge impulse will do.) Turn on SI main breaker and wait for voltage across precharge resistor to taper off to about 0.5V

This should charge up the caps. Then hard-connect SI (+) to battery. Then push button to start up SI.

It is possible there is a transformer inrush when SI starts producing AC. That would usually be hit or miss, although it is possible this inverter has consistent turn off and turn on timing.
 
1/ SMA tech support is of no use. They will either not reply at all, or it takes them 1 month to reply with a message like “sorry we can’t help. These batteries are not supported”. Terrible.

2/ I don’t think this is a inrush issue. I installed an easy start system from battleborn to avoid exactly this scenario.
 
"Other" is not an option for battery type with my model SI and whatever firmware version I'm using.

Try a DC load on the battery, in case that tricks it into behaving differently. e.g. light bulb or space heater.

Do you have 48V of lead-acid batteries, e.g. four car batteries, you could try, make sure the SI functions properly?

Another idea would be tell SI the battery is VRLA. People have used them that way. I think it may need to be something other than 48V nominal (24 cells) to have voltage range match LiFePO4. SI can be set 42V to 52V nominal.
 
1/ SMA tech support is of no use. They will either not reply at all, or it takes them 1 month to reply with a message like “sorry we can’t help. These batteries are not supported”. Terrible.
Thanks for confirming that. Something awful is going on with SMA. Their support was never good pretty now it's totally MIA.

I've been waiting a month to get a straight answer about which firmware version is required for the Sunny Boy to talk to their cellular modem. These are two new in the box SMA products purchased directly from an authorized wholesaler and SMA is refusing to support them.

Our national wholesaler is dropping them because of this, and we will never do another SMA product either. We've got 4 projects to finish up with SMA gear but we're two far along to change inverters now, trust me, we've discussed it.
 
I have 90KWh of lithium on my 4 SIs running VRLA now for 9 months off grid mode, but they automatically connect to grid when low SOC happens... Works perfectly except the low end of SOC can get off sync from reality and couple days ago the SI drained them to 0 instead of connecting to the grid so I need to switch to Litium mode and get the SOC to the SIs some frickin how..

Toms, retrofitting all my 17 batteries with new BMSs would be very costly and a major pita, plus I don't hear much good about those BMSs... Hope to get my victron smartshunt to give the overall system SOC to the SI... I've seen one guy here that has done it but he did not use the SI for charging, so the implementation is incomplete... errr.




here are a few settings I use, took many hours to determine..

Battery type VRLA
Nominal battery voltage 48
Battery boost charge time 60min
Cell charge nominal voltage for boost charge 2.33
Automatic equalization charge no
Cycle time equalization charge 15552000
Battery equalization charge time 1hr
Cell charge nominal voltage for equalization charge 2.33
Cycle time full charge 2592000
Battery full charge time 1hr
Cell charge nominal voltage for full charging 2.33
Max. battery charging current 325A
Battery temperature compensation 0
Cell charge nominal voltage for float charge 2.29




 
I thought I tried to figure out what high/low settings of VRLA would fit LiFePO4, and the limits of settings didn't line up for 48V nominal.
I you told SI it was a 50V nominal VRLA battery (or maybe 46V nominal), can you get it so as voltage goes down lower knee of the curve, SI realizes it has lost track of SOC and connects to grid, or at least shuts off?

I assume all battery charge and discharge currents are known to SI.
Maybe it cycles too much without getting reset in the upper knee.
 
I have 90KWh of lithium on my 4 SIs running VRLA now for 9 months off grid mode, but they automatically connect to grid when low SOC happens... Works perfectly except the low end of SOC can get off sync from reality and couple days ago the SI drained them to 0 instead of connecting to the grid so I need to switch to Litium mode and get the SOC to the SIs some frickin how..

Toms, retrofitting all my 17 batteries with new BMSs would be very costly and a major pita, plus I don't hear much good about those BMSs... Hope to get my victron smartshunt to give the overall system SOC to the SI... I've seen one guy here that has done it but he did not use the SI for charging, so the implementation is incomplete... errr.




here are a few settings I use, took many hours to determine..

Battery type VRLA
Nominal battery voltage 48
Battery boost charge time 60min
Cell charge nominal voltage for boost charge 2.33
Automatic equalization charge no
Cycle time equalization charge 15552000
Battery equalization charge time 1hr
Cell charge nominal voltage for equalization charge 2.33
Cycle time full charge 2592000
Battery full charge time 1hr
Cell charge nominal voltage for full charging 2.33
Max. battery charging current 325A
Battery temperature compensation 0
Cell charge nominal voltage for float charge 2.29




I'm going to try 4p16s with one REC standalone (Toms suggestion). One respected retailer says it should be fine. So you might get buy with fewer REC BMS slave units plus a master? Maye 2p16s? Or 3p16s? Depending on quality of the cells.

I am buying five of the standalone BMSs though...as they can be flashed into slave modules from what I understand. These are backup, or if things go south with that many cells, and I'm going 208v with three sunny boys and will be poised to add a multicluster box if needed with another three Sunny Islands plus another 4p16s battery. And can add on more as required.

Take absolutely every single thing I am saying with a salt mine as I'm in discovery phase and I know less than John Snow.
 
Last edited:
So prevent the SI from shutting down, couldn't we lie to it about the AH capacity of the battery? I have 1120AH battery, couldn't' I just tell it the battery was much much larger? It might use SOH to adjust but calculated SOC shouldn't get down to 20%. Or I suppose I could set all the shut down menu items to 0% SOC??? just brain storming. In an ideal world, it would be nice if the Victron smart shunt would talk to the SI.
 
SI will figure out battery is drained at some voltage.
I have an old 100 Ah AGM bank, testing capacity with 600W load I got something like 40 Ah or maybe 80% of 40 Ah = 32 Ah before SI shut down after some 2.5 hours +/-.

If you are using LiFePO4, should be able to set voltages and nominal number of "lead acid" cells so it operates to voltages part way past the knees.
 
Maybe there's some combination of settings that will let the system run forever, but I've tried many things and no matter the SI eventually gets off sync from the actual SOC. It's impossible to use the lower part of my battery safely without a potential shutdown or my batteries drained to 0.

Madsci1016 could get the Victron smart shunt to work fully. I've tried to reach out to him but no response. If someone is able to do the coding and get it working I would be glad to pitch in or buy the code. I may build a new bank and use a SI supported BMS. One battery should mostly reflect the over all SOC, at least better than spoofing the SI. Why SMA won't support lithium directly, stinks.
 
Back
Top