Hi all,
I had a Smart Shunt and Cerbo Gx installed on my sailboat for my 3 x 300ah Relion LiFePO4 batteries. I’m finding as of late the Smart Shunt is not resetting to 100% despite reaching the appropriate parameters (at least I think so). Am I right in thinking the shunt should have reset here? Here’s some recent data with my voltage achieving 14.4v for quite a while and current dropping down. The installer has essentially stopped responding to me. Any help would be appreciated.
On the face of it, the conditions are met, but the odd thing I am seeing is the current goes below zero, so as soon as the batteries get up to 14.4V, they stop charging and start discharging, I have no idea if the synchronisation works if they are discharging. I can only take it that the BMS has stopped charging. In addition, the charging appears to have stopped well before you have recouped your energy.
Something is not right here. One cause could be cells going well out of balance so the charge is being stopped on high cell voltage well before the batteries are full. Hopefully you have Bluetooth BMS boards in your batteries so you can see what the cell balance is like and what is happening.
Another thought is the wiring on the shunt has been changed and is not correctly measuring. If you have changed anything then see this link for wiring. SmartShunt wiring
Finally, but hardly ever seen is the SmartShunt going out of calibration. There is a zero current calibration procedure in the manual.
Check your batteries first.
Relions don’t have a Bluetooth app, but I have checked each battery on many occasions with my multimeter and they’re all measuring the same voltage.
I have noted significant drift (approx 20% greater than expected) on the SoC.
Additionally, I have noticed that the batteries start discharging once they hit 14.4v or float stage, despite there being enough solar to meet demand. I’ve also noted my mppts keep showing absorption at 14.4v and not changing to float.
I just checked my settings again on VRM. Looks like the mppts were set to some “user defined” parameters with an absorption voltage of 14.4 (not my doing, and previously confirmed to be the present LiFePO4 with abs of 14.2v). That could be an issue contributing.
I am not on about the 3 batteries being unbalanced BUT the 4 cells within each individual battery, you need to open the battery yo measure those. Why companies sell batteries without communications is beyond me.
Interesting observation from a couple days ago - the shunt will reset the SoC to 100% when hooked up to shore power. The voltage rises to 14.2v and the current stays positive in single digits territory, triggering the reset. However, now, just on solar, as soon as the voltage hits 14.2 it starts discharging the batteries. The mppts show absorption mode. Never a float mode.
Screenshot your mppt settings including expert settings. The Smartshunt charged voltage is best set 0.1 to 0.2V below absorption.
I have the shunt set to absorption at 14.2v and mppts set at 14.25v, as I thought the same thing.
Same pattern happened today on solar. Mppts were in absorption for 2 hours, despite the discharge starting once 14.2v was achieved. Mppts just went to float. Once in float I manually reset SoC to 100%.
I can not see anything wrong with your settings, Suddenly going into discharge when 14.2V is reached is very odd as noted above, the MPPT should continue to supply the loads for the 2 hours and show a small positive current into the batteries. I can not explain this, other than as noted above, the battery BMS stopping charge, but it would need all 3 batteries to do the same. Or you have something else in the system that is stopping charge or a large load starting.
A new observation is that the shunt will reset when hooked to shore power, but only in the hours when the mppts are off (ie early morning when it’s still dark). Not sure if the mppts may be culprit.
Have you got an MPPT with a load output and have something on the load output that turns on when the MPPT goes to float or gets above a certain voltage or certain SOC. If you do then that would explain the sudden change from charging to discharging when nearly full, which is why the SmartShunt can not reset to 100% on solar. This could also be controlled by other means. A lot of boaters have so called solar dumps that when the solar gets the batteries almost full some energy is sent to a hot water tank (calorifier).
The thing I would say is this.
LiFePO4 is extremely efficient in terms of charging 95-99%.
LifePO4 can be considered fully charged at 3.45v per cell (99%) (4 x 3.45v = 13.8v).
The final 1% takes forever in ram it in for almost no gain at all, and in fact some say can reduce cycle life, and there is some evidence to support this from various sources (off-grid garage and willprowse have good channels that discuss and test this)
Your tailcurrent also seems high (2% of a 900Ah battery is 18A).
I personally would set the following:
Tailcurrent 1% or 1.5% (I’d set 1% to start with)
Charged Voltage set at 13.8v
I’d be happy with all the other settings.
Remember to fully charge battery 1 x per month (until the charge current drops to almost nothing maybe 1 - 2A). This helps keep your shunt in sync.
If my suggested setting changes don’t make a difference I think this could be a BMS issue as sugested by @pwfarnell, with a cell or cells triggering a BMS disconnect on 1 or more safety parameters for the BMS.
Unless you have good active balancing within the battery BMS 14.4v is (14.4 / 4) 3.6v per cell. That is very close to the maximum cell voltage of 3.65, which almost 1 cell would be hitting at 14.4v and then trigger the high cell voltage protection cutting off the charging MOSFETs.
With a 900Ah bank why not just be more conservative? Longer life, easier charging.
I’d add to my previous post that maybe set the Absorption voltage to 14.2 down slightly from 14.25
I’ll try your suggested settings. Heck, I’ll try anything.
1 Like
Downstream from the mppts is a Lynx Power In bus bar. I don’t have any load dumping or the like.
I tried your settings and no success. Once the shunt detected the target voltage the mppts went into absorption, but something is throttling the solar (must be the mppts). When this happens the solar production significantly reduces and the battery starts discharging since consumption then exceeds production. No change in sun intensity either - getting full sun nearly directly overhead here in The Bahamas.
But what do you have on the load output, it sounds like battery life algorithm keeps the load turned off until the MPPT believes the battery is charged then turns the load on which is what you are seeing, just as it hits charged the system goes into discharge and it does this before the shunt synchronises. Reduce the charged detection time, increase the tail current or reduce charged voltage to get quicker sync to 100%.
Only you know what your system set up as. Nothing else to add without more info on the system.
From the manual
3.8.3. BatteryLife
When the solar charger is not able to recharge the battery to its full capacity within one day, the result is often that the battery will continually be cycled between a ‘partially charged’ state and the ‘end of discharge’ state. This mode of operation (no regular full recharge) will destroy a lead-acid battery within weeks or months.
The BatteryLife algorithm will monitor the state of charge of the battery and, if needed, day by day slightly increase the load disconnect level (i.e. disconnect the load earlier) until the harvested solar energy is sufficient to recharge the battery to nearly the full 100%. From that point onward, the load disconnect level will be modulated so that a nearly 100% recharge is achieved about once every week.
I think I misunderstood earlier. I’m just learning all of this since my installer went MIA. I have 3 100/20 mppts with no load ouputs. I have 3 100/30 mppts and I don’t have a virtual load output on these. The 100/20 mppts are each connected to a 175w panel and each 100/30 is connected to a 360w panel (36v). The mppts seem to just be feeding the battery. Any changes to load output settings?