I'm curious if anyone else has seen this with older Hubble AM2's, specifically v1.5 BMS examples - subtle but definite bulging of the housing? The bulge isn't immediately obvious - but tapping the top cover gives a non-hollow sound indicating something (obviously the cells) is pressing against the metal. Tapping a healthy battery gives a hollow sound. I've come across two cases of older battery pairs, both from 2021 (Serial number starts with HL21...) that were upgraded to the newest firmware version when the additional batteries were added, but have the oldest BMS version (v.1.5), also showing reduced battery capacity. All indications are the battery cells are degraded (as per the BMS alarm report Hubble kindly provided as proof the batteries are being "overcharged" above 53.6V - which is interesting considering the BMS always requests 53.8V. The batteries are part of a larger bank that was expanded over time - not sure if that has anything to do with it at this stage.

Batteries have been removed for obvious safety reasons - busy with the usual displeasurable task of having to deal with the Hubble service center on the matter.

Edited May 5May 5 by AlexTZA

My inverter data shows Battery Voltage going to 54.1V. Is that a problem?

3 hours ago, frivan said:

My inverter data shows Battery Voltage going to 54.1V. Is that a problem?

I presume you have a Sunsynk or Deye inverter connected to the battery BMS? It seems to be a peculiarity of the Sunsynk to report a voltage about 0.3V higher than the actual voltage at the battery terminals even when little or no current is flowing. I have my inverter set to voltage mode to charge to 53.5V and the inverter reports 53.8V at full charge but the voltage on the battery terminals never goes above the expected 53.5V to 53.6V at most. Incidentally, Hubble support told me to manually set the charge voltage to 54-54.2V at a low current to try and sort out a cell balance issue with one battery that was causing an internal battery over voltage alarm/protect event when the voltage was only 53.4V at the terminals so there is definitely headroom in the charge voltage before an external overvoltage BMS protection would kick in. However, I have some doubts whether the BMS requested voltage is in fact suitable/correct.

Edited May 2May 2 by AlexTZA

5 hours ago, AlexTZA said:

Sunsynk or Deye inverter connected to the battery BMS

Yes. Sunsynk inverter with Solarman dongle. What port and software is used to interrogate the battery?

18 hours ago, frivan said:

Yes. Sunsynk inverter with Solarman dongle. What port and software is used to interrogate the battery?

Nothing specific. The inverter is normally connected to the CAN port of the master battery which allows the battery to communicate a few key metrics to the inverter, such as SoC, requested charge voltage/current and so on. You can see some of this info on the LiBMS screen. It is possible to read much more info through the RS232 port using something like Solar Assistant but be advised, Hubble expressely prohibits this and can use it as grounds to void your warranty except if you use their own CloudLink product (which isn't very good in my opinion).

I must add I am seeing much better results with the AM2's with the inverter set to AGM voltage mode. Here it is important to set the battery capacity correctly on the Battery screen as the portal uses this to determine the SoC based on actual energy charged/discharged which seems to work quite well. I've used 100Ah per battery although the battery states 120Ah on the label and the BMS is configured for 100Ah. Who to believe, without being able to confirm the actual cells used. I currently set the inverter target voltage to 53.5V and that normally ends up as 53.5-53.6V at the terminals which matches the voltage specs on the battery label rather than the 53.8V the BMS requested for whatever reason. At this stage, it wouldn't surprise me if that is a configuration error. I'm also left wondering if the battery's internal SoC calcs are even correct as although Hubble specifies the battery as 120Ah, from the internal battery data logs I've seen the battery BMS is configured as 100Ah so near 44V, the battery believes it has < 10% remaining yet the battery label specifies 42V as the cutoff. There are so many "loose" inconsistencies in this battery's technical specs/configuration that also makes me wonder if the cells (NMC chemistry) will get anywhere near the 3000 cycles (full 100% to 0% supposedly counts as a cycle according to Hubble's own website) claimed. The bulging batteries I've seen are between four and five years old and have not had a hard life yet everything, including the BMS logs pulled from the battery showing large variances in cell voltages, indicates the cells in these particular batteries are deteriorated. I can only hope this was caused by bad batch of cells or a BMS/firmware issue or will we soon be seeing large amounts of premature AM2 battery failures? Scary thing is right most people probably won't notice due to the lack of load shedding and I don't think I need to reiterate that bulging lithium batteries are not a safe thing to have in your home.

Edited May 2May 2 by AlexTZA