Two years after installation, my IP22 charger lost its user defined configuration after a power interruption and reverted to a lead-acid charge profile - which would have destroyed my LiFePO4 batteries had the BMS not gone into protect mode. The charger ran an output of 14.4 volts for 8 hours, and 13.8 volts for another 8 hours.
Is there are replaceable memory battery inside the charger?
Not a good look and I hope you find a fix or it never occurs again.
I doubt that 14.4 and 13.8V for 8 hours or even a day or two would have harmed either our 9 year old 300Ah Sinopoly or 1 year old 280Ah EVE LiFePO4 batteries. No internal BMS involved.
Searching around here, I see this same failure mode has happened to others. The story ends with the charger losing its user defined settings every time it has to do its job after the power fails. If the charger now decides to go into recondition mode (over 16 volts), all that stands between it and my lithium batteries going up in smoke are the BMS protections. If there isn't an internal memory-retention battery that I can replace so I can trust it again, it's going in the trash.
The charger was performing perfectly up till now. My configuration was set to charge to 14.4 volts to permit cell rebalancing, and when the charge tail current tapered to 7.5 amps: terminate charging. Properly charging LiFePO4 batteries isn't only a matter of controlling the terminal voltage, it also requires terminating the charge when the charge current has fallen to 0.03C. Just holding the battery at 14.4 volts for hours on end after the battery is fully charged is abusive treatment. Technically, you can overcharge a 4S pack with as little as 13.6 volts (3.4V/cell) if it's maintained for an excessive amount of time. As just happened here.
The damage is done, all I can do to prevent another abusive event from causing further damage - is cross my fingers there isn't another power interruption, and travel 7,000 miles to the site before the charger executes a factory default 7-day reabsorption charge. The timebomb is ticking on my batteries.
This is why the reverted-to factory default settings will damage my battery: LiFePO4 cells simply don’t really charge at voltages up to 3.3V (13.2V pack voltage) and then fully charge if kept for a prolonged period at 3.4V (13.6V pack voltage) and upwards. The transition is so abrupt that claiming to control the charging process by adjusting the voltage alone is purely and simply bound to fail. You can get away with overcharging for a while, with each charging cycle causing only some electrochemical damage, until one day the lithium dendrites short a cell or the cells swell beyond usability.
Charging at reduced voltages, down to 3.4V/cell, only increases the absorption time and therefore the overall charging time, but achieves strictly nothing in terms of preventing the battery from getting fully charged and then overcharged. You need to terminate charging when the charge current tapers off and before the charge current reaches zero. The tail current charge cutoff function of the charger is necessary to achieve that goal. The BMS won't do that for you. It's only there to prevent a catastrophic failure from overvoltage events.
There's an excellent explanation of LiFePO4 charging here: https://nordkyndesign.com/charging-marine-lithium-battery-banks/
The charger lasted two years in service. Maybe that's the best I can expect. I'm considering replacing it with a Phoenix Smart Ip43 Charger. At least I can access it remotely via VRM. Hopefully, that won't lose its mind after two years in service.
A bit more: The graph below, excerpted from https://nordkyndesign.com/charging-marine-lithium-battery-banks/, shows that charging LiFePO4 batteries at any current below 0.027C (I have 250 AH of batteries, so that's 0.027 X 250 = 6.75 amps) at 3.6V/cell will overcharge the batteries. Holding the voltage of a 4S pack at 14.4 volts (3.6 V/cell) for 8 hours at near zero charge current, followed by 8 more hours at 13.8 volts, is absolutely damaging to the batteries. It is the combination of voltage and cutoff current that determines when to stop charging.
The Victron predefined settings for "Li-ion" disable the tail current charge cutoff (termination limit - as shown above), so the user is forced to use "user defined" settings. The predefined settings for "Li-ion" also force a repeated absorption every 7 days, which is completely unnecessary for lithium batteries (it's a lead-acid battery requirement to prevent sulfation). So using the predefined setting clobbers the battery every time it recharges and, just for good measure, whacks it again once a week. Not good.
It's sad because Victron could provide a really good predefined setting for LiFePO4 batteries. It's just software. First, change the name. "Li-ion" describes a whole range of battery chemistries. Then, if a smart shunt is installed, take charging current data from that. Allow the users to define the capacity (AH) of the battery and calculate the charge termination value according to the correct algorithm (see link above). If a smart shunt isn't installed, use the charger output current instead and the worst thing that can happen is the battery is slightly undercharged at termination - unlike lead-acid batteries - that will do no harm. Disable repeated absorption (and temperature compensation - which is already disabled). And now you have a real LiFePO4 charge setting. Not the hybrid of lead-acid settings that exist now in the "Li-ion" predefined settings.
Follow up 2 months later. After traveling 7,000 miles to physically access the chargers, I found that both of my Blue Smart IP22 chargers will intermittently revert to the "normal" factory charge profile after an AC power interruption. "Normal" means lead-acid battery settings. My custom settings are not lost under these conditions, but they become inactive (deselected). Since the whole purpose of the chargers is to recover from a power interruption, they are unusable with my lithium batteries, and I was forced to take them out of service. The reversion to factory profile appears to occur more frequently when the power interruption is short, indicating to me that the fault has something to do with power-up initialization circuitry or algorithms, i.e., a design flaw.
No answer at all from Victron…as always…
Welocome private equity…more and more products show significant design flaws…
DC2DC charger output only 27A and not 30A and get very hot for 300Euro and nothing is smart, no ve direct connect, no current reduction…nothing that a renogy 30A for 99Euro cannot do either and that delivers 32A…
The Phönix smart charger gets to 70 degrees Celsius in normal operation mode confirmed by Victron as I wanted to return it…cannot be healthy for semiconductors inside. And well cables in boats if not in an engine room have a 60 degrees Celsius rating from factory.
The 30/50 100 MPPT remote is an always on means if your remote cable gets disconnected the MPPT is on and BMS cannot switch it off anymore in Desaster Modus…
In conclusion, unless Victron corrects the reversion-to-factory-default-settings problem, I advise against using the Blue Smart IP22 chargers with LiFePO4 batteries. I substituted an IP 43 12 | 50 charger, which I run at one-half output current (25 amps), and it can be manually configured to support correct settings for LiFePO4 batteries and has behaved correctly for 6 months now. I can also access that charger remotely with VRM, so if a lost configuration ever occurs, I have a chance to correct it from 7,000 miles away.
Charging a 4 series cell (12.8 volts nominal) LiFePO4 battery to 13.8 volts (3.45 volts per cell) without a tail current cutoff will overcharge and damage your cells. Any charge voltage over 3.375 volts per cell will eventually overcharge the cells.
Victron never contacted me in any way about the issue.
I'm an electrical engineer who has done circuit design for 40 years. Designing a proper circuit to retain user settings and apply them reliably after a power down is essential. This wasn't accomplished here. As an aside, a model of the JK BMS has just been discovered to have a similar settings retention problem that poses a serious safety hazard. You can see more about that here and here.
In the meantime our 6 year old Victron Blue Smart IP22 30/15A battery charger continues to operate perfectly with no such hiccups - ever.
It still maintains the original manual settings. Sure I have tweaked them slightly over the years.
14.4V or equalisation mode at the terminals of our 9 year old 300Ah 4 cell LiFePO4 battery with no internal BMS could (hopefully) never occur for more than a few seconds without a disconnect. That indeed has never occurred during our full-time RV travels for at least the last 8 or so years.
In the early days for about a year the direct alternator 75-80A output would force the battery terminals to 14.4V for several hours while driving. Now dropped to 14.1V by the series installed Victron Argodiode battery isolator.
Adverse effects?
Well the battery still delivered the equivalent of 299Ah at the last November annual C/10 capacity test. A mere c5% loss since the first test 9 years ago. It also starts our 3.9l turbo diesel engine often multiple times a day.
I'm not too concerned.
