Welcome to denMike's tiny page!
Home
Guestbook
Contact
 
Husky
  Boattent

  Repairs

  Navigation information

  Svanemøllehavnen

  Vision LiFePO4

 
Omnia opskrifter
 
3D Printing
Electronics
Programming
 
Corona virus
Fantasy
mtech.dk
Space
Tilbud
Various Links
 
USA 2001
ISU MSS03
USA 2006
South Africa 2008


Last updated: October 16th, 2023
Vision Iron-V LiFePO4

SOC Issues
So I purchased a set of two Vision Iron-V LiFePO4 100 Ah batteries for my boat. Apparently it was too early, as the software running on the BMS and the software communicating with the BMS on your phone are not exactly polished.

The first problem is to find out which app to use. To save you the effort, here are the links for the IOS and Android versions:

IOS: https://apps.apple.com/ca/app/ll-bms-monitor/id1624879621
Android: https://play.google.com/store/apps/details?id=com.BLUEDC.OSAppCode&hl=en_US

The Android version does not seem to have been updated much, and the only updates for the IOS version seems to be bluetooth related - for good reasons, as some of the versions didn't work on iphone 14, in June 2023 (when the phone was 9 months old...).

Another problem is that the "set bluetooth name" feature hardly works. After you press ok the first time, then you need to wait a few seconds, and then press ok again, before the new name is actually saved.

The next problem is that if you have two batteries in parallel then it seems that the SOC of the batteries is not computed nor reset properly, so after a little while, then one battery is at e.g. 75% while the other states 100%, even though the voltages of both batteries and all cells indicates that both batteries are fully charged, here is an example:

   

Why does this happen? Because the current measurements of the BMS are not particularly accurate, and they are done with much too slow frequency, so apparently the BMS thinks that significantly more current has been discharged than charged into the battery. A reasonable design would have been to reset SOC to SOH when the voltage was high enough / when cell balacing has finished.

Please note that the cells in these batteries are actually very well balanced, and the efficiency of a discharge / charge cycle is as good as expected for LiFePO4 - about 5% loss according to my Nasa BM1 battery monitor.

How do you reset the SOC? This guy has found that on some other batteries the SOC is only reset when an over voltage condition is triggered on either battery or cell level.

It turns out that something similar is going on with the Vision batteries. According to the datasheet they should be charged with 14.0 to 14.6 V. On two different boats we have charged them with 14.4 V and 14.6 V respectively. It turns out that SOC is not reset until the battery voltage has reached 14.7 V, so you actually need to charge the batteries to a HIGHER voltage than the datasheet specifies! At least the over-voltage protection has been set to not kick in until a battery voltage of 14.8 V, but that equals a cell voltage of 3.7 V, which is a bit high in my oppinion.

So, does it matter that the SOC is reported incorrectly? Good question, there are two options:

  1. If the BMS only uses voltages to trigger turning off at "fully discharged", and the SOC is only a number displayed by the app, then no, it does not matter. I have a battery monitor which shows how many Ah have been discharged from the battery (although it is from my old AGM installation the Ah measurement works fine, but the "remaining battery capacity in percentage" obviously does not work, but I can live with that).

  2. If the BMS however turns off when SOC reaches 0%, then yes it is a problem, as the BMS in the example in the screenshots above will then end up turning off when there is still 100-75 = 25% capacity left!
Luckily it seems the first option is the case.

For instance for these SOK batteries the SOC is not used for turning the battery off: http://support.sokbattery.com/bluetooth-battery-how-to-calibrate-soc-and-actual-capacity-readings/

The SOK battery does have a SOC calibration process as described in the above link, but it requires the battery to become fully discharged, something that is 1. very inconvienient if you're on a long trip, and 2. even for LiFePO4 not really good for the cells.

So these are the tests that we have done:

  1. Try to "overcharge" and see if an over-voltage error resets SOC. Done, result: It resets SOC to 100% at a battery voltage of 14.7V.

  2. Try to fully discharge, and see if the battery shuts down when SOC reaches 0 or if it either synchronizes with the other battery in the end, or just continues discharging until the cell and/or battery voltages indicates fully discharged. Done, result: The battery does not use SOC to shut down, it only looks a the cell and battery voltages.
So in the end everything works fine. It is just a shame that Vision does not have proper documentation nor customer support.

Installation Setup
Regarding my installation I have:

  1. Brought new Vision Iron-V LiFePO4 batteries.
  2. Brought a Victron Orion DC-DC converter to limit the charge current in order to protect the alternator.
  3. Removed the old battery bank seperator, as the Orion does this.
  4. Moved the alternator voltage sense wire from the auxiliary batteries to the starter battery (this wire could also just be connected directly at the alternator, as the battery bank seperator has been removed).
  5. Installed the Orion and starter battery on each side of an existing switch, so the Orion does not discharge the starter battery when the boat is not in use.
  6. Installed a remote on/off wire to the Orion and (through a relay) D+ on the alternator, so it starts charging when the alternator is running (ie the engine is running).
  7. Brought and installed a switchable 70A fuse in front of the Orion so I can turn it off if needed.
  8. Reused the existing Philippi 32A lead acid charger, which charges with 14,4 V / 13,8 V, when I
  9. removed the temperature compensation sensor for the charger, so it always charges with 14,4 V / 13,8 V.
  10. Kept the old Nasa BM-1 battery monitor. It does not show the correct remaining battery percentage, but the current flowing to/from the battery and the accumulated Ah works perfectly fine, and that is all I need. A new Nasa battery monitor for LiFePO4 is too expensive to make sense to me.
  11. Kept the cabeling installed as (ok, but not optimal):
         --+---+
           +---+--
All in all it is a bit of work re-routing the cables and adding new cables for the Orion, but the cost of the extra equipment (aside from the batteries) is not staggering. Others have made calculations that show that even with the higher price of LiFePO4 batteries compared to AGM, LiFePO4 batteries are still the cheapest option in the long run, as their life expectancy is significantly longer (more cycles and longer lifetime; at least 10 years, and likely in the order of 20 years). Add to that that the effective capacity is essentially doubled as AGM batteries should not be discharged below 50% (even for "deep-cycle" AGM batteries deep discharge cycles should not happen too often).

Note that when reusing an existing charger, that is not prepared to work with LiFePO4, then it will most likely NOT start charging if the batteries have entered the low-voltage protection mode. The reason being that the batteries will have zero volts over the terminals, so the charger thinks that no battery is connected, and hence won't start. There is an easy fix for this: Turn on the charger, and take a wire and short circuit the 12V terminal from your starter battery with the 12V terminals of the LiFePO4 batteries. The batteries will now see a voltage and re-connect, and the charger will see a voltage and start charging. Remove the wire. A fraction of a second should be enough. I am not entirely sure what happens first (the battery or charger connecting), but it doesn't really matter in which order it happens.

You could, of course, also just start your engine and let the Orion initiate the charging process.