Power-Sonic provides good advice on charging LFP batteries (
https://www.power-sonic.com/wp-content/uploads/2019/04/How-to-charge-LiFePO4-batteries.pdf) and good discharge profiles (
eg https://www.power-sonic.com/wp-content/uploads/2020/07/PSL-12500-Technical-Document.pdf). I emailed them for further advice and received the same response (none) provided by other experts mentioned above.
Therefore I experimented with the Renogy 12v 50Ah LFP battery (which I bought on sale for less than either a 45Ah 345LN1 lead-acid battery or a 6.4Ah Cellink Neo LFP dashcam battery). There may be room for dual batteries in a larger ‘truck’, but not in a rav4. Therefore I replaced the 345LN1 (using a jump starter to retain ECU settings, after the vehicle had been off and the battery ventilated for a long while, being very careful to avoid sparks or short circuits during the whole process). The 50Ah LFP battery is a bit smaller than LN1, so I used some of the polystyrene packing from the Renogy shipment for a snug fit. Toyota also uses polystyrene to make the rav4 holder fit LN1. I used M8-SAE post adapters from Fullriver (which fit the Toyota battery post clamps and also provide M8 threads into the brass SAE posts). Sealed LFP batteries do not vent hydrogen or toxic fumes ( see-https://gwl-power.tumblr.com/post/141674260966/ ), so they do not need the rav4 vent tube, which I secured aside using a cable tie. I live in a mild climate, so the LFP temperature constraints (0~45ºC for charge; -20~60ºC for discharge), and the usual temperature compensations in lead acid (LA)-optimised chargers (-16 to -32mv / ºC), are fine for me.
To 70% discharge (about 13v at 25ºC with the expected low current drain) this battery should power my A129 plus duo in 1fps time-lapse parking mode (2.9W) for more than 6d. My 256 Gb SanDisk Max Endurance micro-SD card will hold about 18d of 1fps parking-mode recordings before overwriting. (Viofo claims that this unit does not overwrite any recording that includes an event that triggers its impact monitor, but that is not my experience in time-lapse parking mode). The “sweet spot” may be 2fps. The remaining 30% battery charge should hold the rav4 (at up to 120mW average commensal drain) for several months without jump starting. I have not experimentally tested these calculated times. They will vary with the efficiency of voltage converters, regulators etc. Cigarette lighter-USB converters that I tested varied from 2.8-20mA current draw (36-260mW) with no load, and 72-95 % efficiency with loads of 0.4-1.8 A. My one test with dashcam load gave battery voltage drop from 13.20v to 12.95v (~35% of battery charge at 23ºC) at 2fps parking mode (with lots of door opening = vehicle commensal drain) over 48hr.
I decided against the Viofo HK3 after reading about its poor efficiency and voltage regulation at
https://dashcamtalk.com/forum/threads/hk3-hard-wire-kit-terrible-voltage-regulation.44244/. With no load, HK3 draws 86mW before and 35mW after cut-off. It does not have suitable voltage options for LFP, I prefer thicker wires, and I don’t want to stand on my head to find the right fuse sockets then leave the fuse box cover off because of piggy-back fuse connectors. So I decided on a Victron Smart BatteryProtect, which has the advantage of more adjustable cut-off and restore voltages. It is wired after a 10A fuse near the battery, then double-insulated 1.5mm2 (AWG16) wires run to a cigarette lighter-USB converter in the console box. My dascham will not draw more than 1A through this 7m circuit, so there should be no problem with voltage drop or heating of the cable. BatteryProtect power use is very low (16mW before and 9mW after cut-off, plus 3mW if Bluetooth is enabled). Cheaper units using a continuous relay, instead of periodic sampling and a MOFSET transistor, have much higher drain. However, MOFSETS can be stressed by voltage spikes from inductive loads, and if they fail (which may be silent) they normally keep power to the controlled circuit.
Wires were fairly easily routed under the trim: forward from the auxiliary battery to the side of the NiMH traction battery, along the front base of the NiMH battery to the centre, under the carpet and felt beside the tunnel, and under the console trim to enter the console box with the dashcam cable. (Pop off the rear seat and a few trim panels, but there are no screws or bolts to undo. The tunnel was trickiest because of cross-braces, but no big deal. I used a packing strap from the Renogy battery to find the route. An electrician’s plastic guide would also work, just twist and push gently to get past any obstructions). This allowed me to use a USB Y cable for automatic parking mode as described at
Front and Rear Dashcam Installation in Rav4 Hybrid....
After the battery change, I conducted tests like the one described above, at several starting voltages:
Starting at 12.95v battery voltage (estimated 25% SOC), it was a bit scary to watch 52.3A at 14.14v flow into the battery. While many LFP batteries (including the Renogy 50Ah) are specified for charging up to 1C (50A in this case), and the battery got barely warm to touch at the side after 30min, I would not like the charging current to go any higher. I assume the fast charge is simply a result of the low resistance presented by the LFP battery. The rav4 does have a (Hall effect) current sensor on the battery, but how this is used does not seem to be made public. Some car alternators might struggle with 50A continuously for 30 min, but the rav4 hybrid DC-DC converter seems fine with it. Toyota hybrid vehicle DC-DC converters are typically ≥100A (
Please explain 12V system to me--how to keep battery charged while using accessories?), and the rav4 12v circuit fusible link is 120-140A (
pure sinewave inverter for AC power), so there should be enough headroom for all other current draws. The ICE ran more frequently while the current into the auxiliary battery was high (for about 3min every 15min while current was 40-50A). The charge voltage dropped quite suddenly to 13.6v when the current was below 1A (at 50-60min in the charts above). This is almost no charge into the LFP battery (and the recommended current for charge termination), so I turned off the vehicle at 60-65min, at which time the battery voltage was 13.5v.
Then I kept watching battery voltage and current flow. In the first experiment, the commensal current draw by the rav4 dropped to about 100mA within 10min, about 15mA within 20min, and below my measurement limit of 5-10mA within 3hr (when battery voltage was 13.34v). In the second experiment, the commensal current draw by the rav4 was about 450mA after 10min, about 100mA after 20min, and below my measurement limit of 5-10mA within 30min (when battery voltage was 13.38v). The times to reduced commensal drain varied somewhat between experiments, for reasons I do not understand. But the stepped levels of commensal drain from the LFP battery were the same as those from the OEM LA battery, within my measurement errors. In essence, commensal drain takes the top off the fully-charged battery voltage when the car is stopped, then drops to a very low level until a door is opened.
On these results, the rav4 is fine with an LFP battery that is kept above 25% SOC. At 14.2v and ≤1C during the bulk/CC phase, there is no danger of damage to a properly top-balanced battery through overcharging, and the vehicle does exploit the fast-charging capability of LFP chemistry. I might be happier at 0.2C for longest cycle life (by reducing the possibility of irreversible plating of lithium metal onto the anode), but dascham users who need to recharge the battery during short drives will appreciate the faster charge near 1C (if the battery manufacturer allows that). The ‘automatic’ drop in charging voltage at 50-60min (presumably triggered by the charging current sensor reporting ~1A) is reassuring. Any accessories that drain current while the vehicle is running should be wired with return/ground connection downstream of this charging current sensor, so as not to interfere with it. Trickle charging of LFP batteries should be avoided, but the rav4 ‘float/trickle’ voltage of 13.6v does not raise any concerns for durations of a few hours. For long drives, a lower float voltage (≤13.2v), or no charge, would be better for LFP life. There were no problems with fusible links, sensors, relays or fuses. No errors have been displayed. The vehicle entered ‘READY’ mode as usual from the LFP battery at 25-95% SOC. If anything, fuel economy should improve due to the light weight and efficient charging of LFP, but this may be hard to see among other driving variations. The rav4 at 14.2v charged LFP to an estimated 90-95% SOC at 23ºC, which is fine for most purposes and good for battery life. Those who want 100% SOC (or more ‘absorption/CV’ time at >14v for cell balancing or erasing any memory effects) will have to use something supplementary, like a good LFP wall charger.
I have a Victron Blue-Smart IP65 12v/10A wall charger that can be programmed for the optimal charging specs from Renogy, except that it lacks the ability to set absorption charge termination based on charge current (try Enerdrive chargers for this feature). For charge termination based on current, a Victron user would need to watch charge current by Bluetooth (reported to 0.1A), and turn off the charger at the wall when it reached the desired level (1A = 0.02C for 50Ah). Charge termination based on
cell current additionally requires a BMS with this output (which is not in the Renogy 50Ah) and an interface to the charger, which is lacking in everything I own. The Victron charger has split leads that can permanently attach to the LFP battery (as in the photo) and it includes the ability to report charging status and history via a smartphone. In default Lithium mode (14.2v charge, 2h absorption) it added only 0.2+0.4 = 0.6Ah to the charge from the rav4. At 14.6v (with 1 hr absorption for full charging and cell balancing) it added 10.3+0.2 = 10.5Ah (much more than I expected). The Blue-Smart IP65 can also maintain an LA battery (now my spare 345LN1, which at 14.4v charge it reported as 3.7+11.1 = 14.8Ah ie 33% undercharged by the rav4, perhaps a reflection of the low charging efficiency of LA chemistry). At the 14.7v charge recommended for lead-calcium batteries, it added another 0.0+0.2 = 0.2Ah.
![165810]( "165810")
The caveat is that the initial charging current in the rav4 rose to over 73A when the LFP battery was discharged further, to around 12.7v (~20% SOC at 23ºC). The unmodified rav4 is not suited for this task, as the applied current at >1C is above the level that will quickly damage the battery. Toyota implies that using other power-consuming parts of the vehicle (headlights, aircon and blower, rear window defogger, radio) can reduce charging current to the battery (
Do I have to take special steps to protect the battery of my hybrid vehicle if I plan to store the vehicle for an extended period?). In my experience the maximum effect is a few Amps, which is not enough to avoid battery damage in this situation.
As the rav4 has a current sensor on the auxiliary battery, it would be ideal if Toyota could program the vehicle for a maximum auxiliary battery charging current (10-50A for 0.2-1C charging of a 50Ah LFP battery), and at 1A either terminate charging or ‘float’ at 13.2v (which will stop any current flow into the LFP battery). Maintaining supply to the vehicle 12v system; while accurately sensing, controlling and terminating 12v battery charge current; may require separation of 12v battery charging and load circuits. This is simple enough to design, and already at least partly implemented, but the available circuit diagrams are somewhat ambiguous (
Please explain 12V system to me--how to keep battery charged while using accessories?). It would not be trivial to implement after manufacture. The “float” approach seems easiest, as it only requires appropriate programming, and 13.2v is fine for both the LFP battery and the remainder of the nominal 12v electrical system. But it leaves the challenge of providing an optimal LFP charge current below the potential demand of other 12v electrical components while the vehicle is running.
The rav4 current sensor is also said to include a temperature sensor, but I can not establish how it is used in the hybrid model. Charge voltage to an LFP battery should
not be altered according to temperature, but at the usual compensation of ~-27mv/ºC for a 12v battery it is not a big deal at normal cabin temperatures. Most ‘smart’ battery chargers simply disable this feature for LFP charging. Maybe LFP compatibility will be included by design in future rav4s?
In the meantime, if an LFP battery might be drained below 25% SOC, the best approach seems to be to limit charge current by interposing a DC-DC charger. This can ensure that the charge profile is optimal for LFP. But these are one-way devices, so they cannot provide power from the battery through the charging cable to the vehicle electrical system, as in the OEM rav4 wiring. At least an added cable with a power diode will be needed. This will need input from an automotive electrician. On preliminary advice, it is feasible, but expensive.
