LTO (Lithium Titanate Oxide) – The Ultimate Battery for Dash-Cam parking mode (DIY)

[EricSan]

A faulty soldering gun just proves that the right tools make a big difference: they make the job easier and improve its quality. I've encountered this many time.

I've learned this lesson over and over again. If you're going to do ANY job, just spend the money and get the proper tools. They really do make a difference. Do the job correctly the first time and you won't have to spend more time and money later on to re-do that same job.

A nice benefit of having moved the eyelet is that the battery remains future-flexible. I can restore the original battery configuration simply by adding a jumper across the busbar that I severed.

On a different note, a close examination of the battery pack reveals that all of the cells are in excellent physical condition - the outer perimeter of the battery pack is straight and true with no bulging along any of the sides or edges.

 

[EricSan]

Well, I ordered a JK BMS with attached wires:

...and the one that showed up today looks a bit different. I even selected one from a vendor that advertised as "USA Seller" figuring that if someone here in the states had inventory, that would increase the likelihood that they had a pile of the "older" ones, not the newer ones without the wires. Turns out, the USA seller took my order and my money and forwarded the order back to China for shipping. Sigh...

Oh well, time to start planning out my "auxiliary box" that will get strapped to the top of my battery box.

 

[GPak]

It's unfortunate when you order a part, wait a long time for it to arrive, just to discover they sent the wrong one.
Hopefully, JK-BMS has improved the BT range; otherwise, this BMS is just as good as the one with the soldered wires.

Also, the plastic 'auxilary box' shouldn't block the BT signal as much as the metal one.
When I tested mine, the BT range doubled with the top cover of the metal case removed, just enough for in-car use.

 

[EricSan]

Well, on the bright side, I don't have to worry about the pre-attached BMS wires being too short and not reaching the battery terminal from inside the auxiliary box. I think maybe I'll connect the wires and see what kind of heat the charger board generates with different charge rates. That will be my guide for the final choice of plastic vs aluminum for the extra box.

 

[EricSan]

A bit off topic from my new LTO battery construction, but the gas furnace for the first floor of the house recently stopped working. Usually, replacing the glow igniter or the flame sensor gets it back up and running again, but not this time. So, the next logical place to look is at the control board. After thoroughly photographing everything, I removed all of the wire harnesses and pulled the board from the furnace. Here is what I found on the back of the board. Looks like a solder joint that didn't receive enough solder during manufacturing (25 years ago), cracked with age, and recently has been arcing and spitting out hot bits of solder. The burned landing receives the leg of a 120v 20A relay that turns on and off the air circulation fan.

A few passes with a brass brush removed all of the burned bits and cleaned up the landing pad. Judging by the missing volume, it's clearly been spewing burned solder for a while.

A quick pass with the soldering iron and some new solder made it look and work like new again! And just in time, too. Daytime high temps are supposed to drop to the low 30s (0c) by the end of the week!

 

[EricSan]

With all of the necessary pieces in hand and a furnace that works again, I started building my new parking battery.

First up, I wired up a set of speaker connection terminals to use for constant power and the ACC signal to toggle Drive Mode and Parking Mode. I bent the tabs down, snipped off the long ends and soldered it together.

Then I mounted the BMS and Charger boards in the box. Since the Charger Board gets the hottest, I mounted it to the top of the ABS plastic box, figuring the top of the box would have access to free air for some cooling. I figured if it ran too how, I'd cut a hole in the lid and install a heatsink to help with cooling. The ABS plastic is good until about 100c and early tests with a 10A charge rate, I'm only seeing a +14c rise in temp over ambient, so I'm thinking I won't need to worry about heat. I had a pair of 5-way binding posts in my parts box, so I added these as grounding lugs for the hardwire kits.

The BMS board sits in the bottom of the plastic project box.

Here it is all wired together and charging. The fuseholders are 10g silicone wire that is SUPER flexible - highly recommended! The heavy black wire is leftover from when I did the hardwire to my car's start battery, which is really nice wire, but is rather stiff by comparison. I just ordered some additional 10g silicone wire so I can redo the heavy black wiring. This will stress the connections less as I close the box.

And a few more shots of the two-box setup:

Some extra-wide terminal rings fit the battery main terminals perfectly:

And the protective plastic clips snap right back into place after the wires are attached:

Here is the top of the auxiliary box. I painted the black spring terminals yellow for the ACC wire and added a 3A slow-blow fuse for the cameras.

First impressions are good: the Charger Board seems pretty efficient and exhibits a fairly small temp rise even at a 10A charge rate. The batteries, though, take quite a bit of time to charge when compared to my 6S battery 😉

 

[EricSan]

They've added a ton of new features to the newer BMS. I'm not sure what all of these are, so I need to spend a little bit of time learning. The BMS I bought last time around came with an instruction sheet and links to some online docs. This newer BMS came with no instructions at all.

Here is the entire Status Menu:

One super cool feature is the "Time Enter Sleep" function. It displays in units of seconds, but you can only set it in units of hours. Pretty funny. I was SUPER excited to see that it works to put the BMS to sleep after 1hr (it's lowest setting). The LED ring light around the power switch turns off and the BMS does indeed shut down! But, it does NOT wake up when you reapply an input voltage (like when the car starts). Total bummer! They are half way there, at least it can put itself to sleep... sigh...

 

[EricSan]

Here are the "Settings" menu options. Some of the screen caps overlap features, so there is a little bit of redundancy.

 

[EricSan]

And here is the third menu, for Control features:

 

[EricSan]

And the firmware version of the new BMS:

 

[EricSan]

Here are some images of my newly completed LTO parking battery. All of the battery terminal screws have blue thread locker applied and are taped over with Kapton tape to help prevent things from coming loose and rattling around inside. There are too many available amps inside the box to risk having anything moving around inside. I added a protective sheath to the BMS balance wires so they don't get snagged inside the trunk. The main battery terminals accept a 5mm hex wrench, so they can be made good and tight.

Here is the auxiliary box before it gets closed up. I added some thermal grease between the charger board and the top cover of the plastic box before fully securing the mounting screws. I taped one of the temp probes from the BMS to the edge of the Charger Board where it meets the plastic case. The location shows a higher temperature while charging than the top of the inductor on the board. The second temp probe is taped to the top of the battery pack, under the top cover. Everything is fused, so this should keep the fireworks to a minimum if something goes wrong. I used a 20A fuse between the XT60 input and the charger board, and a 15A fuse between the charger and the battery. The inline automotive fuse holders are 10g wire. The black wires are 11g and left over from adding the direct-to-battery wire for charging my original parking battery. It had a double sheath on it that made it VERY stiff, so I peeled the outer layer and left the inner layer of insulation. This resulted in making the wire far more flexible and easier to close up the box. The connection points for the dashcams are wired so that when the BMS is shut down, no power flows to the outputs. The only thing wired directly to the battery is the voltage meter.

The overall wiring diagram follows the schematic that I posted earlier in this thread, except that I did not build in the 12v to 5v regulator for the dashcam hardwire kits:
https://dashcamtalk.com/forum/threads/lto-lithium-titanate-oxide-–-the-ultimate-battery-for-dash-cam-parking-mode-diy.50484/post-643986

And here is the completed pair of boxes. The fuse holder is recycled from an old amplifier that I took apart and red & black 5-way binding posts that serve as the grounding lugs for the dashcam hardwire kits have been in my parts box for more than a decade - finally got to use them! The auxiliary box is attached to the top of the battery box with two strips of Velcro. In the trunk, another set of Velcro straps will hold things in place.

Grand total for this 12v, 4P6S, 80Ah, 1104Wh parking battery is about $350.

$205 for the battery pack, including delivery
$10 for the auxiliary project box
$55 for the charger board
$43 for the BMS
$15 for the voltage meter
$7 for some crimp ring terminals
$8 for the speaker terminal clip

For now, I set the charge rate to 10A, or about 145w of input power. It looks like the BMS starts ramping down the charge rate as the battery pack charge level surpasses about 14.5v. It seems to ramp down smoothly and predictably and eventually the charge rate went to 0A, so I'm assuming I have the output level of the charger board set appropriately for the battery pack. It looks like the cells took a full charge up to 2.65v (15.85v total). After resting for an hour or two, the voltage level had fallen to about 15.7v - so it looks like the batteries are in pretty decent shape given that they are used/recycled.

Time to do some testing to see how long the battery actually runs and calibrate the charge/discharge readings in the BMS app.

EDIT: Here is an image of my new LTO battery installed in the trunk of my car.

 

[GPak]

....
One super cool feature is the "Time Enter Sleep" function. It displays in units of seconds, but you can only set it in units of hours. Pretty funny. I was SUPER excited to see that it works to put the BMS to sleep after 1hr (it's lowest setting). The LED ring light around the power switch turns off and the BMS does indeed shut down! But, it does NOT wake up when you reapply an input voltage (like when the car starts). Total bummer! They are half way there, at least it can put itself to sleep... sigh...

My "Smart Sleep" function is also not reliable to waking up with ignition, sometimes it works sometimes it doesn't.
Try increasing the "Vol. Smart Sleep (V)" parameter to 2.65V so that it is active across the entire voltage range.
Also, based on testing my BMS, the minimum detectable current draw for "Smart Sleep" is 0.57A to remain awake.
If the current draw is less than 0.57A, the BMS enters "sleep" mode regardless of the load.
In my case, two Mini 2s consume about 4.5W in LBR mode, which equates to a current draw of approximately 0.28-0.39 A, so the BMS goes into "Sleep" mode even when the dashcam is in LBR mode.
Therefore, when using "Smart Sleep", I set it to 6 hours, according to my work schedule. This essentially replaces the dashcam's parking time setting.
However, currently I have it disabled, because it doesn't always wake up with the ignition.
Perhaps we should contact JK-BMS about this.

Here are some images of my newly completed LTO parking battery....
.....And here is the completed pair of boxes....
.....Grand total for this 12v 4P6S 80Ah 1104Wh parking battery is about $350...

Awesome, it looks nice and clean, as if it was just assembled at the factory!!

... It looks like the BMS starts ramping down the charge rate as the battery pack charge level surpasses about 14.5v. It seems to ramp down smoothly and predictably and eventually the charge rate went to 0A, so I'm assuming I have the output level of the charger board set appropriately for the battery pack. It looks like the cells took a full charge up to 2.65v (15.85v total).

The "Vol.Cell RCV (V) setting is the BMS request for Charger's CV setting, your screenshot shows 2.68V for cell which is 16.08V total for Charger's CV setting.
I think this will delay a ramping down a little.
I set mine to 2.67V and 16.02V accordingly.

Here is AI response:
"Vol.Cell RCV" is a setting in a JK-BMS that stands for Voltage per cell, Requested Charge Voltage.
This is the target absorption voltage that the BMS requests the charger to maintain, which is essential for a 100% charge.

 

[EricSan]

My "Smart Sleep" function is also not reliable to waking up with ignition, sometimes it works sometimes it doesn't

Bummer about the smart sleep not actually being very "smart" or useful... I had good hopes for newer hardware/firmware to make this work better. I suppose our use case is sufficiently distinct from their target audience that it wont suit our needs. Good to know about the minimum current for the BMS to recognize a power draw.

Awesome, it looks nice and clean, as if it was just assembled at the factory!!

Thanks - I was trying to keep it neat and tidy. I was originally planning on running the BMS balance wires straight down through both boxes and not having them show at all, but I figured that would complicate future maintenance, so I just bundled them together. I really wanted just to have two wires, one for each battery terminal. I think I'll cover the screws with some tape in the event the plastic caps get broken. The fuses on the wires won't protect exposed battery terminals.

I like the idea of inviting the JK people to the forum. Perhaps they can help us decipher some of the functionality and we can suggest some improvements.

 

[EricSan]

Here are some of the descriptions of JK BMS Parameters that I found with Google AI. Guess I need to experiment a bit more:

A "par-limiter" (parallel limiter) for a Battery Management System (BMS) is a function that regulates or limits the current flow between parallel-connected battery banks to prevent damage and ensure balanced charging. This prevents a situation where one battery bank, which may be at a different state of charge, overcharges the other, potentially damaging the cells.

How it works:

• Prevents overcurrent: When batteries are connected in parallel, especially if they have different charge levels, a large current can surge from the more charged to the less charged battery.
• Regulates current: A parallel limiter activates when it detects a high current flow between banks and reduces it to a safer, lower level.
• Protects the BMS and cells: By controlling the current, the limiter prevents the BMS from disconnecting the charger and protects the battery cells from damage caused by excessive charging or discharging rates.
• Utilizes PWM controllers: In some systems, a PWM (Pulse Width Modulation) controller is used to rapidly switch the power supply on and off, thereby regulating the current flow.

---------------------------------
The "SCP delay" in a battery management system (BMS), which stands for Self-Control Protector, is the time delay before a short circuit protection (SCP) feature trips and shuts down the system. This delay is adjustable in some BMS units, like the JK BMS, and is used to prevent the system from shutting down during momentary high current spikes, such as the inrush current when a large inverter is turned on. Increasing this delay, for example from 1500μs (1500 mu s or 1500𝜇𝑠) to 20,000μs (20,000 mu s or 20,000𝜇𝑠) (0.02 seconds), can solve problems where a high-power appliance trips the protection.

---------------------------------
RCV (Requested Charge Voltage) time in a battery management system (BMS) is a configurable timer that starts after the battery pack voltage reaches a set RCV voltage, indicating it's fully charged. The timer's purpose is to hold the battery at this full voltage for a specific period, allowing for final cell balancing and ensuring stability before the BMS moves to a different charging or resting state.

Purpose and function

• Cell balancing: Holding the voltage for the RCV time allows the BMS to perform final balancing on cells that may have been slightly out of sync.
• Stable state: It ensures the battery pack reaches a stable, full state before the system transitions to a float charge or other modes.
• Preventing rapid cycling: A sufficient RCV time prevents the BMS from rapidly cycling between protection and recovery modes due to minor voltage sags or fluctuations right after the pack is charged.
• State of Charge (SOC) reset: It is crucial for resetting the SOC to 100% in some BMS systems, as it confirms the battery has reached its full capacity after being held at a high voltage for a set duration.

How it works with other settings

• Inverter communication: When the BMS communicates with an inverter, it requests the inverter to maintain the RCV voltage. The RCV time timer starts once the inverter actually holds this voltage.
• Open-loop systems: In systems without communication, the inverter's own settings for charge voltage and timer become the primary control. The BMS RCV time is then a failsafe to ensure full charge if the inverter's timer is too short.
• Rebulk State of Charge: Some modern systems have a "rebulk" function that monitors the SOC. After the RCV timer expires and a full charge is achieved, the battery might drop to a float voltage. If the SOC drops by a set percentage during this float phase, the system can initiate a "rebulk" to return to the absorption voltage to top off the battery, making the RCV time an important first step in this process.

Practical considerations

• Choosing a time: A longer RCV time, like 1 hour, is often recommended to allow for better balancing, especially in systems without inverter communication. Setting a very short time can be problematic.
• Voltage calibration: Differences in voltage calibration between the BMS and the inverter can cause issues. To compensate, it is often advised to set the BMS RCV trip voltage slightly lower than the inverter's absorb voltage setting.

 

[EricSan]

I had to break this up into multiple posts - the forum has a 9999 character limit per post. Who knew??

RFV (Request Float Voltage) time is a setting on a battery management system (BMS) that determines how long the BMS will stay in the float charging stage before returning to the bulk charging stage. This duration is set to provide a sufficient amount of time for the battery to stay at a steady float voltage to finish charging, with a recommended setting of 6 to 8 hours.

How it works

• Float Voltage: After the battery is charged to its full capacity, the system enters the float voltage stage, where it maintains a lower, steady voltage to keep the battery topped off.
• RFV Time: This setting defines how long the BMS will keep the battery at this float voltage.
• Daily Recharge: After the RFV time expires, the BMS will return to the bulk charging stage the next day, ensuring the battery is fully recharged.

Why it's important

• Full Recharge: It ensures the battery stays at the correct float voltage for a sufficient period, which is important for a full charge.
• Battery Health: By allowing the battery to remain at a steady voltage, it helps in balancing the cells and maintaining the battery's health and lifespan.

---------------------------------

A "dry alarm" for a battery management system (BMS) is a signal from the BMS to an external device, acting as a switch that sends an alert when a specific condition is met, such as a low voltage or high temperature. "Dry" means the relay provides no power itself and simply closes or opens a circuit, allowing for integration with external monitoring systems like Building Management Systems (BMS) or other control panels.

How it works

• Dry Contact: The BMS uses a "dry contact" output. This is a set of terminals that are not connected to a power source, functioning like a simple switch.
• Closed Circuit: When the BMS detects a specific event (e.g., a low voltage), it "closes" the circuit between its two dry contact terminals, connecting them together.
• External Monitoring: An external device, like a PLC, alarm panel, or Building Management System (BMS), receives the signal because the completed circuit allows it to detect the closed state.
• Notification: Based on the closed circuit signal, the external system can trigger a local audible/visual alarm, send a remote notification, or initiate another action.

Common applications

• Public Safety Systems: Used to comply with regulations like the NFPA® 72 & 1221 codes for public safety communications (BDA and DAS networks) by providing critical alarm closures for DC UPS systems.
• Building Management Systems (BMS): Allows for remote monitoring and control of battery banks integrated with a facility's overall management system through interfaces like SNMP or Modbus.
• General Monitoring: Provides a flexible way to monitor battery status and connect to a wide range of external devices, including lights, buzzers, or automated shutdown sequences.
• Preventive Maintenance: Can be used to proactively alert users to issues like low electrolyte levels or to remind them to replace a battery before it becomes a problem.

Key features

• Flexibility: Dry contact alarms are versatile and can be integrated with many different types of external hardware.
• Safety: They are often used in safety-critical applications because they are reliable and can be configured for 24/7 monitoring.
• Integration: Modern BMSs often include features for remote monitoring, data logging, and reporting, in addition to the dry contact alarm output.

 

[EricSan]

It seems that the larger 25A charger behaves a little differently than the smaller 20A one as exhibited by its higher temperature. Here is a comparison of the two charger board temps from the JK App.

Here are the temps from my new, larger battery:
MOS Temp is the temp of the BMS itself.
Battery T1 probe is attached to the edge of the charger board.
Battery T2 probe is attached to the top of the battery, under the removable lid.
The Charger board is elevated by 6c above room temperature. I wonder how much power it's burning... The new BMS indicates a power consumption of 0.8w.

Here are the temps from my original, smaller battery:
MOS Temp is the temp of the BMS itself.
Battery T1 taped to crevice between two batteries
Battery T2 taped to the edge of the charger board, same metal panel shared with BMS
It seems that I reversed the T1 and T2 probes between my two battery packs - oops...

I presume the difference is that my smaller battery has an anti-backflow diode between the Charger and the BMS, so it prevents the charger board from receiving any power from the battery. Without the diode in the larger battery, the charger seems to be drawing some power from the battery.

 

[EricSan]

The "Vol.Cell RCV (V) setting is the BMS request for Charger's CV setting, your screenshot shows 2.68V for cell which is 16.08V total for Charger's CV setting.
I think this will delay a ramping down a little.
I set mine to 2.67V and 16.02V accordingly.

2.68v is the default setting for LTO batteries in the BMS. I didn’t change it. I wonder if this means the adjustable voltage output setting on the charger is less critical? Lots more to learn with this newer BMS…

 

[GPak]

The default 2.68V for the "RCV(V)" setting is a appropriate value for the default 2.7V for the "OVP(V)" setting.
Consequently, the charger's constant output voltage (CV) should be adjusted to 2.68V * 6 = 16.08V or slightly higher.

I set mine to 2.67V because I also limited "OVP(V)" to 2.67V.
And I adjusted charger's CV to 2.67V*6=16.02V.
I prefer to increase the "RCV(V)" value as close to the "OVP" as the BMS allows, to delay the onset of charging current reduction, (for our application we almost never charge the battery for several hours continuously).

P.S.
As you know, I later replaced this latest BMS with an older model due to the limited Bluetooth range.
The older model does not have the "RCV (V)" setting, however I maintained my original OVP setting and the charger's CV output.

 

[GPak]

It seems that the larger 25A charger behaves a little differently than the smaller 20A one as exhibited by its higher temperature. Here is a comparison of the two charger board temps from the JK App.

Here are the temps from my new, larger battery:
MOS Temp is the temp of the BMS itself.
Battery T1 probe is attached to the edge of the charger board.
Battery T2 probe is attached to the top of the battery, under the removable lid.
The Charger board is elevated by 6c above room temperature. I wonder how much power it's burning... The new BMS indicates a power consumption of 0.8w.
View attachment 88310

Here are the temps from my original, smaller battery:
MOS Temp is the temp of the BMS itself.
Battery T1 taped to crevice between two batteries
Battery T2 taped to the edge of the charger board, same metal panel shared with BMS
It seems that I reversed the T1 and T2 probes between my two battery packs - oops...
View attachment 88311

I presume the difference is that my smaller battery has an anti-backflow diode between the Charger and the BMS, so it prevents the charger board from receiving any power from the battery. Without the diode in the larger battery, the charger seems to be drawing some power from the battery.

Maybe I'm missing something, but for the 25A charger, I'm seeing a battery voltage of 15.42V, a current draw of -0.23A, and a corresponding power of 3.5W, not 0.8W?
I don't understand this relatively high current draw of -0.23A if nothing is connected to the battery's output terminals?

I have the same 25A chargers installed on both 6S and 5S batteries without any diodes, and I don't see any reverse current or power draw by the charger itself.

Here's the 6S battery in my Jeep, which has been parked in my garage for almost four days now, and all three temperature sensors are showing roughly the garage temperature.

And here's my 5S battery in my home office, and all three temperature sensors are showing roughly my room temperature.

 

[EricSan]

I made a few current measurements with my DMM in the battery loop so I can get the most accurate readings of the self-discharge rate of my completed batteries.

The older JK BMS + 20A charger board in my smaller battery pack together consume 0.73w (47mA at 15.14v) with the Active Balance turned OFF and 0.88w (58mA at 15.14v) with Active Balance turned ON. Overall, not too bad, but it does represent constant power draw as long as the BMS is turned on.

The newer JK BMS + 25A charger board in my larger battery pack together consume a bit less at 0.70w (47mA at 14.88v) regardless of whether Active Balance is on or off. Thus, the newer BMS is slightly more efficient in two different ways - a small, but nice, efficiency improvement with the newer BMS.

The digital voltmeter from Supnova consumes 0.14w (9.51mA at 14.88v) when the display is on and continues to consume 0.03w (1.96mA at 14.88v) AS LONG AS IT IS CONNECTED. For this reason, I moved the wire connection to the "input" side of the BMS so when the BMS is turned off, there is zero current draw on the battery.

Maybe I'm missing something, but for the 25A charger, I'm seeing a battery voltage of 15.42V, a current draw of -0.23A, and a corresponding power of 3.5W, not 0.8W?
I don't understand this relatively high current draw of -0.23A if nothing is connected to the battery's output terminals?

Ah, that's my fault - the power draw you are seeing is with one of my dashcams attached. I was attempting to calibrate the current draw and have been unable to do so. How did you manage to calibrate the power draw figure? It won't seem to allow me to make any changes at all, it keeps telling me "send failure" when I click on OK after changing the Calibrating Current to anything at all. I am also unable to enter a negative sign - I tried changing the current draw to both a positive and negative number. Neither worked. Any insights??