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

Interesting thermal behavior, thanks for sharing your data! Cooler temps are always nicer when dealing with current handling! There seems to be a sweet spot in there in the context of current vs thermal rise (thermal efficiency vs your reported electrical efficiency). Dividing thermal rise by current flow, we get some interesting ratios from your test data (smaller numbers represent greater efficiency):

8A = 5.25 c/A
10A = 4.80 c/A
12A = 4.42 c/A
16A = 4.31 c/A
20A = 4.50 c/A

So, it looks like there is a sweet spot in terms of thermal rise/current somewhere between 12A and 18A. Using thermal paste to couple the back of the charger to the inside chassis wall will likely result in lower temps for the charger board as the heat is spread over a larger surface area. Thermal behavior at 8A to 10A (typical car charging rate?) seems like a nice improvement over the original, smaller charger board. Not really a surprise given the more robust construction of the newer charger, though.

In the context of my own LTO battery box, I was not a very good scientist: I made three different changes all at once, so I'm not sure which one of these changes is responsible for my lack of current backflow. Ooops....

Here are the changes I made:
1) Switched from utility outlet charging to direct from battery charging
2) Added a time delay relay so the LTO doesn't draw power until about 8s after starting the car
3) Added an anti-backflow diode between the charger board and the battery pack

With these three changes in place, I'm not seeing any current backflow, even with a fully charged parking battery at 16.0v. Since my anti-backflow diode was exhibiting extremely high temperatures (over 100c) while charging from my SMPS power supply, I tried an experiment: When I bypassed the diode with a wire jumper, there was still no current backflow. Looking at the BMS log file, the time-based "current limiter" did not activatie, either.

From the perspective that fewer parts is likely to result in better reliability over time, I'm thinking I'll just remove the anti-backflow diode and leave the rest alone. The disappointing part is that I'm not sure which change led to the elimination of the current backflow. It was one of the following: a) switching from the utility outlet to direct to battery charging, b) adding the time delay relay, or c) the combination of both.

Regardless, I now have two different cars where a 16v LTO parking battery is wired directly to the car battery through a time delay relay. Neither parking battery backflows current anymore.
 
Last edited:
If we subtract the 25°C initial ambient temperature from the overall temperature, the trend in thermal efficiency will be more in line with electrical efficiency.
8A = 2.13 °C/A
10A = 2.3 °C/A
12A = 2.33 °C/A
16A = 2.75 °C/A
20A = 3.25 °C/A

.........
I think it was the time delay and the diode that helped solve the backflow problem.
The diode is self-explanatory because its main purpose is to block backflow, and it always works.
The time delay helps by cutting off the initial phase of starting the car when the alternator is not yet providing stable power, but the time delay will not help if the ignition is on, but for some reason we delayed the engine start beyond the time delay.
The diode is not expansive and is very efficient, I think it is better to have both, the time delay and the diode.
 
GPak said:
If we subtract the 25°C initial ambient temperature from the overall temperature
Click to expand...
Duh, I missed this step 🙄 . Makes more sense now that both efficiencies track together.

I have two more diodes arriving shortly. I’ll compare to see how things play out. Probably makes sense to use both a belt and suspenders 😉.
 
I've been tracking voltage for the LTO battery pack in my son's car almost two weeks now that it is charging consistently. I've just about got it dialed in to an "equilibrium" point where the power draw on a weekly basis matches the weekly charging rate from driving. I'm guessing at this point, it's best to draw down the battery pack so that the typical charge level floats back and forth closer to its "nominal" voltage (about 13.8v) rather than keeping the battery pack closer to it's maximum 16.0v charge level.

Is this a good strategy for long term battery health? Or should I just keep the battery fully charged?
 
Today, I received two diodes that I ordered a little while ago. One was the 80v 50A little black rectangular board that @GPak already tested (in post #195), so I skipped testing this one. The other diode is this 60A dual device purple board.

Best Diode.jpg

https://www.aliexpress.us/item/3256805499983692.html

An interesting feature of this ideal diode board is that there is no ground connection, there are just two connection points: voltage in and voltage out.

This little guy is a champ:
With an 8A charge rate, the temp rise is only 4c over ambient.
A 10A charge rate only produced a 6-7c rise over ambient.
At 11A, thermal rise was only 9c! At this point, the board was passing more than 150w of power.

Measuring the difference of Vin to Vout showed the board never dropped more than 0.01v.

I haven't tested backflow performance in my car yet, but I'm impressed with its behavior so far!
 
Last edited:
wow this is a really impressive diode i like the fact that no grounding is required and i definitely like its very high efficiency.
 
EricSan said:
I've been tracking voltage for the LTO battery pack in my son's car almost two weeks now that it is charging consistently. I've just about got it dialed in to an "equilibrium" point where the power draw on a weekly basis matches the weekly charging rate from driving. I'm guessing at this point, it's best to draw down the battery pack so that the typical charge level floats back and forth closer to its "nominal" voltage (about 13.8v) rather than keeping the battery pack closer to it's maximum 16.0v charge level.

Is this a good strategy for long term battery health? Or should I just keep the battery fully charged?
Click to expand...
In general it is good to keep the battery between 10%-90% or even 20%-80% SOC if it fits your use case.
Although it is not so critical for LTO compare to other chemistry.

The upper SOC limit should be synchronized and set by BMS and by Charger.
The lower SOC limit is set by BMS.

However, I am not sure that Float Charging the lithium batteries is a good thing.
It is too late today, tomorrow I will post my thoughts on this and some of my BMS settings.
 
At 150+w of charging current, I did notice that my 12g automotive fuse harness from the XT60 input to the charger board was warming up a bit. I'm thinking I might try to find a more robust fuse housing for this one, since it's likely to see the highest current draw. Because the charger board runs at higher voltage, up to 16v, the fuse from the charger to the BMS sees a somewhat lower amperage flow and stays cooler than the input fuse.

I'm really pleased with that little purple diode board, though. Pretty clever overall design!
 
GPak said:
In general it is good to keep the battery between 10%-90% or even 20%-80% SOC if it fits your use case.
Although it is not so critical for LTO compare to other chemistry.

The upper SOC limit should be synchronized and set by BMS and by Charger.
The lower SOC limit is set by BMS.

However, I am not sure that Float Charging the lithium batteries is a good thing.
It is too late today, tomorrow I will post my thoughts on this and some of my BMS settings.
Click to expand...
Here are my BMS settings, to limit/prevent the floating charging.

There is a difference between max charging cut-off voltage and charged resting voltage after charge is complete.
Here are my observations:
2.7V cell – is charging cutoff voltage - 100% SOC
2.6V cell – is charged resting voltage, about 60 min after charge is complete - 100% SOC

In order to prevent rapid charging cycles, it is important to set OVPR (Over Voltage Protection Recovery) at charged resting voltage or slightly below, so recharge cycle will start not sooner than in about 60 min.

Here are my BMS settings (so far):
Cell OVP (Over Voltage Protection) - 2.7V (charger set to 2.7Vx 5S = 13.5V)
Cell OVPR (Over Voltage Protection Recovery) - 2.6V
SOC-100% (100% State Of Charge) - 2.65V (charged resting voltage)

Similarly:
There is difference between cut-off discharge voltage and discharge resting voltage after discharge is complete.
Here are my observations:
1.85V cell – is discharge cutoff voltage
1.9V cell – is discharged resting voltage, about 60 min after discharge is complete

In order to prevent rapid discharging cycle it is important to set UVPR (Under Voltage Protection Recovery) noticeably above 0% SOC discharged resting voltage

Here are my BMS settings (so far):
Cell UVP (Under Voltage Protection) - 1.85V
Cell UVPR (Under Voltage Protection Recovery) - 2.0V
SOC-0% (0% State of Charge) - 1.9V (discharged resting voltage)

These settings are basically for 0%-100% SOC
I may actually adjust settings to 5%-95% or 10%-90% SOC, but the principles of the BMS settings will stay the same
 
Last edited:
Ah, your settings are a little different from the "Default" JiKong BMS settings when you choose LTO chemistry. My present settings are:
Cell OVP = 2.70v
Cell OVPR = 2.650v
and
Cell UVP = 1.80v
Cell UVPR = 1.850v

I've also seen the "voltage recovery" measurements after the passage of time that you're indicating. Your settings essentially leave the "hard stops" in place for OVP and UVP, but narrow the operational range just a bit. This seems like a good idea to keep SoC from hitting the hard stop limits of the battery chemistry. It's easy enough to adjust my settings to increase the "padding" of these parameters a bit. I'll make this change. Thanks for sharing your settings.

What I'm wondering next is where is the "ideal" SoC for long-term use within these parameters? I've adjusted my son's LTO battery pack to an 8.4A charge rate. He works 4 days per week and has about an hour drive each day. The camera runs in parking mode for about 9.5hrs/day while at work. He turns off the camera when he gets home and the car sits in the driveway. The BMS is always powered on. This results in the following power use characteristics:

230w consumed by 9.5h park mode * 4 day work week * ~6w per hour of camera draw
220w consumed by BMS that runs 24h * 7d * 1.3w
Total power drawn from LTO battery every week = 450w

Total amount of LTO charge every week is 8.4A * 13.5v from car's electrical system * 4h of drive time = 450w

He starts the working week (after 3 days of the car sitting in the driveway) with a 6S battery charge level of about 14.4 to 14.5v. By the end of the working week, the battery is charged back up to about 15.8v or 15.9v. Thus, his battery voltage only ever varies by about 1.5v (from 14.4 to 15.9v), which means the battery is nearly fully charged all of the time. From what I've read, keeping batteries fully charged (or fully discharged) seems to create a situation of greatest "stress" on the battery in terms of relative distribtuion of ions on either side of the permeable barrier that separates the anode from the cathode. This seems to be generic advice for "Lithium Ion" batteries. I haven't found any specific advice for LTO chemistry, though.

So, is it better for the life of the LTO battery to maintain a higher voltage level (14.5v to 16v) as it is now, or is it better for the life of the battery so the SoC is maintained close to 13.0v to 14.5v (keeping the 1.5v weekly swing "centered" on the 13.8v Nominal voltage level)?

Or maybe I'm just splitting hairs that don't need to be split here...
 
Last edited:
Usually all cells are in good balance until the cell voltage approaches its maximum or minimum level.
I think a good SOC cutoff level would be the voltage at which the imbalance starts to increase noticeably between any two cells.
In my case it is 2.65V and 1.95V so I can limit the SOC to:

Cell OVP (Over Voltage Protection) - 2.65V (charger set to slightly above of 2.65Vx 5S = 13.25V)
Cell OVPR (Over Voltage Protection Recovery) - 2.60V
SOC-100% (100% State Of Charge) - 2.60V (charged resting voltage)
and
Cell UVP (Under Voltage Protection) - 1.95V
Cell UVPR (Under Voltage Protection Recovery) - 2.0V
SOC-0% (0% State of Charge) - 2.0V (discharged resting voltage)

This is 98.7%-0.65% SOC for total of 1.95% or 4.5Wh capacity reduction, or about 54 min of parking time reduction out of 46 hours, assuming 5W parking consumption.
Note, that in this case the difference between OVP and OVPR is only 0.05V on both ends - farther we are from the extremes less bounce in voltage.

Edit: Based on the test data below I updated the numbers. (GP)
 
Last edited:
Finally completed the capacity tests for the 5S LTO battery pack at different charge and discharge cut-off voltages (OVP and UVP).
The charge current was around 8.4A up to 85% SOC, then gradually dropped automatically to 0A at 100% SOC.
The discharge current was 4A up to a 2V cell, then manually changed to 1A at 2V cell and below.

The total full baseline capacity is 230Wh for the 2.7V-1.8V per cell cut-off range, and the parking time is about 46 hours, assuming 5W parking consumption.
1% capacity reduction is equal to 2.3Wh or about 27.6 min of parking time, assuming 5W parking consumption.

The top end voltage cut-off options (OVP):
2.7V - baseline full capacity
2.65V - is about 3Wh or 1.3% capacity reduction compare to baseline
2.6V - is about 9.2Wh or 3.9% capacity reduction compare to baseline
and
The low end voltage cut-off options (UVP):
1.8V - baseline full capacity
1.85V - is about 0.2Wh or 0.17% capacity reduction compare to baseline
1.9V - is about 0.5Wh or 0.22% capacity reduction compare to baseline
1.95V - is about 1.5Wh or 0.65% capacity reduction compare to baseline
2.0V - is about 3.5Wh or 1.5% capacity reduction compare to baseline

Conclusions:
1. The difference between 2.7V and 2.65V cut-off is about 3Wh/1.3% or about 36 min of parking time, but at 2.65V the battery cells stay in balance, so I think it is better to keep the OVP at 2.65V cut-off setting.
2. The difference between 1.8V and 1.95V cut-off is only 1.5Wh/0.65% or about 18 min of parking time, but below 1.95V the imbalance is increasing rapidly, so I think it is better to keep the UVP at 1.95V cut-off setting.

The imbalance limit between any two cells is 0.01V before balancing is triggered.

Based on this test data I updated my previous post for the preferred settings.
 
Last edited:
Awesome data, GPak! Thanks for sharing. I was planning something similar (and surely less thorough than what you did), but haven't had the time recently.
 
Just wanted to add:
The above Wh capacity and parking time reduction are specific to 5S configuration.
However, the percentage capacity reduction is applicable to any xS configuration relative to the base capacity.
So for the 6S configuration the full/base capacity is 276Wh, and the 1% capacity reduction is equal to 2.76Wh or about 33.12 min of parking time, assuming 5W parking consumption.
 
I was playing around with the voltage setting on my charger board and the BMS setting with the battery pack installed in the car. The behavior is a little different with a direct battery connection compared to when I was powering it from the car’s utility port. The utility port provides a pretty constant 13.4v (I adjusted the voltage output of my SMPS in the house to match the car’s utility port output) whereas the direct battery connection provides 14.2v under load (closer to 14.7v without a load) when the LTO battery is charging. I was wondering why the overall current draw had dropped from 8.0A in the house with the SMPS to 7.5A in the car… Now I know (diff voltage levels).

Anyhow, watching power behavior in the app as the battery neared a full charge, I was noticing the charger board cycling on and off more frequently than I thought it should. This led me to reduce the OVPR setting a bit, leaving a difference of 0.10v between OVP and OVPR rather than the default 0.05v. The new settings are OVP=2.65v and OVPR=2.55v. The wider margin results in less charge on/off cycling for the LTO battery as it becomes fully charged. This seems potentially more gentle to the batteries.

I was also noticing the charge current (as reported by the BMS) seems to bounce around quite a bit as the LTO battery pack approached being full. Rather than slowly tapering off the charge current, it seems to do a random walk: 2.4A then 5.2A then 3.6A, then 0.4A, then 4.8A, …. Eventually turning off as the batteries are topped off.

Any ideas why the charge current bounces around instead of exhibiting a smooth decline?
 
Last edited:
In my case I have an integrated "Dual Source Auto Power Switching Module" (UPS), so the dashcam is powered directly from the car’s electrical system when driving/charging (bypass).
And a fully charged battery, without power draw for dashcam, takes about 1 hour to drop from 2.65V to 2.6V and start re-charging, when driving.
However, I actually like the wider spread between OVP and OVPR and will probably change it to 2.65V/2.55V as well.
The 2.55V would be still a fully charged battery at resting voltage level, if it actually gets that low in my case.
===
I am not sure why the current is jumping in the final stage of charging, it has been a while since I tested it but as you mentioned it should be dropping slowly.
Is your charger CV set to 15.9V (6 x 2.65V)? That is how I would set it.
If so, maybe try playing with it in the 15.9-16.2V range.
 
I’ve played with charger voltages of 15.9 to about 16.2, but it’s hard to dial in that voltage with consistency and precision. It seems to “float” a little from one day to the next.

The last time I played with a UPS module in my power supply chain, the Viofo 139 Pro didn’t like it and it kept rebooting rather than just switching over. It seems to be rather sensitive to the power supply’s consistency.
 
Last edited:
Yes, I have tested many UPS modules and all Relay based units worked unreliably or did not work at all.
The Relays are not fast enough, depending on how sensitive the particular Dash cam is to power interruption.
However, the following two UPS modules are based on the ideal diode principle and work well with a 12V system (tested with the Viofo Mini2, A229 Plus 2ch and A329 2ch):

https://www.aliexpress.us/item/3256..._main.189.68be1802BfLCY6&gatewayAdapt=glo2usa

 
Last edited:
Thanks for the links! Seems that both Amazon and AliExpress have recently changed their sets of permissions. Amazon won’t let you read very many reviews without logging in and AliExpress doesn’t let you do anything without logging in.
 
I finally finished upgrading my 4S (184Wh) battery to a 5S (230Wh) configuration.
Here's a video explaining some of the changes:


Next, I'll post an updated wiring diagram, and hopefully by the end of the week I'll have a short video showing the installation in my wife's Lexus.
 
Last edited:
You must log in or register to reply here.
Back
Top