diy solar

diy solar

LTO battery build with latest tech and teachings for offgrid solar

laraku

New Member
Joined
Dec 26, 2021
Messages
16
Hi all,
I've been following the forums for awhile now and finally made an account to participate in hopes of creating an up to date Lithium titanate (LTO a.k.a. Li4Ti5O12) battery build thread. There haven't been a lot of threads on this recently it seems (I've read most of them now). I really appreciate the time many people take to help others and hope to contribute here too.

The situation:

I'm looking to replace my current dying AGM batteries and after many hours of research I'm planning to get some LTO Toshiba cells that I've sourced and build some batteries with a friend who has more electrical know how. (They are not in the scib packs). I'm pretty set on LTOs as I want super safe, long-lasting batteries that can withstand low Ontario temps without heat when I'm away for days or weeks.

My setup:
  • I've been running a 240 watt x 3 24v solar setup since 2016 with AGM batteries (4 12v 240ah batteries configured to 24v).
  • I've just added another 240 watt x 3 string and I'll be adding 3500 watts more of solar panels this year (haven't decided which ones yet in what configuration).
  • I have a 24v Magnum 4024 pae inverter (input battery voltage range 18vdc-34vdc, so should be good) and an Outback Flexmax 60 charge controller so far (will be getting another charge controller for the new array or if I need something different for the LTOs).
  • The cells I'm getting are the 2.3v 20ah Toshiba prismatic cells (new condition hopefully? and not in scib packs).
So, some questions if anyone would be so kind as to share their wisdom:

1. If I'm looking for around around 12 kwh (500ah at 24v) what cell configuration do you think I should go with? I was thinking 12s6p as I'm thinking of someday changing to 48v when my current inverter dies and this would give me four batteries of 72 cells with around 288 cells total (I'll also get a few extra).

2. What BMS at what amp rating would you recommend these days given answers to #1? Quality and longevity are important. I understand Chargery, Daly, and Batrium are options from other threads. Any advice on best places to purchase too? @wild01 also wrote recently that a BMS isn't even needed, only a balancer. Thoughts? I'm thinking a 200A or 250A daly 15s LTO BMS. Not sure yet if 250A exists.

3. Does anyone have experience in Ontario acquiring bus bars and other needed items? Not sure how common they are for these size of cells or if I should just make my own. Any advice on sourcing needed materials? Advice on copper vs aluminium?

4. Are there any other issues or advice you can think of to mention or resources you are aware of (e.g. will my charge controller work to charge these?)? @Bob B @Will Prowse @120vjohn @Freibeuter @Claudi @erik.calco @thedodgersonline @toms @wild01 @Brendon @Craig

Thank you so much for your advice and insights in advance! I will document my process and share.

Kind regards,
Trent
--------------------

Resources

For anyone interested working on LTO builds here are some links to threads and videos that I've found:

A bit outdated now but some useful info.
https://diysolarforum.com/threads/lto-cells-on-batteryhookup-com.2541/

Some useful background info
https://www.insightcentral.net/threads/used-lto-scib-questions-comments.126355/page-2

Very interesting points
https://diysolarforum.com/threads/lto-bms-sources.27641/

Functioning build
https://diysolarforum.com/threads/yinlong-lto-bank-for-my-off-grid-cabin.6733/

Consideration?
https://diysolarforum.com/threads/lto-bms.5696/#post-58711

Good info
https://diysolarforum.com/threads/lto-scib-battery-bank-for-off-grid.7969/


A bit of background. Old though and some mentioned issues are not considerations for me or have been worked out

An LTO bank build (different cells than me)

Another build

Another build

This guy has some videos of his LTO set up with some good points (its 48v with the yinlong cells)
https://youtube.com/c/MikeLawrence

Likely not needed?
http://www.ejbattery.com/productview_35.htm
 
Last edited:
Thank you for summing up your research!

(1)
I'd recommend a voltage range of 1.8V - 2.5V per cell.
For designing the system, you can use 1.9V - 2.5V per cell, since the LTO cells are not always the exact same capacity and you don't want to dorp below 1.5V/cell on a daily basis.

6S works with 12V Inverters: Range 11.4V - 15.0V (you might have to set end of charge voltage to 14.6V, if your Inverter does not allow 15V)

13S works well with 24V Inverters: Range 24.7V - 32.5V
12S works well with 24V Inverters: Range 22.8V - 30.0V
11S works well with 24V Inverters: Range 20.9V - 27.5V

24S works well with 48V Inverters: Range 45.6V - 60.0V
23S works well with 48V Inverters: Range 43.7V - 57.5V
22S works well with 48V Inverters: Range 41.8V - 55.0V

(2)
If you want to be on the save side regarding longevity and failsafe, avoid a BMS and just use a active balancer, that kicks in over 2.4V/cell.
You could e.g. use the programmable relay of a Victron Battery monitor (e.g. the BMV-702) to switch on the balancer based on SOC (e.g. > 90%) or based on voltage (e.g. > 29V on a 12S pack) or based on both cases, whatever occurs first.
Without a BMS, make sure to set the end of charge voltage correctly to not more than 2.5V/cell and the discharge cutoff around 1.9V/cell...even if one of the cells goes below 1.5V, there is no permanent damage, as far as I can tell from my experiance.

(3)
I made my busbars from aluminium, because it is easy to work with.
The same busbar in copper can be slimmer formfactor of course.

(4)
As long as you are able to set the maximum charge voltage to not more then 2.5V/cell, you are good to go with any charge controller.
If you have more options, like float, bulk and equalization voltages, try so switch equalization off or just set every voltage to 2.5V/cell.
If you want to be extra, of course you can set the float voltage a bit lower, like 2.45V/cell to increase battery life even more. But it won't change too much. Avoiding high temperatures will do more for longevity.
 
Thank you for summing up your research!

(1)
I'd recommend a voltage range of 1.8V - 2.5V per cell.
For designing the system, you can use 1.9V - 2.5V per cell, since the LTO cells are not always the exact same capacity and you don't want to dorp below 1.5V/cell on a daily basis.

6S works with 12V Inverters: Range 11.4V - 15.0V (you might have to set end of charge voltage to 14.6V, if your Inverter does not allow 15V)

13S works well with 24V Inverters: Range 24.7V - 32.5V
12S works well with 24V Inverters: Range 22.8V - 30.0V
11S works well with 24V Inverters: Range 20.9V - 27.5V

24S works well with 48V Inverters: Range 45.6V - 60.0V
23S works well with 48V Inverters: Range 43.7V - 57.5V
22S works well with 48V Inverters: Range 41.8V - 55.0V

(2)
If you want to be on the save side regarding longevity and failsafe, avoid a BMS and just use a active balancer, that kicks in over 2.4V/cell.
You could e.g. use the programmable relay of a Victron Battery monitor (e.g. the BMV-702) to switch on the balancer based on SOC (e.g. > 90%) or based on voltage (e.g. > 29V on a 12S pack) or based on both cases, whatever occurs first.
Without a BMS, make sure to set the end of charge voltage correctly to not more than 2.5V/cell and the discharge cutoff around 1.9V/cell...even if one of the cells goes below 1.5V, there is no permanent damage, as far as I can tell from my experiance.

(3)
I made my busbars from aluminium, because it is easy to work with.
The same busbar in copper can be slimmer formfactor of course.

(4)
As long as you are able to set the maximum charge voltage to not more then 2.5V/cell, you are good to go with any charge controller.
If you have more options, like float, bulk and equalization voltages, try so switch equalization off or just set every voltage to 2.5V/cell.
If you want to be extra, of course you can set the float voltage a bit lower, like 2.45V/cell to increase battery life even more. But it won't change too much. Avoiding high temperatures will do more for longevity.
Thanks a lot for the advice @Raphael Hofmann !
 
Okay, an update on this project:

I ended up going with only about 9 kwh of batteries rather than 20kwh for now since we've decided to put our river to use with hydroelectric power and thus won't need as much storage for cloudy days. So I'll be using 216 20Ah Toshiba LTO cells, which are now on their way.

I've also decided to go with six 3p12s packs potentially, and possibly just throw a 150A JK BMS on each one of them with 2A balancing, as my understanding is that these LTO cells are pretty tight in terms of similar resistance and capacity, so I'm hoping I don't need a separate BMS for each 12s. I'll capacity test all of the cells and match them based on capacity and resistance too, however, and will keep an eye on things and rearrange them if needed to use a BMS for each 12s. Attached is my design idea for each pack, which will be stacked on top of each other on shelves and connected in paralell (using 1/0 insulated welding wire with tinned copper ring terminals and brass nuts and washers). The purple areas are the series connections and the red and black are positive and negative connections (using tinned copper bus bars rated for 150A). Hopefully everything else is self-explanatory.

Any comments on this idea? @upnorthandpersonal @Raphael Hofmann

Any comments on whether the BMS's will work okay with this paralell design? I know a BMS is not necessary if an active balancer is used, but I do like the idea of having an extra layer of under and overvoltage protection, high temp protection, as well as being able to stop them charging below -15 degrees C or so (they can handle this but Toshiba test charts indicate it may shorten their life a bit). Hopefully the current is high enough to keep them balanced.

I'm also thinking of a work-around with my limited charge controller capabilities (Outback Flexmax 60) to get them to charge only to 90% or so and then cut off charging, without ever putting them into float, using a Victron 712 battery monitor and resistor and relay connected to the charge controller's temperature probe. Bit of a workaround I saw someone else use with LFPs for the same reason. According to Toshiba test charts LTOs don't like to go into float and this will shorten their life.

Anyway, that's all for now, all comments appreciated!
~Trent
 

Attachments

  • 3p12s design.PNG
    3p12s design.PNG
    14.9 KB · Views: 28
Last edited:
12S 3P per pack sounds good! The BMS can be paralleled...no problem.
Also the JK BMS should do the job just fine and 2 A is more then enough balance current, as long as you don't leave the balancer switched on all the time. You can adjust that setting in the BMS...e.g. switch balancer on over 2.4 V/cell.

When you have more constant hydro power production, you can definetly save on batteries.

Well...the Battery monitor is never 100 % accurate and needs to be synchronised to 100 % SOC every now and then.
You could probably do that with the "Equalisation" setting in the charge controllers...just set it to 30V once a week for 1h.
Your normal "absorption" voltage can be set to e.g. 29.4 V and your "Float" voltage to 28.8 V.
 
12S 3P per pack sounds good! The BMS can be paralleled...no problem.
Also the JK BMS should do the job just fine and 2 A is more then enough balance current, as long as you don't leave the balancer switched on all the time. You can adjust that setting in the BMS...e.g. switch balancer on over 2.4 V/cell.

When you have more constant hydro power production, you can definetly save on batteries.

Well...the Battery monitor is never 100 % accurate and needs to be synchronised to 100 % SOC every now and then.
You could probably do that with the "Equalisation" setting in the charge controllers...just set it to 30V once a week for 1h.
Your normal "absorption" voltage can be set to e.g. 29.4 V and your "Float" voltage to 28.8 V.
Thanks for the advice @Raphael Hofmann !
 
I know the cycle is life is ~20,000, but I'm not sure about the calendar aging of LTOs. If they'll last 50 years being cycled once a day then they'd be great for solar/wind if you have the space for them. But, if the shell or other parts disintegrate within a decade because they were designed for a regenerative breaking app with 100's of cycles per day, then not so good. Have you by chance seen anything regarding their calendar aging?
 
I know the cycle is life is ~20,000, but I'm not sure about the calendar aging of LTOs. If they'll last 50 years being cycled once a day then they'd be great for solar/wind if you have the space for them. But, if the shell or other parts disintegrate within a decade because they were designed for a regenerative breaking app with 100's of cycles per day, then not so good. Have you by chance seen anything regarding their calendar aging?
In the datasheets you can see values from 25 years to 30 years for calandar lifetime.
Those LTO cells are commercially available since 2009.
 
12S 3P per pack sounds good! The BMS can be paralleled...no problem.
Also the JK BMS should do the job just fine and 2 A is more then enough balance current, as long as you don't leave the balancer switched on all the time. You can adjust that setting in the BMS...e.g. switch balancer on over 2.4 V/cell.

When you have more constant hydro power production, you can definetly save on batteries.

Well...the Battery monitor is never 100 % accurate and needs to be synchronised to 100 % SOC every now and then.
You could probably do that with the "Equalisation" setting in the charge controllers...just set it to 30V once a week for 1h.
Your normal "absorption" voltage can be set to e.g. 29.4 V and your "Float" voltage to 28.8 V.
I was thinking of trying to rig something so I would not have to pass charge or discharge current through the BMS's and thus not rely on a BMS to interrupt charge or discharge. However, if I use a 200A JK BMS, maybe it's not a big deal, and I should just use the BMS's to cut off charge at a certain voltage like 2.6v or whatever voltage is needed (will experiment) to not charge them quite to 100% -- since I am not sure I'm up to rigging up control relays, resistors, and what not yet as @Johncfii mentioned in this thread (plus not sure how to modify this solution for LTOs). I would rarely use even 100A that will be what the bank is rated for, so maybe it's okay. I am tempted to figure it all out and save some money on the BMS's like he did, though. Like he says in the thread I'm not sure how I feel about relying on the BMS's for charge and discharge current. Or am I just worrying about it for nothing?
 
Last edited:
Darn, I was just ordering the JK BMS and realized it only supports 14s minimum with LTOs according to documentation. Anyone know of a workaround for it to work with 12s? I guess it's a voltage thing?
 
Darn, I was just ordering the JK BMS and realized it only supports 14s minimum with LTOs according to documentation. Anyone know of a workaround for it to work with 12s? I guess it's a voltage thing?
Usually, you can use a boost converter to boost up the voltage for the BMS.
You connect all sense wires normally.
And for the power supply of the BMS (usually the last positive wire), you can use a DC-DC step up / boost converter.

JK-BMS DC-DC.PNG
 
I also in progress to build a LTO for another purposes (car starter battery)
My plan is to use protection board of Super capacitor to regulate the charging voltage....
Afaik the protection board will be limiting the maximum voltage but there is no current regulating....
And since it using TL431, there is possibility to "adjust" the target voltage output by change the value of resistor R1 and R2...
 
I also have a 6S LTO Battery (40 Ah Yinlong) in my car. I use it without protection or balancing.
It works well for half a year already. But it is everything else than professional ?
My car battery died and a needed a quick replacement.

signal-2022-02-08-090557_002.jpegsignal-2022-02-08-090557_003.jpeg

I use it in a Volvo XC70 from 2009, drive e version.
Longterm, I want to replace the 40 Ah Yinlong cells with smaller 15 Ah or 20 Ah cells an place them into a closed battery box.
 
Last edited:
I also have a 6S LTO Battery (40 Ah Yinlong) in my car. I use it without protection or balancing.
It works well for half a year already. But it is everything else than professional ?
My car battery died and a needed a quick replacement.

View attachment 82913View attachment 82914

I use it in a Volvo XC70 from 2009, drive e version.
Longterm, I want to replace the 40 Ah Yinlong cells with smaller 15 Ah or 20 Ah cells an place them into a closed battery box.
wow... nice... (y)
the battery that I'm going to use is Lishen with 16Ah capacity, but have nominal voltage at 2.5v...
So it will be going to 5S
1644310026700.png

1644310082904.png

with discharge spec, hope it will enough to start the car... ?

already ordered, but due to chinese new year holiday, the shipment become postponed ?
 
I wonder, if the datasheet is actually correct. [Edit Spoiler: Yes, it is! See post below.] It must be a different chemestry, when the nominal voltage is 2.5 V instead of 2.3 V.
for the Yinlong LTO cells, the cut-off voltage is also 2.9 V. But you should not absorbe at 2.9 V. Absorption should be 2.6 V or 2.5 V.

In a regenerative breaking scenario, 2.9 V is fine as the maximum charging voltage. But in a solar installation or as a car battery, that is charged with a constant voltage of e.g. 14.4 V, it is better to use a 6S instead of a 5S battery pack.

With 14.4 V constant charge voltage of the car alternator:
5S: 2.88 V / cell
6S: 2.40 V / cell

The Yinlong LTO cells won't last long as a 5S configuration.
But maybe the Lishen cells have actually a different chemistry and it is just fine to absorb at 2.9 V.

Please keep me updated on your experiences with the Lishen cells!
If you find a discharge curve of a Lishen cell, please feel free to post it.

Discharge curve of Yinlong 40 Ah LTO cell with ZKE-Tech EBC-A40L:
40Ah LTO grade A discharge curve.jpg
As you can see, the average voltage during discharge is 2.16V @ 1C.
I am curious to see the average voltage of a Lishen cell @ 1C discharge rate.
 
Last edited:
I wonder, if the datasheet is actually correct. It must be a different chemestry, when the nominal voltage is 2.5 V instead of 2.3 V.
for the Yinlong LTO cells, the cut-off voltage is also 2.9 V. But you should not absorbe at 2.9 V. Absorption should be 2.6 V or 2.5 V.

In a regenerative breaking scenario, 2.9 V is fine as the maximum charging voltage. But in a solar installation or as a car battery, that is charged with a constant voltage of e.g. 14.4 V, it is better to use a 6S instead of a 5S battery pack.

With 14.4 V constant charge voltage of the car alternator:
5S: 2.88 V / cell
6S: 2.40 V / cell

The Yinlong LTO cells won't last long as a 5S configuration.
But maybe the Lishen cells have actually a different chemistry and it is just fine to absorb at 2.9 V.

Please keep me updated on your experiences with the Lishen cells!
If you find a discharge curve of a Lishen cell, please feel free to post it.

Discharge curve of Yinlong 40 Ah LTO cell with ZKE-Tech EBC-A40L:
View attachment 82917
As you can see, the average voltage during discharge is 2.16V @ 1C.
I am curious to see the average voltage of a Lishen cell @ 1C discharge rate.
Here's the 18ah Lishen 1C discharge curve

JimJr
 

Attachments

  • 2022-2-8-18ah 1c.jpg
    2022-2-8-18ah 1c.jpg
    138.6 KB · Views: 37
  • 2022-2-8-21 18ah.txt
    32.9 KB · Views: 21
@JimJr
Thank you so much! The voltage is indeed higher!
So we can conclude, that the Lishen cells actually have a nominal voltage, that is about 0.2 V higher than the Yinlong cells.

@-sgt-
With the Lishen cells you can only use 5S as a car battery replacement. (y)
But if you can adjust the alternatervoltage, try to adjust it to 14 V or 13.8 V.
With 6S Lishen cells you would need a alternatervoltage of 15 V to make it work.
 
Last edited:
Back
Top