Honda Insight Forum banner

LTO "Lithium" battery conversion thread

165K views 860 replies 51 participants last post by  insightbuyer 
#1 · (Edited)
I am starting a LIthium titanate (LTO) battery conversion on my 2000 Insight.
I will try and give details and pictures about the whole process.

The conversion will be a 3.3KWH, 72S, ~185 volt 90lbs pack.

UPDATE: Upgraded it to a 3.86KWH 84S, ~210 volt pack by adding 12 extra cells.

I will be using the BCM fooler + BCM interceptor, as they are simple to install and should make the whole conversion integrate nicely.

The battery packs are from a Honda Fit EV, which was produced in the USA from 2013 through 2015 and available by lease only.

Packs come in 24 20AH cells, 1.1KWH (they actually test at 22AH but rated sticker is 20AH)
2.3 volt nominal,
voltage range : 1.5 - 2.7 *most 95% of capacity is between 2.0 - 2.5 volts
Weight of pack: 30.2 LBS
Deminsions: ~10.5" x ~9.5" x ~5.5" *Each 24 pack can be divided easily in half.
Continuous charge: 160 amps
Continuous discharge: 160 amps
Temperature Performance -30C to 55C


Voltage DROP across 3 cells at resting 7.17 volts was .16 volts at 61.8 amps. Compare that to a 2 volt drop on one of my NIMH sticks which was fully charged.

From my testing of the 3 packs 72s, there was only 1 bad cell. The other cells tested fine and moved in perfect balance with each other.

The active BMS located on the packs is not usable and seems it needs to be plugged into a honda fit car so that it can communicate. So far there are really not many options for a BMS.

The reason I decided to use these cells is they are very robust, extremely long life, work in freezing temperature, Very safe from fires. They simply make a perfect NIMH hybrid battery replacement. The drawback is they are heavy, weighing in at almost NIMH weight. A 72S 3.3 KWH pack will weigh 90 LBS.

Thorough testing at 20 amp charge/discharge of all the cells.


Each pack has removable black covers that go over the top.


30.2 LBS per pack, which is heavy considering you need 3 of them.


Voltage and capacity out at 20 amps discharge rate, test starts at middle of paper below or 50 MAH.
 

Attachments

See less See more
5
#2 ·
Advice needed on my conversion:

What wire gauge should I use connecting these individual packs in series? Is 10 AWG Gauge stranded enough or should I use a thicker gauge wire?

The wire runs are so short so I am not sure if I need to use thicker than 10 gauge.

Thanks.
 
#4 ·
#8 · (Edited)
Really, I would be so happy to get a current mod. The LTO cells can easily handle, I believe, 160 Amps continuous, and Toshiba says it can handle around double that for short bursts. However, I'm not 100% sure about these specs as it was listed someplace for 20 AMP hour cells, and another for 40 AMP hour cells or 2P. Can't confirm for sure the current handling.

Either way I should easily be able to handle 160 amps for 10 second bursts of assist.

Here is where a current spec is listed if anyone is interested.
https://www.toshiba.com/tic/datafiles/SCiBBrochure2014Final.pdf

The good thing is you can beat these cells all day at high amps and they will take it. I also believe you can beat the electric motors all day with no issues. I push 7500KW on my ebike all day and that is considered low power compared to what many other people are pushing on these small motors with not very good cooling. I could imagine a bigger motor, like what would be in the insight could handle much more power without any problems, especially considering you are not using continuous power at high levels.
 
#9 · (Edited)
retepsnikrep, what do you think if I used 84 cells total, and only charged around 2.45 volts per cell to keep it under the DC-DC converter max 210 volts.

What about if I put like 10 diodes in series to lower the voltage to the DC-DC converter by about 10 volts so that I can charge the LTO cells to the max? I assume the DC-DC converter will not draw too many amps and it will not be a big deal to lose a little bit of power from the diodes to have a larger battery in the car.

Or if worse comes to worse I can just pull out 4 cells from a pack and use 80 cells. The LTO cells max resting voltage is 2.56 so this would work as I should not charge this high anyway.

I could fit 3 packs plus 1/2 packs easily of the LTO cells in the battery compartment, so why not use them? I also plan on buying a spare pack so will have an extra pack anyway sitting with no use.

Should I consider doing this, or will it cause me more trouble than it is worth using such high voltage?
 
#10 ·
The DC - DC converter shuts down around 220V.

Just use 84 cell charged to 2.56v = 215V approx..

Forget diodes to the dc-dc, it draws a fair bit of current..
 
#11 ·
Ordered another LTO pack today, this will allow me to now use 84 cells.

Waiting in the mail for:

BCM FOOLER AND INTERCEPTOR

My Meanwell power supplies and LED CC supply that should allow me to charge my pack at 2 amps. Will charge a complete empty pack in 10 hours.

Bunch of M6 4 AWG ring terminals to connect the pack in series.
 
#12 · (Edited)
I received my power supplies for my charger, and also the ring terminals.

I have to figure out in what and how I am going to mount those. I also need to decide if I want to leave this in the car, or keep it outside the car.

Weight is about 4 lbs for all 5 supplies, which consist of all meanwell:
3x LRS-100-48
1x LRS-50-12
1x LPC-100-2100

The ring terminal ends seem very long, not sure why they are so long, just wasted weight as they are heavy. I will solder the wire into them as I don't have a crimper. I believe a soldered connection is better anyway.

If anyone could give advice on if I should mount the power supplies in the car or leave them outside and plug into the car from the outside with the 210 volts?

If left in the car and connected to the battery but unplugged, will they cause any errors with the car computer?

EDIT: YOU DO NOT NEED THE 12 VOLT POWER SUPPLY FOR THIS SETUP, THE VOLTAGE IS HIGH ENOUGH WITH JUST THESE 4
3x LRS-100-48
1x LPC-100-2100

I never wired in my 12 volt power supply as just the 4 I listed will charge up to 215 volts.






 
#14 ·
When I made one from those components I just used a case from some old computer part that I had laying around. Many have used power supply cases. I never leave them in the car. Some have gotten CELs if they start the car with the charger plugged into the battery. I think it's better to remove it from the car.

OTOH, I recently bought a car that has one permanently mounted. It's been there for years and works fine.

Sam
 
#16 · (Edited)
YES, in the beginning, I will run this without a BMS. While not recommended, the cells so far all tested to move perfectly together. I am talking .001 volts together and all 3 packs. Other than the one bad cell which I will discard.

A BMS is definitely not needed for balancing.

The big issue without having a BMS is that you never know if a cell will go bad and go out of balance. That is really the big risk of not having a BMS. This chemistry is very safe, you can discharge to 0 volts, but I am not sure what happens if you overcharge a cell.

However, these cells are used and a few years old and most likely a bad cell would have shown signs like the one bad cell I detected. Highly unlikely it will just go out of whack after all this time when the cells are protected from external damage.

But I plan on doing something down the road and more research into a BMS, I'll be doing checks often of the cell voltages and leave my battery compartment opened. Hopefully, Linsight comes out and that will have a BMS, so I will at least wait until I see what is up with that.
 
#17 · (Edited)
I decided I am going to build a clear acrylic case for the charger. It will look cool and also allow for easy adjustment of the voltages. At first I am just going to mount them on wood until I get the materials.

I also plan on adding this 0-300V - 0-100A display voltmeter ammeter, for only $14.80 dollars it will give me an idea of the charge amps, and voltage, amp hours out, the temperature of inside charger case.

It got good reviews and some reviewers say it is accurate, so worth the try. Would be nice to have an idea of what the charger is doing.

https://www.aliexpress.com/item/AC-...1.html?spm=a2g0s.13010208.99999999.263.lfKBjc

 
#19 ·
I would say if the cells move together you should be OK with periodic checks. I have 280 cells in my ebike which I never put in parallel, and the cells are still perfectly balanced over 2 years of use.

The way I look at it is that worst case scenario if you have the aluminum battery cover on, you have a fire and total the car, this would only happen if you were extremely unlucky. Basically, you would be driving and hear or see smoke, and just pull over quickly and exit the car.

I wouldn't chance this in your home but not a big deal in the insight.

I found the same thing as you there really isn't all that much out there with BMS. That is why if Mudder actually made a BMS that was adjustable to different voltages, cut off voltages, etc. and was an active BMS where it charged the weak cells, he would have one of a kind BMS.

It really is surprising that nobody sells an adjustable BMS, all the BMS have set voltages. Even with a simple 3s lipo protection board, they are only available with a 2.5 volt cutoff, which is way too low. I am going to modify one so that it cuts off at 3.3 volts to use with heated electric clothing.


I'll definitely take a look at your thread, I have actually read some of it already.
 
#20 · (Edited)
I posted in that other LTO thread that there's quite a few super capacitor balance boards with balancing voltages that seem appropriate for LTO cells. Wouldn't those work, as far as balance is concerned? You're using the BCM interceptor and fooler, right, so those should take care of top and bottom voltage, etc... What else would be needed - in terms of a basic setup that's more or less safe?
 
#22 · (Edited)
IamIan,

I really like that idea. I would have to test it to see just how sensitive it is to changes in cell voltages. If it could go off with .05 volts difference in voltage than that would detect a single cell going out by about 10% capacity, and would be a good early warning. If it needs like 1 volt to set off an alarm than that may be a little too much.

I bought 4 of these precision voltmeters and will hook them up to my 4 separate packs. They got good reviews and they said they are very accurate, voltage limited to 34.5 but that is high enough per pack. This should give me an idea how well the separate packs move together and will show me if a cell goes out of whack.

https://www.aliexpress.com/item/5-D...32354601094.html?spm=a2g0s.9042311.0.0.s4Y7K3


But monitoring it manually like this sucks, so would want something like your idea eventually that would just shut down the pack.

 
#23 ·
IamIan,
I really like that idea. I would have to test it to see just how sensitive it is to changes in cell voltages. If it could go off with .05 volts difference in voltage than that would detect a single cell going out by about 10% capacity, and would be a good early warning. If it needs like 1 volt to set off an alarm than that may be a little too much.
Usually the simple 1/2 pack difference method is used at higher easier differences nearer to simple 1 cell drop out detection levels.

I have my doubts you can get the simple/basic passive component version easily down to that 50mV.

You could instead get/make a fairly simple logic device down to that level. Detect it , and send a signal.

Lower difference can be attractive , but it has trade offs .. The lower the difference , the more closely matched you have to make sure everything about your two 1/2 packs are .. capacity, internal resistance, temperature, etc .. especially at higher and higher rates power/current.
 
#25 ·
In order to automate the control of the BCM interceptor you will need a microcontroller to measure the voltages and and act on the measurement. One option would be the route I went with a master/slave architecture, but you got me thinking about another possibility. One of the ideas for a diy BMS that I posted in my thread used an arduino as a voltmeter and used relays to change which cell positive and negative was connected to the arduino for measuring the voltage. You would need a voltage divider to be able to measure more than 5v, but since each bank is being measured by the same voltage divider and arduino pin, you should be able to easily get the precision you would need. You can get optoisolated relays cheap in ebay and would need 8 for 4 banks. You could also pair the positive and negative arduino pins with a multiplexer each (again cheap in ebay) so that only one relay can be open at a time and even if your code messes up you won't burn out the arduino.

Since the measurements are not taken simultaneously, and you are looking for such high precision, you would need to make sure that you only go off of ones that are taken when there is no current. If you used a copy of my arduino based BCM interceptor you could get the current from the BATSCI data via i2c and only calculate measurements taken when current is 0.
 
#27 ·
Sounds interesting, I will surely take a look at this. I didn't get a chance to read over your thread yet. I will put a lot of research into the BMS after I get my pack installed, which hopefully will be installed this week, and that is when I'll be doing a lot of research and I'll be reading how your BMS works.

I'm still waiting for the interceptor and Fooler, but I may pull out my NiMH pack tomorrow and start figuring out the wiring, and also figuring out how I'm going to install all these packs in the battery compartment.
 
#26 ·
Just did some more calculations. From what I can find Toshiba rates the capacity of these cells at .2C, which is a 4 amp discharge from 2.7 to 1.5 volts.

I discharged a single cell after making sure it was topped off at 2.7 volts, this gave me 20.4 amp hours at 4 amp discharge.

20.4 amp hours capacity is very good for used old cells. However, I believe some say they are getting 22 amp hours out of these. Can't be sure how they are testing them as I know my testing is accurate.

My battery capacity with 84 cells will be at least: (using 20AH instead of 20.4AH)
(84 cells) (2.3 Volts) (20 AH) = 3,864 watt hours.

The stock NIMH with new cells had:
(120 cells) (1.2 volts) (7) = 1,008 watt hours.

However, the best of my cells only had 6.5 Amp-hours, many had 5.5 AH left.
(120 cells) (1.2) (6.5) = 936 watt hours

This means my new pack will have over 4 times the capacity of my NiMh pack.

Don't forget that after sitting only 5 days you lost about .5 amp hours from your NiMh pack from self-discharge. The BMS also won't let you discharge all the usable capacity. The LTO cells can safely be pushed 100%-10% SOC so you get much more usable capacity.

Here is Toshiba specs on the 20AH SCIB cell
s.


Wide effective SOC*range
SCiB™ exhibits excellent input/output characteristics over a wide SOC* range of 0-100%. This makes it possible to reduce the nominal battery capacity or amount of batteries necessary for a system.
 
#28 ·
I received my PCB's from the United Kingdom today, took 7 days to arrive to my home in USA.

Peter was very nice in that he included a current Hack mod and OBDIIC&C. I will surely use those, especially the current hack mod because my new battery will easily handle increased amps.

I got just the bare PCB boards but you can purchase them built also. I already built the BCM Fooler board.

Connecting everything should be fairly simple so don't let any of this stuff scare you away if you are not experienced with circuit boards.

Now that I received these boards, I will be spending most of my time getting this all hooked up.



 
#29 · (Edited)
Quick question if anyone can help me out:

The Bypass Junction board from what I can determine connects to the Battery pack at 4 locations, splitting the pack into 2 halves.

Since I will use 7 packs of batteries, at around 30 volts each pack I can't easily split this in half as the connections are only at the ends of the packs and I can't easily add new connections to the aluminum cell tabs.

Does it matter if I split it between 3 packs and 4 packs?

This would give me

1st half of pack 90 volts (3 packs)
2nd half of pack 120 volts (4 packs)

I don't see any issues with this, but one thing I've learned is that it is easy to overlook things.

 
#31 ·
Thanks, on to the next issue. This will probably be the most difficult part is figuring out how to secure the packs.

The white paper represents a pack. The packs will fit but that black air duct circled in red in the picture below is in the way.

I assume this vent is needed and should be left in place?

Which would leave me having to build something to elevate the 3rd pack above this black duct.

Please let me know if any of you have ideas or know the deal with that air vent.

Thanks in advance if anyone has ideas.

 
#32 · (Edited)
Get rid of the vent I did.. Just let the MDM fan exhaust into the IPU compartment..
Block up the floor exit hole with a round piece of foam board and some silicone goo.

Unless you live in a mega hot state?
 
#33 ·
Makes sense to just get rid of it, will make my life a lot easier fitting the batteries.

I've yet to hear it ever run and I had the battery compartment opened for a lot of the summer unless it is really quiet.

I've checked the voltages of how the pack is split, it seems that it isn't even split in half. It splits it 64 volts - 96 volts

 
#37 ·
They resize fine on my Firefox?
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top