diy solar

diy solar

Question About LTO Cells

bobkart

New Member
Joined
May 10, 2020
Messages
21
Location
Pacific Northwest
I picked up six 40Ah LTO cells earlier this month and am wondering how tolerant of overdischarge they are.

I was individually charging cells to 2.8V and unplugged the charger overnight, but left it connected to the cell (that was my mistake). I assumed the charger would go into high-impedance mode when not powered on, but there is at least a small green LED that stays lit. Well, in the morning I see that the cell had run down to 1.3V, somewhat below the 1.5V they like to stay above (or is it 1.6V)? I quickly added charge again, and the cell seems to be behaving as it did before, although I haven't put it through much since then.

My online searching has turned up no help on this question. I understand LiFePO4's don't like going below 2.5V, and will be damaged by such overdischarge, so this makes me wonder about the LTO cells.

I guess I could put them in a 6S array and apply a moderate load (after first top-balancing them), then monitor individual cell voltages, to see if the cell in question starts dipping before the others. But hopefully someone more knowledgeable about LTO chemistry can weigh in.
 
I am going strictly from memory ..... so more research is necessary .... but from what I remember when I was researching them, they are extremely tolerant of very low discharge with no permanent damage.
 
Did a little more looking, and found this post by Craig .... he has an LTO pack and actually sells them.
Maybe do a search of the forum with LTO and member Craig. He is the most knowledgeable about this chemistry that I am aware of.
 
Thanks for your help, Bob. It looks like good news. My 1.3V discharge is nothing compared to the 0.5V incidents I see in that thread. Craig also mentions that LTO is the only cell chemistry that can be fully discharged with no damage. Impressive!

I'll still keep an eye on the individual voltages, but can rest more at ease that my mistake didn't cost me a cell.
 
Wondering now about LTO cells being held at high SOC for long duration, especially in elevated ambient temperatures.

I know this is to be avoided for LiFePO4; are there similar considerations for LTO or is this not a problem?

If no problem then this chemistry would seem to be more suited to standby UPS applications than LiFePO4.
 
Occasionally I see mention of the torque limits on those LTO cell terminals . . . warnings about not to exceed them while tightening connections. But obeying these limits might leave connections not as tight as one might like.

I have an approach that sidesteps those limits: before any attachments, run a nut down to the bottom of the terminal (preferably Aluminum as that's what the terminals are made of). Finger-tight. Then add your connection (buss bar or cable lug for example), then use a second nut to tighten against the first nut, instead of against the cell itself. By using an open-end wrench (3/4" or 19mm) on the first nut, you can prevent tightening torque applied to the second nut from making it to the cell. Be sure to use that open-end wrench when loosening these connections, or you run the same risk of damage to the cell.
 
I picked up six more cells, and am top-balancing all twelve. The goal is to build an 11S arrangement, to closely approximate an 8S LiFePO4 arrangement voltage-wise.
 
Occasionally I see mention of the torque limits on those LTO cell terminals . . . warnings about not to exceed them while tightening connections. But obeying these limits might leave connections not as tight as one might like.

I have an approach that sidesteps those limits: before any attachments, run a nut down to the bottom of the terminal (preferably Aluminum as that's what the terminals are made of). Finger-tight. Then add your connection (buss bar or cable lug for example), then use a second nut to tighten against the first nut, instead of against the cell itself. By using an open-end wrench (3/4" or 19mm) on the first nut, you can prevent tightening torque applied to the second nut from making it to the cell. Be sure to use that open-end wrench when loosening these connections, or you run the same risk of damage to the cell.
I think you will get a lot better connection to the bus bar by tightening it against the pad on the cell .... there will also be an extra mechanical connection for the current to pass thru. I guess it kinda depends on how much current there is going to be.
 
Hi again Bob, thanks for weighing in.

I see that concern. I think if I was to go ahead and tighten the first nut down onto those pads (beyond finger tight, but within the torque spec of course) I'd be in better shape, rather than relying on all current flowing through the M12 terminal post. Those M12 terminal posts have 2.3x the cross-section area of 2AWG wire, which in Aluminum and with 90C insulation rating, can safely carry 100 amps. And the 19mm aluminum nuts I'm using increase that area to about 3.6x that of the M12 terminals posts, for a total of 8.3x that of 2AWG wire. My highest anticipated load is less than 200 amps. But of course less voltage drop is better, so just saying 'it will safely carry the current' isn't the same as 'it will carry the current with acceptably-low voltage drop'. On the other side of the argument, the length of M12 terminal post needing to be traversed by that current is so short that wire ampacities don't really apply.

I may give it a try with no leading nut. I just don't want to ruin a cell by overtightening, so I should probably get a good torque wrench whose range covers that spec before I do that.
 
It's pretty disappointing that both the LTO and the prismatic cells have connection issues due to weak aluminum threads .... I didn't realize you were using aluminum nuts, that should help a little.

I would do a test with fairly high current draw and measure voltage drop and temperature of the connection point with and without your extra nut .... please keep us informed how this works out for you.
 
Who Needs A Torque Wrench!

I found that the 7 newton-meter torque spec for those terminals converts to a bit over 5 pound-feet. The 19mm combination wrench I'm using is 8" center-to-center. So I just need 7.5 pounds to hang from one end. The SKB case I'm going to put these cells in weighs 4.8 pounds, and one of these LTO cells weighs 2.7 pounds. So I put a cell in the case and hung it from the far end of the wrench parallel to the ground.

G0019036-crop.jpg

I know this discounts 'half the weight of the wrench'. Should be close enough, as the actual number I want is more like 5.16 pound-feet. I turned (by hand) the cell until the wrench was level.

Another detail: some of the threads don't run all the way down to the terminal pad, causing an air gap between the nut and the pad. So I added some 24mm OD aluminum washers, and the contact between the nut and the pad is good.

Now the detail-oriented person may ask how I'll do this to the last cell, as the weight of a cell with nuts and washers is 'more' than 2.7 pounds. I'm joking really as the weight difference is negligible. But it just so happens that I need eleven cells for this pack I'm making, and I have twelve cells. So that last cell doesn't need this treatment!
 
Very innovative ... I like thinking outside the box. Probably more accurate than your average torque wrench.

At least the torque spec on those cells is a lot better than the prismatic cells .... anything greater than 4 NM or 35 in lbs is likely to cause stripping of the threads on those.
 
Yeah I was looking online about torque wrenches and the question of calibration came up. Turns out the technique I use above is how most people suggest they be calibrated.

I do have a couple of torque wrenches, but they're for much higher torque values than this, like 100-pound-feet (for lug nuts). So I don't really feel they'd be very accurate in this application.
 
Last edited:
I finished building the 11S LTO pack. It's got 3/8" terminals on each side, with red & black 'wing nuts', so no tools are needed to make connections.

Weight is 36.2 pounds. Capacity is 11 X 2.3 volts X 40 amp-hours = 1,012 watt-hours.

I was shooting for at most 30 volts when 'nearly full', and am seeing 29.8V currently.

Nominal will be 25.3 volts, very close to the 25.6 volts for an 8S LiFePO4 pack.
 
Fusing is handled in the cabling between battery and load. In this case my load is a 2.2kW electric outboard motor, which might pull as much as 50A from 48V nominal. I have half of my 200Ah LiFePO4 pack providing the other ~24V for now; eventually it would be two of these packs. I realize that they're not as good energy-to-weight-wise as LiFePO4, but I have a second application, which is a 6kW electric outboard motor, and that will pull as much as 125A, and 40-50Ah LiFePO4s typically can't handle that discharge rate (2C-3C). So building the fusing into the case wouldn't let me change the amperage limit as easily, because I usually use the 'Maxi' fuse style for the smaller load (80A fuse), but will use something more like a breaker for the large load (150A).

BMS is another discussion altogether. Suffice it to say for now that my use case doesn't really need that. I know there will be those out there that can't imagine not using a BMS, but I made it work all last year over a couple dozen trips with 4S LiFePO4 packs.
 
Generally, you want a fuse as close to the main battery connection as feasible. This fuse is dependent on your wire size more than the load .... and is needed in the case of an inadvertent short circuit.
Otherwise, your wire becomes the fuse and bad things can happen.

If you have a fuse at or near your load and the wires rub thru or for some other reason you get a short before the fuse .... the fuse doesn't do anything for you.
 
Back
Top