Ultra Cheap 66160 30Ah LTO - Or Scam?

Sunder

10 MW
Joined
Sep 6, 2011
Messages
3,054
Location
Sydney, Australia
About 3 weeks ago, I took a $550 USD punt. I bought 12 of these:

foto_no_exif.jpg

They arrived neatly packed today. These are Lithium Titanate batteries.

At $22 per 30Ah + shipping, these units are ridiculously cheap at about 70c per Ah. The going rate is between $2-3 per Ah.

These are rated for 300A charging and discharging, with a shelf life of 20+ years, and 30,000+ cycles. Of course, their energy density for both volume and weight is terrible, so they are totally not suited to EV or eBike use, but this is an experiment for off grid power. ONE of these cells weighs a kilo, and is the diameter of a coke can, and about 50% taller - and you need 11-12 to make a 24v battery

The only question now, is have I bought the deal of the century, or have I been ripped off? Right now, being at work, I couldn't even tell you if I've bought a 66 x 160 cylinder full of sand... But if this kind of thing interests you, stay tuned for my testing! (Will probably do a few quick tests tonight, but the really juicy stuff won't happen until the weekend.)
 
Sunder said:
The only question now, is have I bought the deal of the century, or have I been ripped off? Right now, being at work, I couldn't even tell you if I've bought a 66 x 160 cylinder full of sand... But if this kind of thing interests you, stay tuned for my testing! (Will probably do a few quick tests tonight, but the really juicy stuff won't happen until the weekend.)

Check back with us in a few years to see how they are holding up lol. Long term testing is the only real way to know if you got a good deal or not. But in the mean time, have you done any actual capacity testing yet?
 
Unless something major has been left out of the info I casually found on the web, these are not very size appropriate for EVs, however...

I keep hearing about the 10K cycles figure, and even if its actually half that, these sound wonderful for time-shifting energy from a sunny day collecting solar PV, and storing it to power your home for a few hours after you get off work. I think their low price might just be because most people don't know much about them, plus these are slightly "used" creating fear of a bum purchase?

The people most willing to take a risk on used cells of any type seem to be EV builders, and a used Tesla model-S pack would be a much better risk. (I must admit I am curious, though...)
 
Okay, so I charged one up, and it took 35Ah to fill it from empty. That's pretty promising. It's on a discharge cycle now, but with a 5A discharge limit, this will take a while.

I have set the termination voltage to 1.5v at the charger, which is usually a "real" 1.7v at the battery. So I should be getting about 95% of the available capacity. Will report back tomorrow morning with the results.
 
Sunder said:
About 3 weeks ago, I took a $550 USD punt. I bought 12 of these:



They arrived neatly packed today. These are Lithium Titanate batteries.

At $22 per 30Ah + shipping, these units are ridiculously cheap at about 70c per Ah. The going rate is between $2-3 per Ah.

These are rated for 300A charging and discharging, with a shelf life of 20+ years, and 30,000+ cycles. Of course, their energy density for both volume and weight is terrible, so they are totally not suited to EV or eBike use, but this is an experiment for off grid power. ONE of these cells weighs a kilo, and is the diameter of a coke can, and about 50% taller - and you need 11-12 to make a 24v battery

The only question now, is have I bought the deal of the century, or have I been ripped off? Right now, being at work, I couldn't even tell you if I've bought a 66 x 160 cylinder full of sand... But if this kind of thing interests you, stay tuned for my testing! (Will probably do a few quick tests tonight, but the really juicy stuff won't happen until the weekend.)

30.6 Cents per Wh seems like a great deal to me.
Well LTO cells are used in Mitsubishi MiEV IRCC...
So this is around 72 Wh per kilo, which I think is on the higher side for LTO.
Looking forward to your testing resulta.
 
Well, it's passed the first test:

Charger_no_exif.jpg

Unfortunately, ACIR is reading about 16mΩ, so way too high to be able to deliver the claimed 10C reliably. This doesn't mean DCIR is that high, but I thought as a general rule, ACIR is LOWER than true DCIR.

I had hoped to string a few 1000W halogen lights in parallel to get a low resistance load, but doing the maths last night, it looks like I'd need way too many to get even a 1C load, let alone a 10C load. My only other option now is to build it into a pack, and do whole pack internal resistance testing, which is less accurate.

If anyone has any better suggestions on how I can generate a high load on just 2.4v, I'd appreciate advice.
 
Sunder said:
Well, it's passed the first test: Unfortunately, ACIR is reading about 16mΩ, so way too high to be able to deliver the claimed 10C reliably. This doesn't mean DCIR is that high, but I thought as a general rule, ACIR is LOWER than true DCIR.
I am under the same impression. For instance VTC4 NMC cells have 12 mOhms 1kHz AC impedance new, across the whole SOC range, while DCIR varies from 22-24 in the 100-20 % SOC range... So almost the double.

Sunder said:
I had hoped to string a few 1000W halogen lights in parallel to get a low resistance load, but doing the maths last night, it looks like I'd need way too many to get even a 1C load, let alone a 10C load. My only other option now is to build it into a pack, and do whole pack internal resistance testing, which is less accurate. If anyone has any better suggestions on how I can generate a high load on just 2.4v, I'd appreciate advice.

1. Maybe using a transfmormer coil with good ventilation could dissipate enough heat. Basically using the coil as a very low resistance value. You can calculate resistance of the coil from the wire lenght and wire gauge http://www.powerstream.com/Wire_Size.htm WARNING (RISK OF SHORT CIRCUIT !): Use a thin enough wire gauge (so to have at least a few milliohms resitance, otherwise you will short the battery) and use enough turns or the wire in that coil so the the metal can dissipate enough heat... Watch the temps closely (as temps can play on resistance value of the coil)... I used this method to discharge 138 parralled 18650 from 4.2 to 3.65 volts.... But in my cas, the transformer coil was part of an halogen lamps, so the coil was in parrallel with an halogen bulb (1/Rtot = 1/Rcoil + 1/Rbulb, the bulb reistance being the highest one...) Your cells are much higher amp-hours and amps than mine, so I'd watch closely.

2. Heater element, toaster element : Not sure if it would pull enough load at 2.4 volts. My toaster eats 1500W, but a 120V AC.... Maybe 30W at 2.4V I dunno, as this is DC, not AC...

3. I have an old battery car vacum cleaner that runs on a 2.4V 2S NiCa.... That brushed motor pulls 3.5 amps at 2.4 volts... Tried it with 3.7 volts and it was pulling almost 9 amps !
 
I'm now thinking of one of these:

http://www.ebay.com.au/itm/0-01-500-Ohm-100W-Watt-Shell-Power-Aluminum-Housed-Case-Wirewound-Resistor-New-/282356047987?var=&hash=item41bdbb6073:m:mTRFDFt8UANSfNypzYxi-8w

s-l1600.jpg


At 0.01ohms, it would deliver 240A, or 8C - but that's 576W of power. I'm pretty sure even additional cooling wouldn't save that resistor
At 0.035oms, it would deliver 68.5A or just over 2C or 164W or so. I think a good fan could help that survive long enough to do load testing, and I could get 4 or 5 of these in parallel to get to the 10C rate.

What do you think? Would that work? Any faults in my plans?
 
Sunder said:
I'm now thinking of one of these: http://www.ebay.com.au/itm/0-01-500-Ohm-100W-Watt-Shell-Power-Aluminum-Housed-Case-Wirewound-Resistor-New-/282356047987?var=&hash=item41bdbb6073:m:mTRFDFt8UANSfNypzYxi-8w

At 0.01ohms, it would deliver 240A, or 8C - but that's 576W of power. I'm pretty sure even additional cooling wouldn't save that resistor
At 0.035oms, it would deliver 68.5A or just over 2C or 164W or so. I think a good fan could help that survive long enough to do load testing, and I could get 4 or 5 of these in parallel to get to the 10C rate.

What do you think? Would that work? Any faults in my plans?

It's a good idea to equip yourself with these "tools" in my opinion...

It's funny how we get similar ideas sometimes :D

Last febuary 26th, I order these : http://www.ebay.com/itm/0-5-20-R-Ohm-50W-Watt-Power-Metal-Shell-Case-Wirewound-General-Purpose-Resistor-/311651440908?var=610572169517&hash=item488fdf990c:m:mll8IQNqY1EMg38fSlDR4lA (but couldn't wait to discharge my cells to storage voltage, so I used a halogen lamps in the meantime).

I'm still waiting for those resistors to arrive, but I ordered :
0.5 Ohms ±5% - 50 W
1.0 Ohm ±5% - 50W
2.0 Ohm ±5% - 50W
2.2 Ohms ±5% - 50W
4.0 Ohm ±5% - 50W

I plan to test all my VTC4 one by one, measuring voltage drop and current consumption simultanously, in order to estimate DCIR (it's still not a 100% percent accurate methode unless we do multiples mesurements with multiples resistors/combinations of resistors) and discard bad cells for my pack construction.

The thing is, it's best to measure voltage drops at different current load. But apparently, it's not a good to use the float voltage (no load voltage) @0amp for one of the two measures, as at the extremes, measurment do not necessarly fall on the graph's slope that represents the average cell's internal resistance value. (also keep in mind, resistance varies according to temperature too !)

The formula is R = (V1 - V2)/(I2-I1) see here https://technick.net/guides/electronics/bpw/c09_02/

DC Load Test.jpg

The thing is, under high load, voltage will drop AND CONTINUE TO DROP not just because of the load and resistance, but also from the cell discharging itself ! And we have to take that into account

So I believe there is a sweet spot to find between too small load, small load, heaver load and too heavy of a load (to small value resistance will burn, or 18650 cell will get damaged from excessive discharge rate).

Hence I ordered many different values resitors to experiment with them, finding the sweet spot.

Some good videos related to the subject :
[youtube]Uz4X8i9ubeA[/youtube]
[youtube]3OEJ6eok-Zs[/youtube]

NB : Il you parrallel two resistor of same value each , the overall resistance will be twice as small i.e : you'll have double the load on the battery, but each resitance will still see the same watts and current...
BUT, Put them in SERIES, and resitor add up, two same resitor in series and you'll have double the resistor, half the load... For the power consumed, we have to know what the voltage "across each resistor is" and caclulate power from P = V across x I.

I would buy many of these, and also some of different values so you can make combination to get virtually any value you want (parallel combination ; series combinations or both)...
 
I posted a video some years back from EVTV showing a guy doing a discharge test on a 200Ah Lifepo4 "brick" .
He simply rigged up a pair of buss bars from the terminals, and pushed them to touch each other in a big tub of water !! :shock:
An ammeter read 2000+ amps for a minute or two with a beautiful plasma arc in the water :lol:
Pity i cannot find that vid again. :cry:
 
Hillhater said:
Wire them up, and drop them into a bucket of water !
...but you are going to need a bunch of those (1.0 ohm) to get down to 0.03 ohm ?

They also come in 0.035ohm size. In fact they come in 0.01 ohm size too.

I think I'll pass on the plasma arc. :) I think these with good airflow and short tests - allowing cooling between tests, will do just fine.
 
Hillhater said:
I posted a video some years back from EVTV showing a guy doing a discharge test on a 200Ah Lifepo4 "brick" .
He simply rigged up a pair of buss bars from the terminals, and pushed them to touch each other in a big tub of water !! :shock:
An ammeter read 2000+ amps for a minute or two with a beautiful plasma arc in the water :lol:
Pity i cannot find that vid again. :cry:

Crazy what cells can do ! I recently bought Sony VTC4 cells from DoctorBass. We live in the same city, so I came to pick the Makita battery Pack. TO show me what the cells are capable, he grabbed a old pack he made with some older cells. It was a 1S5P pack, with spot-welded tabs soldered on bussbars.... Just this... (Actually, I think it's this exact one on this picture)...

He use a 0 or 00 welding cable to short the positive bussbar to the negative bussbar while a clamp-ammeter was aroung the beafy copper cable. 765 AMPS BURST !!! FOR 5 cells In parallel !!!! I could not beleive it before i see it.
 

Attachments

  • P1080389_800x600.jpg
    P1080389_800x600.jpg
    48.3 KB · Views: 5,972
Sunder, dont forget to take into account the voltage loss from the normal cell discharge curve, otherwise would will end up overvestimating the internal resistance.

Finally found the video I wanted to show you. This guy talks about it at 7:45 :
[youtube]XHQzIy7OPvo[/youtube]
 
LTO discharge curves are extremely flat. Even at 300A for 5 seconds, a 0.85Ah draw from 30Ah battery is not going to be substantial, as long as I start off with a 2.3-2.4v charge.

If I start fresh of the charger at 2.7v or so, yes, 0.85Ah @ 300A will instantly drop the voltage to 2.4v and stay there. But if I start at 2.40, and draw that amount, it might rebound to 2.39v or so after the test, being nothing more than a rounding error.
 
Sunder said:
LTO discharge curves are extremely flat. Even at 300A for 5 seconds, a 0.85Ah draw from 30Ah battery is not going to be substantial, as long as I start off with a 2.3-2.4v charge.

If I start fresh of the charger at 2.7v or so, yes, 0.85Ah @ 300A will instantly drop the voltage to 2.4v and stay there. But if I start at 2.40, and draw that amount, it might rebound to 2.39v or so after the test, being nothing more than a rounding error.

Sounds good ! Can't wait to see the results !
I'll also post my results with VTC4, when I can finally get my hand on those resistors... Dunno if it's still gonna take long for the mail to arrive.
 
For developing a load to load test the cells, I think generally those resistors are way over rated in their power handling?

I've heard of people using heaters or hair dryers because many of those don't care if you hook them up to AC or DC. But obviously be very careful and do your calculations if you do decide to go with a hair dryer or heater.

AC IR is always lower than DC IR, well at least in my experience, although I'm sure their are some strange exceptions.

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

765A from VTC4's is quite a bit. I know a lot of power tool cells can do up to 100C if dead shorted.
 
redilast said:
765A from VTC4's is quite a bit. I know a lot of power tool cells can do up to 100C if dead shorted.

It is ! It was for a fraction of a second though... and he had to try to short it 3 or 4 times so the clamp meter could record the surge !
But 5 cells parallel cells at 2Ah each is 10Ah.... So at 100C, that'd be 1000A !!
 
I dont think the hair dryer load method.... https://endless-sphere.com/forums/viewtopic.php?f=14&t=74567&p=1196522&hilit=Hair+dryer#p1196522
...is an option for single cell tests. Not enough voltage.

Sunder, are those 0.01 ohm resistors still 100W rated ?
..and i was serious about using water as a heat sink for continuous tests. :wink:
 
Hillhater said:
I dont think the hair dryer load method.... https://endless-sphere.com/forums/viewtopic.php?f=14&t=74567&p=1196522&hilit=Hair+dryer#p1196522
...is an option for single cell tests. Not enough voltage.

Sunder, are those 0.01 ohm resistors still 100W rated ?
..and i was serious about using water as a heat sink for continuous tests. :wink:

I believe so. 2.4v would melt that thing in a few seconds though.

I was thinking of using a metal container full of water for more thermal mass, but not putting the resistor in the water. I seriously do not want any arcing to occur.
 
These batteries will be great if they last 30,000+ cycles i'm very interested in your results.

Something to consider.

A cheap and powerful load could be created with lots of air flow and a lot of short paralleled nichrome wires .

Creating a load with nichrome wire is easy.

You can tune it with length and parallel the wires for more amps when using low volts.

Nichrome wires is easy to get for free its inside toaster ovens, hairdryers ect.

Nichrome wire keeps its resistance pretty constant when it heats. Using fans to remove the heat allows you load up the current and turn energy into heat as much as you want.

For example a toaster oven is tuned for lets say 800 to 1200 watts 110 V AC, they do this by the length of the wire.

For DC you would need Double the voltage or cut the wire in half and parallel the connections. 110V dc paralleled to get the same 800 to 1200 w

Cut the same wire into 4 parallel connections and you get same watts at 55v divide it into 8 and 27.5V and so on.

The shorter the wire the bigger the load so active cooling is necessary to keep the wires from melting.

I use crimp connections.
 
Yeh I use nichrome wire for all my test loads. I have two basic setups a high powered and low powered rig. Then I just clamp where I want for the specific draw. Too easy. I can go up to ~2000w without a problem- keeps my shed warm!! Or all the way down to just a few amps ( actually I do have a couple of halogens in series for low draw)
I can pop it in series with my RC dis/charger as an extended load for discharging.
It's good stuf.
 
For such a low voltage I'd consider a simple coil of steel bailing wire dropped into a bucket of water as a resistive dump load.
 
mushymelon said:
For example a toaster oven is tuned for lets say 800 to 1200 watts 110 V AC, they do this by the length of the wire.

For DC you would need Double the voltage or cut the wire in half and parallel the connections. 110V dc paralleled to get the same 800 to 1200 w

Cut the same wire into 4 parallel connections and you get same watts at 55v divide it into 8 and 27.5V and so on.
.

This gets a little tricky for us guys down here on 240v domestic mains....900W toasters
..and even more so if you are testing a single 2.4v cell like Sunder has.
I think it would need A lot of parallel sections to get to a 3v DC, 800-1200W system ?
 
Back
Top