diy solar

diy solar

15AH LiitoKala 33140 LFP review and test results

Anybody knows why they say this needs to be spot-welded with laser spot welder? Anybody tried a high-current spot welder?
 
I received 60 cells made in April 2021, I am testing them at 35A constant current discharge, 3.5V end of charge, 2.8v cutoff, and the best is giving 15.2Ah, the worst 14.4, typical is 15Ah. to make this test I just solder 4mm² short wires with 4mm gold connectors on each side, and then connect them to the output of the charger. My soldering iron is a 80W weller, soldering take less than 5 seconds, I do not use a 450°c solder, more 350°c (old one). I am a laser service engineer but I do not have this kind of source for my hobby...
 
Anybody tried a high-current spot welder?
I do not have this kind of source for my hobby...
 
Thanks, I do have a kWeld. What i dont know is if anybody has tried resistive spot-welding on these kind of cells and why they recommend "against" that
 
Thanks, I do have a kWeld. What i dont know is if anybody has tried resistive spot-welding on these kind of cells and why they recommend "against" that
I only know that the issue with adding connectors is doing so without getting the cells too hot.
 
Today the price is 20% higher than the OP, $8 is now $10- at least with shipping to Canada anyways.
Nobody tried the $16 DIY spot welder on these? (note this needs a small resister mod and possible isolating the logic chips) https://www.aliexpress.com/item/32836588140.html

I saw a video showing the cells can heat up to 125 C even with a quick solder job of 3 seconds or less. Typical was 60 C. Since the life of LFP is reduced significantly over 40 C that may be something to consider. How about freezering the batteries before soldering? Maybe the case will heat up and leave the contents frozen while we solder.

Making a 30amp Ebike 52V 16s battery pack with BBSHD Bafang motor, daly 40a bms.
Also will be testing with a shottky diode sbr60a60ct when using 2 packs at the same time in parallel and different SOC.
EDIT: There are nice brackets available for these now. 30 for $5.
 
Last edited:
I tried resistive spot welding with over 1,000amps. So much current that the supply leads (from 460AH LFP bank) leap into the air from the induced magnetic field.
Not successful due to the volume of copper in the tab. So I have ordered a large 300W soldering iron and will try to put some heat sinking to each side of the copper tab and then solder as has been done by some folk here. I would much rather spot weld if I had the gear but laser is out of my price range and would be a one off as I bought 18 of these.
 
How about tinning the copper really quickly with solder and then trying to spot weld it?

What kind of bus bars do you have that can 1,000a without having holes blown into them?
 
If I can tin it really quickly then I shall solder it. Just need a much more powerful soldering iron as 100W and small tip are too small either to tin or solder. I think the spot welding problem is caused by the amount of copper sucking the energy away and preventing it melting and fusing with the nickel. It might also require tungsten or similar electrodes rather than the copper ones I have (which are perfectly okay on nickel to nickel such as 18650 and similar cells).

For bus bars, I use copper bars about 25mm x 6mm. The current is only for a fraction of a second. My 460AH cells can deliver over 12,000amps for a short while if short circuited. Vaporised part of a spanner once......
 
I have a $20 little spot welder and 2mm nickle strips coming soon that I will modify for longevity and 32 33140 cells. I have to say, your experience spot welding them is definitely not welcomed news. I will still give it a go though!
 
I was full of confidence.....until it didn't work. Let us know what happens.
I have 18 cells.
 
Received the cells and they are all identical and the date says 11/21 so these were stickered within 2 weeks or less of my ordering them. All at 3.23 volts give or take 0.01v. So they match very very well from their 1st appearance, the shipping took 6 weeks from my order date and the shipping address is from Ontario, Canada despite my ordering from aliexpress.

There are 2 notable things:

1. The copper terminals now have been laser welded with a sharp little square piece of nickel. So I should be able to spot weld them with a $20 spot welder when it arrives in the mail!

2. The internal resistance is advertised at under 8 m.ohm. The cells all measure exactly 16.12 m.ohms.

Is that still small enough resistance to not be concerned? More resistance = more self discharge I just read on this site, is that true? I know that more resistance = more heat loss when discharging. 8 miliohms sounds like a small amount to worry about but it is DOUBLE what is advertised. How do your cells measure?
 

Attachments

  • 33140 cell signal-2021-12-19-220850_001.jpeg
    33140 cell signal-2021-12-19-220850_001.jpeg
    179.2 KB · Views: 6
  • 33140 cell signal-2021-12-19-220923.jpeg
    33140 cell signal-2021-12-19-220923.jpeg
    166.6 KB · Views: 6
Last edited:
Internal resistance will not affect self discharge. It only affects losses when you draw a lot of current & similarly if charging at high amperage. So, at 20amps, 0.016ohms will drop 0.32volts and cost 6.4watts (so you will lose 6.4 of your nominal 3.2x20=64watts. This will manifest as heat.

Also, I would question how you are measuring this? Have you welded or soldered on contacts because if not, your measurements might not be accurate. It does not take much of za poor contact to add a few milliohms. Also, are you measuring DC resistance or using an AC impedance technique?

The fact that the cells all measure the same is good. However they are shipped with about 20% SOC for longevity and I would therefore expect them to be extremely similar because 20% is on the flat part of the LFP discharge curve.So, do not read a lot into the same voltage. On capacity, mine were all 15-16AH as advertised.

That's interesting about the nickel. That should make things easier!
 
It is a battery operated digital multimeter so that should be a DC measure, I don't have any fancy equip for AC resistance.
Those are some great tips, thanks doggy. When I touch the probes together they zero out so the wire resistance should be adjusted for already.

I will be discharging with an ebike with a 52V @ 30amp controller. Your calculation states a 10% power loss to heat, just in the battery pack. Is there a decimal out of place? I guess 90% efficiency is decent for batteries but then the motor has another 10% too. 6.4 watts of heat loss x 16 cells is 102.4 watts. Then I should be able to ride in the average Canadian winter days (-20C average) assuming the battery started at room temp.

How much resistance do your cells give?

I put the multimeter probe into to 10A connector and measured current. I got a peak of 30amps. That is only 2C. I have no idea is that is the peak current the battery can give, I guess I would have to find a better load.

5C discharge and only 0.33C for charging.
Input Charging Current5A
Continuous Discharging Current75A
Max Current Discharge5C
 
No, decimal place is right. So, power loss is correct but only if your IR is right.

I have 18 cells. I did not measure IR but I did measure capacity. I'm only discharging at 5-6amps so IR is not a concern.

I just grabbed one which I had taken to 80%SOC. The internal resistance is approx 6 milliohms (0.006ohms) at 25degC.

So methinks there could be something wrong with your measurement?

Is the capacity of your cells okay?

30amps is a lot to pull from these cells (1.5kW). Certainly their discharging should keep their temperature up in a horrible -20degC! I think the 5C rating (75A) is a bit "enthusiastic" from Liitokala.
 
My trusty old 20 year old digital multimeter said 16.12 miliohms, it was a decent quality meter from it's day with diode and current testing. I have a new $35 multimeter coming in the mail but that is a month away at best. I just got the 2nd one so I can calibrate my new bench power supply a little more accurately.

I don't have any reason to distrust the meter until now, I hope those cells can perform as good as yours 0.006 ohms is great. What kind of meter are you using, anything special?

I don't know how else I could measure mOhms. I got a precharge resistor, I will try adding it in series and see if that makes a difference when I sum the battery plus the resistor then subtract the resistor ohms from the total.
 
I have a precision multimeter (but not really necessary).

I connected a constant current load of 1amp, waited for the surface charge on the cell to drain off and the cell voltage stabilise. Measure the cell at two points where the current is not flowing (i.e. away from the terminals connecting to the load). Measure voltage. Then remove load and check voltage, reapply load and check voltage. The difference in the voltages for load on and off is the internal resistance (to the precision of the constant current load and the meter).

Of course, the internal resistance will vary with SOC, size of the actual load and cell temperature.
 
Regarding: " I got a peak of 30amps. That is only 2C. I have no idea is that is the peak current the battery can give, I guess I would have to find a better load."
I doubt that you were being restricted by the cell/s. Given the good and consistent reports about the Liitokala original store and this product, I would guess that the cells really can deliver 75 amps. If you got only 30amps, that signified that your load plus internal resistance was 107 milliohms.
I suspect your project is going to be okay. Also, I cannot imagine your bicycle would be pulling 1.5kw very much. That's two horsepower which would presumably make a bike into a mini-rocket (but I have no experience of electric bikes). So presumably your average power is going to be a couple of hundred watts?
 
At full speed the motor will take 1500W from a 52V 16S battery pack. That would be around 35mph. That is constant when riding at top speed on pavement. For off road then the speed varies a lot.

I have been trying to measure the internal resistance with what I could rig together. I got a new model coulometer that I got a discount on (I posted the discount for others in another thread) and I was excited to hook it up but at 3.2V with a tiny load was all I could rig together. This meter tells me the internal resistance of the battery automatically. Need to wait for spot welder to arrive, in the meantime I will also order some alligator clips so I can at least get up to 12v so I can find a compatible automotive load. I had no idea that is how you measured the internal resistance, I thought you just check the ohms between + / -

I am excited that these seem to be new cells. I was thinking ALL the cells on aliexpress were heavily used and had downgraded capacity.
I also ordered some 14 ah rectangular cells. from a seller with a similar name but ppl are reporting only 12ah capacity (used?) and the same price with much slower shipping. The 33140 definitely seem the be the better choice and there are nice modular battery brackets.

Sorry to put you out, doggy, you have shown your thorough knowledge about these cells, may I ask you a direct Q? I wasn't able to find any 52V fuse for the 52V 30amp ebike system. So I went with a 12V automotive fuse I see ppl using for RV solar builds. I ordered a 12V 40amp for the job, and I have a 60amp version onhand already. Do you think these will be able to stop the 52V arcing over the broken fuse conductor? I have a cheap hardware Daly BMS that will be installed but I would like to use the fuse as on ON/OFF battery switch. Also, since FETS on the BMS fail in a closed state, this will protect the battery from a short circuit caused by some crazy vibration water bike trail situation.
 
Last edited:
I think you'll be okay with that device which seems to be a circuit breaker and not a fuse. In the description it mentions 48V even though rated for 12-24V so I do not think 52 will be a problem. Also, it seems to be electromechanical not electronic so unlikely to be voltage sensitive.
 
Back
Top