LiFePo4 Battery Voltage drop at high discharge

[aimthiazz@gmail.com]

Hi

I'm using Turnigy Nanotech 3000mAh 20C battery from hk (https://www.hobbyking.com/hobbyking/store/uh_viewitem.asp?idproduct=23826). When the load discharges at 25A, I noticed that there is a serious battery voltage drop (from 6.6V to 4V). Any particular reason why it is happening, since i am operating battery half of it's continuous discharge capacity(20C constant 3Ah so 60A of continious current it must deliver)?

 

[eva-michael]

Single cell low voltage cut off for discharge is 2.0V for LiFePO4. That means two cell should stop discharge when it reachs 4V.
Generally LiFePO4 energy discharge stage is from 3.3V to 2.0V.

 

[aimthiazz@gmail.com]

But Battery voltage should be constant right? (since i'm discharging only 25A, even though they are rated for 60A continuous discharge)

 

[aimthiazz@gmail.com]

After digging little further, I noticed that the batteries which i have is a receiver pack battery, and receiver pack is not designed to give high currents. How true is this statement considering the battery discharge spec is 20C?

 

[dogman dan]

Well, my rule of thumb for real world use on e bikes is cut spec in half. So ideally, if you have some 3c cells, keep the amp spikes below 1.5c on starts.

But then there is hobby king spec, :wink: more like cut by 3/4. So you want to use those at a 5c spike, expecting some noticeable sag then, and run with less amps at cruise speed.

So 3 ah, at 20 amps,, supposedly ok to 60 amps. But cut to 5c, now you have 15 amps. And a 25 amps controller. It will work a lot better with a larger size battery. 10 ah at 5c, 50 amps.

 

[eva-michael]

However, high quality ev power LiFePo4 cell will give a much better performance for high current discharge. Lower quality LiFePO4(actually use cheaper LiFePO4 material and manufacture craft) is mostly rated at Max 1 C.
Even though, it can handle 20C discharge, it will reduce the cycle life of cell faster and make increase temperature of cell largely. Personally, I don't advise for that without proper temperature control solution.

 

[teklektik]

But Battery voltage should be constant right?

Not even close...
And so we get to the missing spec: IR (internal resistance)...

Open circuit battery voltage is always reduced by the internal voltage drop due to the voltage impressed across the internal resistance by the discharge current.

Vsag = (IR) x (discharge current)
so:
(more discharge current) = (more voltage drop)

In your case it appears you are seeing (6.6-4.0)V/25A ~= 100mOhm for 2s or 55mOhm per cell -- which is huge. Headway 38120s are about 5-6mOhm and they are known for being 'saggy'. These cells may be able to deliver the stated C rate, but the resulting voltage drop may well make them unsuitable for high discharge applications. So - the spec may be correct, but they left out the other half of the tale...

 

[Peter.bp]

Headway 38120S (10Ah) cell has less than ~6mOhms only when it is over ~80% SOC...
This cell has internal resistance over ~25mOhms when it is discharged (30% or less SOC)...
When it is half full, I measured 12-15mOhms at current 10A. I did this measurement couple years ago on over 50 cells, randomly taken from 600 pieces batch.
This cell hardly can give more than 3C continuous current even when it is full.
Simple calculation for 10C:
P=I^2*R = 100^2 * 0.006 = 60W. Plenty power loss...

 

[dogman dan]

Bottom line, if you start at 48v on a bicycle, and sag more than 5-6 v under load, you can't use them well. They might be ok for a very low c rate use, like .5c with that much internal resistance, but to get 25 amps from a battery like that, you need 50 ah of it.

2-4 v of sag, that you can live with. (from 48v)

Since you start at 6v, even one volt of sag would be a lot.

 

[aimthiazz@gmail.com]

No, there is always a discharge curve

.

I checked A123 26650cell's discharge curve. I noticed a 0.5V drop at 20C. In my case it is 1.3V drop per cell.That's high.

 

[aimthiazz@gmail.com]

55mOhm per cell

I measured Internal resistance using Reaktor BMS (also from hk :wink: ) The Internal Resistance is 14mOhm per cell at no Load. To compare i checked A123 26650 cell Internal Resistance. It is around 6mOhm. So I guess the batteries which i bought is not a good one. should consider buying high-quality LiFePO4 batteries.

 

[aimthiazz@gmail.com]

@All

Thank you for your inputs, Got some interesting insights. I'm planning to build 10S1P setup (99Wh -33V-3Ah). Is there any off the shelf "good quality" batteries available in the market which i can use.

TIA

 

[999zip999]

What controller and motor are you using ?

 

[aimthiazz@gmail.com]

What controller and motor are you using ?

I'm building a skateboard. controller is VESC and motor is Turnigy SK3 6374, 149-KV.