30A LG 18650 cells' capacity test: HB4 vs HB6

thunderheart

100 W
Joined
Oct 8, 2017
Messages
219
Location
Armenia
Hi guys!

I've tested 2 old but still good 18650 batteries from LG - HB4 and HB6. They are both rated at 1500mAh and 30A continuous. 1500mAh seems to be ridiculous nowadays but if you buy them at a good price they can be a good choice for using in many high-drain applications (like power tools, for example).

My Li-ion battery supplier is Queen Battery, as always, and the tests were done using ZKETECH EBC-A20, which supports up to 20A discharge, 4-wire measurement and is PC-connected.
100_9607.jpg


I'm still using the v2.5 of my battery holder which is going to be upgraded to v3.0 soon.
160967_original.jpg


I've followed all the prescriptions of the IEC61960-2003 standard concerning battery's capacity measurement. Before each discharging cycle each battery was charged at standard current mentioned in its datasheet to charge end voltage (4.2V) (cut-off at 0.1A, which is the lowest supported by EBC-A20). Before each discharging or charging i've held a 1-1.5hrs pause. The environment temperature was about 25°C. To be sure in results i've done each test minimum twice.

LG ICR18650 HB4

This guy has the following marking: LGDAHB41865 Q193G124AC.
188277_original.jpg


The main specs from its datasheet:
Minimum capacity: 1500mAh
Nominal voltage: 3.65V
Standard charge current: 0.75A
Maximum charge current: 4A
Charge end voltage: 4.2V
Charge cut-off current: 50mA (std chg) / 100mA (fast chg)
Standard discharge current: 0.3A
Maximum discharge current: 30A
Discharge cut-off voltage: 2.5V
Maximum weight: 46.1g

My sample's measured weight is 43.27g
188668_original.jpg


Test results:
187807_original.png


Good results. At 0.2C the capacity is slightly higher than declared and at higher discharge rates it plays around 1400mAh. Nice curves without ups and downs.

LG 18650 HB6

The HB6 cell i've got is marked LGDAHB61865 Q091D014A1
188866_original.jpg


The main specs from its datasheet:
Nominal capacity: 1500mAh
Minimum capacity: 1400mAh
Nominal voltage: 3.65V
Standard charge current: 0.75A
Maximum charge current: 4A
Charge end voltage: 4.2V
Charge cut-off current: 50mA (std chg) / 100mA (fast chg)
Standard discharge current: 0.3A
Maximum discharge current: 30A
Discharge cut-off voltage: 2.0V
Maximum weight: 48g

My sample's measured weight is 43.13g
189095_original.jpg


Test results:
188008_original.png


Pay attention to the discharge cut-off voltage - it's 2.0V here. At 0.2C HB6's capacity is much more than declared - 1603mAh. There is one interesting moment here - all the way from 5A to 20A the capacity rises but the energy, of course, not. I'd prefer all the batteries to be marked using energy values, not capacity, but we've got what we've got.
1.5Ah and 4.6Wh at 20A discharge looks pretty good, better than 1.4Ah / 4.2Wh of HB4. So let's compare them.

COMPARISON

I'm not going to compare at 0.2C / 0.3A because 0.2C discharge test i'm doing for checking the nominal capacity only (the standard requires that). So let's start with 5A discharge test:
186639_original.png


The curves are quite close. HB6 is ahead with its 1455mAh/5093mWh but even being discharged down to 2.5V it would still be ahead giving out 1435mAh vs HB4's 1401mAh.

10A discharge:
186906_original.png


Here we can see the same picture but with bigger difference. HB6 does a good job!

15A discharge:
187389_original.png


HB6's curve looks better in the beginning and looks more linear compared to HB4's curve. The difference keeps on rising.

20A discharge:
187606_original.png


No doubt HB6 is the winner. Just look at the curves and the numbers.

CONCLUSION

Yes, HB6 is the clear winner and it has 3 advantages over HB4 which has the same rating:
1. HB6 can be discharged down to 2.0V even at 30A, while HB4 is limited by 2.5V
2. Even at 2.5V discharge cut-off HB6 is better than HB4
3. HB6's curves look more linear with less voltage sag compared to HB4.

The only advantage of HB4 is the slightly lower price - US$2.05 vs $2.25 (shipping not included).

I hope this test was interesting to read/watch and i'll be happy if it was useful for someone:) Here is the video version of this review


P.S. Your likes, comments and subscriptions are highly appreciated!
 
Great data, thanks...
A few comments..
🚩 No mention of cell IR.. either AC or DCIR from the tests,..but it seems to be similar for each cell at, 0.025 - 0.030 ohm.
🚩 I cannot share your liking for the 2.0v lv cut off ability ! That is implying a "working voltage range of 2.0+ volts...such that a typical 20S (72v nominal) pack would need a controller/motor system able to function from 40v to 80+volts . :shock:
🚩 The result also highlight again the huge voltage sag typical of most 18650 cells (due to the high DCIR)
🚩 Why do the charts have such odd voltage scale incriments ??? :?:
 
Hillhater said:
No mention of cell IR.. either AC or DCIR from the tests,..but it seems to be similar for each cell at, 0.025 - 0.030 ohm

I just don't have any more or less accurate equipment to measure the IR. All the tests i'm doing at home, i don't have a lab (yet) ))))

Hillhater said:
I cannot share your liking for the 2.0v lv cut off ability ! That is implying a "working voltage range of 2.0+ volts...such that a typical 20S (72v nominal) pack would need a controller/motor system able to function from 40v to 80+volts .

It's not mandatory to use all the 2.0-4.2V range:) But if HB6 cell provides 400 cycles at 4A charge and 10A discharge down to 2.0A, you can be sure that discharging it down to 2.5V would incrase the cycle life. HB4 at 4A charge and 10A discharge down to 2.5A does only 300 cycles. So the 2.0V cut-off is another big advantage of HB6 which makes it even better than HB4.

Hillhater said:
Why do the charts have such odd voltage scale incriments ???

Let's thank the EB Tester Software for that))) BTW, it was even worse. I tried to make it more or less readable.

Hillhater said:
Great data, thanks...

Thank you too:)
 
Hillhater said:
No mention of cell IR.. either AC or DCIR from the tests,..but it seems to be similar for each cell at, 0.025 - 0.030 ohm
thunderheart said:
I just don't have any more or less accurate equipment to measure the IR. All the tests i'm doing at home, i don't have a lab (yet) ))))

Some EBC testers like EBC-A05 can do simple DCIR test. You simply set the load current and it shows DCIR calculated from voltage drop.

Hillhater said:
Why do the charts have such odd voltage scale incriments ???
thunderheart said:
Let's thank the EB Tester Software for that))) BTW, it was even worse. I tried to make it more or less readable.

EBC tester software using graph scaling with fixed 10 rows, so the solution is to set "Curve - Coordinate Settings" 4.4V to 2.4V which gives you 0.1V scale. :?

Hillhater said:
Great data, thanks...
thunderheart said:
Thank you too:)

Agreed, great data, thx :)
 
Pajda said:
Some EBC testers like EBC-A05 can do simple DCIR test. You simply set the load current and it shows DCIR calculated from voltage drop.
They also can charge Ni-MH/Ni-Cd and do increase the current automatically by the given increment, but EBC-A20 does not.

Pajda said:
EBC tester software using graph scaling with fixed 10 rows, so the solution is to set "Curve - Coordinate Settings" 4.4V to 2.4V which gives you 0.1V scale.

When you discharge down to 2.0V it doesn't work :D But it works when you discharge down to 2.5V like in the case of HB4 (look at the first chart in this test)

Pajda said:
Agreed, great data, thx

You too:)
 
thunderheart said:
Hillhater said:
No mention of cell IR.. either AC or DCIR from the tests,..but it seems to be similar for each cell at, 0.025 - 0.030 ohm

I just don't have any more or less accurate equipment to measure the IR. All the tests i'm doing at home, i don't have a lab (yet) ))))
DCIR is a simple calculation from the voltage and current test data.. DCIR = 🔺V/🔺I,...and could be easily generated from the log data .
 
Pajda said:
Some EBC testers like EBC-A05 can do simple DCIR test. You simply set the load current and it shows DCIR calculated from voltage drop.

I'm sorry for being wrong, my A20 also does it with v1.8.5 (2018-02-01) software. Later today i'll publish the IR values for HB4 and HB6.
 
So after fully charging the cells and waiting for more than 30min i started the test. I made 5 iterations at 1C (1500mA) for each cell with 10-15 second interval.
The results:
HB4: 32.7, 30.7, 30.7, 23.3, 27.3
HB6: 23.3, 21.3, 20.7, 21.3, 23.3
 
Hillhater said:
The DCIR will change with the state of charge

Of course it will. That's why i tested both cells at the same SOC - full charge

Hillhater said:
it needs a graph /plot line on the discharge graph.

Taking into account the accuracy of DCIR measuring it's useless. And my equipment can't measure the DCIR during discharging so it's impossible too.

Hillhater said:
Do you have temperature monitoring ?

I have a digital thermometer without logging for monitoring the air temperature.
 
I often measure DCIR with just 2 load resistors and multimeter and no other tools.

You need to take the measurement every 10% SOC or so, and the cell temp must be constant within +-1degC for each reading to be useful data.

Load the cell with a resistor, the voltage drops a bit, load the second resistor, it drops further. Use your meter to observe the change in voltage between adding the second load resistance and removing the second load resistance, do the V=IR math and you get DCIR. Some people use 1sec loaded for "DC", some use up to 10-15sec loaded time, i typically use 5-10sec depending on what best matches the application.
 
Why should i complicate things doing all that manually when the same thing does my EBC-A20?

liveforphysics said:
You need to take the measurement every 10% SOC or so, and the cell temp must be constant within +-1degC for each reading to be useful data.

What's the accuracy of your resistors, multimeter, timer and overall manual measurement? I doubt it's very accurate and i think it's a useless hassle. Just buy a Miboxer C2-4000, set the discharge current and watch the DCIR displayed every 3 second.
 
You're welcome to use what you please. The method described above done with a 6.5digit meter yields plenty accurate results. That little charger and all the hobby pack Ri measurement tools only give useful data for high impedance cells, as they don't drive enough current to get dynamic range on the voltage deltas. The 2 resistor method can easily be sub 1% error, the cheapy testers seldom can achieve even +-5mOhm.


thunderheart said:
Why should i complicate things doing all that manually when the same thing does my EBC-A20?

liveforphysics said:
You need to take the measurement every 10% SOC or so, and the cell temp must be constant within +-1degC for each reading to be useful data.

What's the accuracy of your resistors, multimeter, timer and overall manual measurement? I doubt it's very accurate and i think it's a useless hassle. Just buy a Miboxer C2-4000, set the discharge current and watch the DCIR displayed every 3 second.
 
hey thunderheart
whats your experience with queenbattery been like, have you ordered much stuff through them?
im looking for another chinese lithium wholesaler cause the one I normally use isn't responding (evva).
its a minefield out there!!
cheers
 
liveforphysics said:
The method described above done with a 6.5digit meter yields plenty accurate results.

Even if you could achieve 0.1% accuracy with ultra-extra-super equipment, what's the reason for having such an accurate DCIR data? For what? All I need is:
1. discharge cut-off voltage
2. nominal/max charge current
3. max discharge current
4. curves to see the cell's behavior
5. temperature curves at different discharge rates
6. dimensions and weight

I'd like to have an accurate temperature logger rather than IR meter :)

kdog said:
hey thunderheart
whats your experience with queenbattery been like, have you ordered much stuff through them?

Hi kdog!
I've already made 7 or 8 orders and everything was excellent while i was in Europe. Now i'm living in Russia and the shipping became twice as longer as to Europe, but that's not QB's fault, of course. Russian logistics sucks :(
 
Whats the wattage rating of the resistors that you use?


liveforphysics said:
I often measure DCIR with just 2 load resistors and multimeter and no other tools.

You need to take the measurement every 10% SOC or so, and the cell temp must be constant within +-1degC for each reading to be useful data.

Load the cell with a resistor, the voltage drops a bit, load the second resistor, it drops further. Use your meter to observe the change in voltage between adding the second load resistance and removing the second load resistance, do the V=IR math and you get DCIR. Some people use 1sec loaded for "DC", some use up to 10-15sec loaded time, i typically use 5-10sec depending on what best matches the application.
 
To get an idea of the general state of health your battery is in, rather then just going by reduced Ah capacity. Also you'd want the batteries with the same Internal Resistance all placed together.

thunderheart said:
what's the reason for having such an accurate DCIR data? For what? All I need is:
I'd like to have an accurate temperature logger rather than IR meter :)
 
The wattage of resistors depends on the cell you're measuring. Many times I'm measuring a large format cell with >30Ah capacity and 10C+ rate capability, so the resistors used need to be in the 1kW range for maximum dynamic resolution.

For 18650's, resistors in the 100W range are generally adequate.


markz said:
Whats the wattage rating of the resistors that you use?


liveforphysics said:
I often measure DCIR with just 2 load resistors and multimeter and no other tools.

You need to take the measurement every 10% SOC or so, and the cell temp must be constant within +-1degC for each reading to be useful data.

Load the cell with a resistor, the voltage drops a bit, load the second resistor, it drops further. Use your meter to observe the change in voltage between adding the second load resistance and removing the second load resistance, do the V=IR math and you get DCIR. Some people use 1sec loaded for "DC", some use up to 10-15sec loaded time, i typically use 5-10sec depending on what best matches the application.
 
Back
Top