Internal Impedance Measurement Sharing - Lots of Batteries

Hi All,

I already share the following data in a Taiwanese forum, so there is watermark on the pictures~
http://www.wii.tw/~fogerdis/viewthread.php?tid=48943&page=1#pid253384
Not using a very professional equipment, the impedance meter is named "ZY05L" :
http://www.aliexpress.com/item/ZY05...ltmeter-nickel-metal-hydride/32371737562.html
It uses Kevin clip 4-wires AC sensing, so I think the results still comparable between all the batteries.
Last thing is I use the GW Instek GTL-108A test lead.





Sorry for my bad English .......
Any comments and suggestions are welcome !!

Update:
I'll try to find the datasheet or dealer's description of all the batteries:

As you can see, although it's not really useful for the "DC" application,
the internal impedance value measured by my setup is close to the datasheet or dealer's description.

10/3 Update:
Got 4 AA NiMH speical design for Mini 4WD racing，although the capacity is small (1600mAh)，
but the 1kHz impedance is really quite low (only 4.8mΩ) !! Even lower than Li-ion batteries.
30A discahrge curve by the supplier in post #84:
http://tcfans.org/bbs/dispbbs.asp?page=9&id=2250&boardid=9


26650:
Panasonic - 26650 5000mAh - 27~29mΩ
https://item.taobao.com/item.htm?spm=a1z09.2.0.0.g2tg4C&id=37736658887&_u=fjtd1e8768d
KeepPower - IMR26650 5200mAh - 25mΩ
https://item.taobao.com/item.htm?spm=a1z09.2.0.0.g2tg4C&id=42076860730&_u=fjtd1e8365f
KeepPower - IMR26650 4200mAh - Max. 25mΩ
https://item.taobao.com/item.htm?spm=a1z09.2.0.0.g2tg4C&id=41520919785&_u=fjtd1e862c8

18650:
XTAR - XTVTC4 - N/A
http://www.xtarlight.com/05-chanpin/p-001-1.asp?styleid=335
KeepPower - IMR18650 2500mAh - Max. 18mΩ
https://item.taobao.com/item.htm?spm=a1z09.2.0.0.g2tg4C&id=41358133103&_u=fjtd1e8e05b
KeepPower - IMR18650 3000mAh - N/A
http://www.keeppower.com.cn/products_detail.php?id=509
Sanyo - NCR18650GA - <38mΩ @AC 1KHz
http://www.hurt.com.pl/prods/bat/_li_ion/ncr18650ga.pdf
LG - ICR18650HE4 - <=20mΩ @AC 1KHz
https://cdn.shopify.com/s/files/1/0674/3651/files/LG18650_HE4.PDF?828
LG - INR18650HG2 - <=20mΩ @AC 1KHz
http://keeppower.com.ua/download/2015-06/datasheet-LG_INR18650HG2-1.pdf
Samsung SDI - INR18650 30Q - <=26mΩ
http://www.hurt.com.pl/prods/bat/_li_ion/18650-30q.pdf
SONY - US18650VTC5 - 13mΩ(Typ.) @AC 1KHz
https://cdn.shopify.com/s/files/1/0674/3651/files/VTC5.pdf?740

Others:
KeepPower - 16650 - N/A
http://www.keeppower.com.cn/products_detail.php?id=335
KeepPower - 18350 - N/A
http://www.keeppower.com.cn/products_detail.php?id=359
KeepPower - IMR18350 - Max. 35mΩ
https://item.taobao.com/item.htm?spm=a1z09.2.0.0.6m49mQ&id=41131664831&_u=fjtd1e8ac6a
KeepPower - Sanyo UR14500P - N/A
http://www.houseofbatteries.com/documents/UR14500P.pdf
KeepPower - 16340 - N/A
http://www.keeppower.com.cn/products_detail.php?id=354

Soshine - LiFePO4 16340 - N/A
http://www.soshine.com.cn/a599.aspx
Soshine - LiFePO4 CR2 300mAh - N/A
http://www.soshine.com.cn/a600.aspx
Soshine - LiFePO4 CR2 400mAh - N/A
http://www.soshine.com.cn/a579.aspx

Panasonic - BK-3HCCE - Approx. 25mΩ @ 1kHz
http://www.master-instruments.com.au/files/data_sheets/BK-3HCCE Bulk.pdf
Panasonic - BK-4HCCE - Approx. 40mΩ @ 1kHz
http://www.master-instruments.com.au/cgi/ajax/get_file/63247/1
Tianhao T.H.M - SHV1600 - 6.5mΩ

Hey thank you for posting.

That is an interesting battery tester. I have my doubts about some of their claims, i.e. the odds of the Fluke and their volt meter measuring exactly the same (0.0025% accuracy) is highly suspect and means they probably are calibrating theirs to match the Fluke .. at least at that voltage. That said, measurement functions are pretty good.

It would be great if you could compare your measured ESR versus the manufacturers claims.

What method was used to measure internal resistance? The numbers are meaningless without that information. The results vary widely depending on the method (and equipment/setup, etc).

That's a really good point !!
I'll try to find all the datasheet and update in the 1st post. :thumbsup:

SemiMan said:

Hey thank you for posting.

That is an interesting battery tester. I have my doubts about some of their claims, i.e. the odds of the Fluke and their volt meter measuring exactly the same (0.0025% accuracy) is highly suspect and means they probably are calibrating theirs to match the Fluke .. at least at that voltage. That said, measurement functions are pretty good.

It would be great if you could compare your measured ESR versus the manufacturers claims.

Click to expand...

Actually I already mentioned the equip. & method in the 1st post, I also update the test lead I used.
Equip. name : "ZY05L"
Test lead : GW Instek GTL-108A
Method : Kevin clip 4-wires AC sensing
Please check the link for the detailed information in the 1st post.

Gauss163 said:

What method was used to measure internal resistance? The numbers are meaningless without that information. The results vary widely depending on the method (and equipment/setup, etc).

Click to expand...

Interesting! The only battery I have on this list is the NCR18650GA. I think my Opus 3100 charger measured that (brand new cell) around 40. Obviously not great for measuring resistance, but good enough for a ballpark. My old Eagletac 3400 cells are up to about 120 or so! They've sat idle and charged for almost 2 years, though.

Uxorious said:

Actually I already mentioned the equip. & method in the 1st post, I also update the test lead I used.
Equip. name : "ZY05L", Test lead : GW Instek GTL-108A, Method : Kevin clip 4-wires AC sensing

Click to expand...

If "AC sensing" refers to AC high-frequency (1 kHz) impedance tests then that is a useless metric. See here for explanation.

Really thanks for your information !!
I checked the link you provided, it seems that impedance doesn't do well for the normal 'DC' use.....
But from HKJ's information:
http://budgetlightforum.com/node/20997
Maybe I can use the value to track the batteries' state of health ?
Thanks again !!

Gauss163 said:

If "AC sensing" refers to AC high-frequency (1 kHz) impedance tests then that is a useless metric. See here for explanation.

Click to expand...

I also has Powerfocus BC3100 v2.2 (looks like the same as Opus 3100), but I haven't use it to measure internal resistance yet.
Maybe I'll try it next time and give you some numbers for comparison.

markr6 said:

Interesting! The only battery I have on this list is the NCR18650GA. I think my Opus 3100 charger measured that (brand new cell) around 40. Obviously not great for measuring resistance, but good enough for a ballpark. My old Eagletac 3400 cells are up to about 120 or so! They've sat idle and charged for almost 2 years, though.

Click to expand...

Uxorious said:

I checked the link you provided, it seems that impedance doesn't do well for the normal 'DC' use.....
But from HKJ's information:
http://budgetlightforum.com/node/20997
Maybe I can use the value to track the batteries' state of health ?

Click to expand...

For reasons explained in the link I gave (among others), high-frequency AC impedance tests have little (if any) value for such purposes.

Gauss163 said:

For reasons explained in the link I gave (among others), high-frequency AC impedance tests have little (if any) value for such purposes.

Click to expand...

AC 1Khz impedance tests though are still the industry standard. I've always seen a direct correlation between AC 1Khz resistance tests and DC load curves from cells.

jasonck08 said:

AC 1Khz impedance tests though are still the industry standard. I've always seen a direct correlation between AC 1Khz resistance tests and DC load curves from cells.

Click to expand...

You're comparing apples vs. oranges, i.e. a defect test useful in manufacturing vs. a health test useful to end users.

As explained in the post I linked above, the AC 1kHz impedance test is used during manufacturing because it is a cheap and quick way to screen-out massive failures and help grade cells. Such defect analysis is not so useful to end users. Rather, they are interested in tracking the health of their cells under conditions that are close to actual usage, for which (low frequency) DC resistance test provides much more accurate data. The measurements of the two tests are not simply related, and the former can differ from that latter by an order of magnitude (factor of 10), so it does not generally yield a sufficiently accurate way to approximate the latter.

Gauss163 said:

You're comparing apples vs. oranges, i.e. a defect test useful in manufacturing vs. a health test useful to end users.

As explained in the post I linked above, the AC 1kHz impedance test is used during manufacturing because it is a cheap and quick way to screen-out massive failures and help grade cells. Such defect analysis is not so useful to end users. Rather, they are interested in tracking the health of their cells under conditions that are close to actual usage, for which (low frequency) DC resistance test provides much more accurate data. The measurements of the two tests are not simply related, and the former can differ from that latter by an order of magnitude (factor of 10), so it does not generally yield a sufficiently accurate way to approximate the latter.

Click to expand...

I understand what you are saying. But for me personally if I test a power tool grade battery and I do a 1Khz impedance test on say the Samsung 25R, it registers around 14.8 mili ohms. Then I do one on a VTC5 and it registers about 12 milio ohms. If I do a discharge test and compare the voltage of the cells under load (using an electronic load), the Sony VTC5 will win out and will maintain higher voltage at the same discharge current. So for me anyways with 18650 cells, I've always notice an almost exact correlation between a batteries high discharge output and its 1Khz AC resistance.

Also if I test lets say a NCR18650B (laptop grade cell), it usually is about 38 mili ohms. If I test a NCR18650BE (10A mid-power cell) its around 25 mili ohms. Then as stated above, true power tool grade cells with 20A-30A max discharge ratings will be even lower generally 12-15 mili ohms. The article you link seems to suggest that the AC 1Khz resistance readings can be random and that you cant directly compare a cell from company A to a cell from company B. This may be true with some battery types or configurations, but it is nothing I have ever seen. If you can provide any cell model examples where the resistance reading is high but the cells high power output capability is good, then I'd like to hear about it.

Here is an example of what that article seems to suggest:

A 18650 cell with a 30A discharge rating reads 50 mili ohms on a 1Khz AC impedance test. But it outperforms a 20A rated cell (holds its voltage much better under load) with a measured AC 1Khz impedance of 15 mili ohms. I've never seen a single occurrence out of hundreds of li-ion cells I've measured and tested over the years.

@Jasonck09 The point is that the DC resistance tests are much closer aligned with the type of loads (low-frequency DC) used by most end users, so they will provide much more accurate means of tracking health. Internal resistance depends on many parameters (state of charge, temperature, steady stateness, etc) which makes it difficult if not impossible to accurately correlate AC 1kHz impedance with DC resistance. High frequency impedance may sometimes be useful for very rough analyses, but it will not prove as useful for finer-grained analysis of cell performance under low-frequency DC loads. For example, because it can differ by an order of magnitude from DC resistance, it will not yield a useful estimate of voltage sag under (low-frequency DC) loads, which is one of the most important factors governing health.

Gauss163 said:

High frequency impedance may sometimes be useful for very rough analyses, but it will not prove as useful for finer-grained analysis of cell performance under low-frequency DC loads. For example, because it can differ by an order of magnitude from DC resistance, it will not yield a useful estimate of voltage sag under (low-frequency DC) loads, which is one of the most important factors governing health.

Click to expand...

I'd love to see some examples of where it can be off by orders of magnitude. The site you linked to has some really wild results such as a cell that has roughly 0.8 mili ohms DC resistance has around 15 mili ohms AC resistance. But it does not state the cell type/chemistry used in the test.

10/3 Update:
Got 4 AA NiMH speical design for Mini 4WD racing，although the capacity is small (1600mAh)，
but the 1kHz impedance is really quite low (only 4.8mΩ) !! Even lower than Li-ion batteries.
30A discahrge curve by the supplier in post #84:
http://tcfans.org/bbs/dispbbs.asp?pa...2250&boardid=9

It highly depends upon your application, which kind of impedance measurement you will need to estimate the real behaviour during discharge of a battery. Most applications will not be approximated by any AC tests. If you need low AC impedance you will use a capacitor as near as possible at the device instead of a battery with ultra low AC impedance.
For most applications DC tests are better suited. But most testers ignore, that DC impedance depends on time and temperature too.
An exotic example on how temperature alters impedance is shown in this thread.