Measuring battery internal resistance (always around 50-60 mOhms)

10 posts / 0 new
Last post
keenox
Offline
Last seen: 17 hours 29 min ago
Joined: 02/21/2021 - 16:22
Posts: 4
Location: Bucharest
Measuring battery internal resistance (always around 50-60 mOhms)

Hi,

I recently bought some VTC6 cells and wanted to test if they’re fake. I compared them with known original cells from nkon.nl (HG2, VTC4 and VTC6).
I first measured the DC IR and seem to always get around 50-60 mOhms for all the cells measured. I get about the same values with different loads (0.33, 1, 4, 4.7 ohms) and with different multimeters (Agilent U1232A, 121GW and some no-name HQ Power DVM891). The setup is simple: I first measure the no load voltage, then connect the load to the multimeter leads and measure the voltage once more. As a side note, I also get around 50mOhms with the Opus quick test.
I also tried measuring the impedance, but since I don’t have a dedicated meter I used a 1kHz square wave through a MOSFET (60N03L) with a 0.330 Ohm load. With this setup I got pretty close results to the datasheets, but got around 20-22 mOhms for the new cells I bought.
In this post (https://budgetlightforum.com/comment/1558703#comment-1558703) I see that the DCIR should be ~25 mOhms for VTC6.
Since I want to be sure my measurements are correct, I have doubts about the correctness of the DCIR measurement.
Any idea why I always get around 50-60 mOhms and what I might do wrong?

Thanks!

Edited by: keenox on 02/21/2021 - 16:57
thefreeman
thefreeman's picture
Offline
Last seen: 3 hours 17 min ago
Joined: 01/06/2020 - 09:56
Posts: 389
Location: France

I have made measurements like this, with the noload/load method as you described and also light load/high load method ( See HJK’s article ) the results were reasonably consistant and arround the expected values.

It is strange that you always see the same numbers no matter the cells (have you tested know high DCIR cells ?), are you using four separate contacts ? (As shown in the first pic in ”measuring basics”)

Edit : welcome to BLF Beer

docware
Offline
Last seen: 2 days 5 hours ago
Joined: 12/14/2018 - 12:28
Posts: 31
Location: Czech republic

Hi keenox,
Precise DCIR measurent requires some level of basic knowledge and equipment. Resulting values depend on the method. IEC 61960-3 method is giving 18 – 19 miliohm while 10s method yields 25 miliohm for VTC6. Essential is to use four wire measurement. Voltage measurement contacts have to be separated from current load contacts.

keenox
Offline
Last seen: 17 hours 29 min ago
Joined: 02/21/2021 - 16:22
Posts: 4
Location: Bucharest

@thefreeman: I am not using four separate contacts because, as he says, it’s impossible to hold Smile I used alligator clips from the resistor to probes. After you told me, I also tried with some old Samsung cells and indeed I got higher values.
After some more fiddling I think it also has something to do with the probes. I tried with the cheapo multimeter again and also with the probes from the cheapo mounted on the Agilent and I got values very close to the ones in the table in the other thread.
@docware: I think I do have some basic knowledge of electrical theory Smile Why do you say it is essential to use separated contacts? I am thinking that wires and probes have small enough resistivity and the errors should be small enough. If not, why?

thefreeman
thefreeman's picture
Offline
Last seen: 3 hours 17 min ago
Joined: 01/06/2020 - 09:56
Posts: 389
Location: France

Nah, contact resistance is non negligible, as an example I took a small piece of steel (same material as cell’s terminals) and measured the voltage drop across it with 5A current.

With 4 wires measurement : 4.1mV, stable.

With 3 wires (2 on one side, 1 on the other side : current goes through a crocodile clip attached to the end of the voltmeter probe, the probe is making the contact) : 25~60mV depending on the pressure applied. Edit : when the clip makes the contact it’s even worse, up to 80~100mV

It’s difficult to do 4 wires measurement with two hands but not impossible, if you have 2 other hands to help you it’s easier for sure, or you can use a 4 wires battery fixture (buy or make one).

docware
Offline
Last seen: 2 days 5 hours ago
Joined: 12/14/2018 - 12:28
Posts: 31
Location: Czech republic
keenox wrote:
* @docware:* I think I do have some basic knowledge of electrical theory Smile Why do you say it is essential to use separated contacts? I am thinking that wires and probes have small enough resistivity and the errors should be small enough. If not, why?

I don´t want to be offensive but your question itself is demonstration that you need some more learning. Smile

Voltage drops at contact resistance and cables resistance are usually underestimated in mind of people with only general knowledge about electric circuits. I recommend to start web searching with words like „4 wire resistance measurement“ , „four terminal sensing“, „ Kelvin test probes“.
Shortly, 4 wire measuring is must for such low resistance values.

And I didn´t yet mention temperature effect, voltmeter resolution and accuracy, fact that you don´t know real current value which is changing during your measurement, …….

keenox
Offline
Last seen: 17 hours 29 min ago
Joined: 02/21/2021 - 16:22
Posts: 4
Location: Bucharest

@thefreeman: Can you help me with a link for a DIY 4 wire battery fixture? I ordered some from aliexpress, but it takes a long time to arrive.

thefreeman
thefreeman's picture
Offline
Last seen: 3 hours 17 min ago
Joined: 01/06/2020 - 09:56
Posts: 389
Location: France

This is one I made :

I drilled some aluminium screws that I had and used pogo pins for the voltmeter contact :

keenox
Offline
Last seen: 17 hours 29 min ago
Joined: 02/21/2021 - 16:22
Posts: 4
Location: Bucharest

@thefreeman: Thanks for the idea! Unfortunately I don’t have any aluminium screws available, just regular steel ones. I’ll try to crop up something using copper pipe.
@docware: No worries, but I have used 4 wire measurement in the past. One thing. Let’s assume it’s the fault of contact resistance. That means that under load I will measure a higher voltage, hence this will result in a lower calculated IR, not higher. I was constantly getting higher IR. Am I missing something?

docware
Offline
Last seen: 2 days 5 hours ago
Joined: 12/14/2018 - 12:28
Posts: 31
Location: Czech republic

Yes, you get higher values with 2 wire measurement.

Big problem is that you don´t have constant current load, I´m not sure what values you put into the calculation.

Agilent U1232A should have relatively sufficient resolution. I would recommend use logging if possible. You want to have measuring time repeatedly the same, in your case 10 s probably.

You also should have constant temperature of the cell. Even short holding of the cell in hand affect the result.

If you already ordered fixture, it can arrive during 3 weeks, don´t waste your time on the effort to produce something yourself.