Does internal resistance matter?

Announcement - Help us fight the BOTS! Please report all spam including stuff in your inbox!

batteryq

New member
Joined
Nov 30, 2018
Messages
15
When you build a battery pack, how much ofinternal resistance should you tolerate? Let's say if I plan to use thecells at 1C level, would it be a problem if I pair cells that are 100mOhm from each other? How much is too much?
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,653
batteryq
You asked
"how much ofinternal resistance should you tolerate?"

That my friend is the 64 Thousand dollar question.
There are many opinions out there as far as IR is concerned and opinions count as some of them are tied to real life experiences.
There are several on this board that are researching this question with much fervor and are working on an answer.
I am one of those. I have at this time only collected several weeks of data regarding "Heaters" IR,SD and the testing procedurethat I have used to come up with some datasets.
This is by no means complete and unfortunately not ready to publish as we do not have enough data as of yet, to come up with a proper conclusion. We are at a theory stage as of yet.

So the best I can say at this point is stay tuned and some definitive best practices will come of this I am sure.

The best info I can give you for now is that preliminary data shows most (not all) cells which are measured with a Kelvin 4 wire IR reader that are over 70m?tend to be heaters esp. the Sanyo brand.

Thats all I have for now.

Wolf
 

Crimp Daddy

Member
Joined
Feb 21, 2018
Messages
963
Short answer is variable.

The higher the discharge rate, the more IR matters.

You can lower the discharge rate with more cells in parallel. Alternately, experiment with the cells under load to see what type of voltage drop you get and see if it is within an acceptable range for your application..
 

batteryq

New member
Joined
Nov 30, 2018
Messages
15
Sorry, I just realized I didn't phrase my question properly, there are actually two questions. But I think you guys guessed what I meant:
Question 1: How much resistance is too much at what discharge level? e.g.:
At 1C, IR should be < 300mOhm
At 2C, IR should be < 250mOhm
Question 2: How much variance of IR among the cells should you tolerate? e.g.:
At 1C, all cell's IR should be within 50mOhm from each other
At 2C: all cell's IR should be within 20mOhm from each other

I don't fully understand the impact of the 2nd question, what would happen if I pair two cells with very different IR? Is the smaller IR one going to be empty much sooner and then start sucking power from the larger IR cell?
 

daromer

Moderator
Joined
Oct 8, 2016
Messages
5,659
I would say all between 0 and 200 works fine. Higher resistance do mean they will have issues to cope with the load since u=i*r. As said add more cells and the current per cell is lower and the demand for lower IR decreases too.

First of all if you even plan to go IR route you NEED to have a proper tester. You cannot test with an Opus or Lii tester or equal. You also need to test with a decent current to measure correct drop.

Is it needed? Not for powerwall if you ask me. If you test capacity at max current you will use on a cell you also actually test the IR. because if the cell doesnt get hot then the IR is low enough.

If you want to have proper evidence just read any datasheet or scientific data test out there. There are plenty and i say MANY of them allready done in terms of what IR you should have and plan for. Even IR over time.
 

batteryq

New member
Joined
Nov 30, 2018
Messages
15
Cool website!

Is there any device that can test the IR for you? I understand the basic concept of how to measure the IR using voltage drop etc, but don't really wan to DIY that thing.


Wolf said:
batteryq
You asked
"how much ofinternal resistance should you tolerate?"

That my friend is the 64 Thousand dollar question.
There are many opinions out there as far as IR is concerned and opinions count as some of them are tied to real life experiences.
There are several on this board that are researching this question with much fervor and are working on an answer.
I am one of those. I have at this time only collected several weeks of data regarding "Heaters" IR,SD and the testing procedurethat I have used to come up with some datasets.
This is by no means complete and unfortunately not ready to publish as we do not have enough data as of yet, to come up with a proper conclusion. We are at a theory stage as of yet.

So the best I can say at this point is stay tuned and some definitive best practices will come of this I am sure.

The best info I can give you for now is that preliminary data shows most (not all) cells which are measured with a Kelvin 4 wire IR reader that are over 70m?tend to be heaters esp. the Sanyo brand.

Thats all I have for now.

Wolf



Can't wait for your testing result!
 

Korishan

Moderator
Joined
Jan 7, 2017
Messages
6,635
Yes, there are several devices. A standard DMM can measure them. The opus and liitokala, and i'm sure others, can test the IR.

The higher the IR, the more difficult it becomes to pull higher amps. Most IR ratings come into place at .5C or more. If you parallel enough together, you won't even get .5C, but closer to .1C. Or, about 300mA per cell. This is so low that IR doesn't have much of an effect.

Think of IR as water, energy required to move as Amp, and your movement as Capacity. If you move slowly through water, you can do it all day long with little energy use. Try to walk fast, and it becomes more difficult. Try to turn, and you burn your energy rather rapidly. As you can see in this real world life size example, IR has a major effect the faster you move, or with the cell the more Amps you try to pull.
 

Cherry67

Member
Joined
May 13, 2018
Messages
501
And, fo DIY, You can build an IR Meter or swiper by yourself, given You have an Multimeter with AC mV range, or a simple Oszilloskop Like DSO 138. Look into my Kelvin Thread.
 

Cherry67

Member
Joined
May 13, 2018
Messages
501
Korishan said:
Yes, there are several devices. A standard DMM can measure them. The opus and liitokala, and i'm sure others, can test the IR.

The higher the IR, the more difficult it becomes to pull higher amps. Most IR ratings come into place at .5C or more. If you parallel enough together, you won't even get .5C, but closer to .1C. Or, about 300mA per cell. This is so low that IR doesn't have much of an effect.

Think of IR as water, energy required to move as Amp, and your movement as Capacity. If you move slowly through water, you can do it all day long with little energy use. Try to walk fast, and it becomes more difficult. Try to turn, and you burn your energy rather rapidly. As you can see in this real world life size example, IR has a major effect the faster you move, or with the cell the more Amps you try to pull.
I Am sorry to disagree in the fact that Opus ist able to give accurate IR.
More, No Equipment without 4wire Connection give even remotly accurate and stable results, At least Not in the sub 100 mOhm areas.
 

batteryq

New member
Joined
Nov 30, 2018
Messages
15
Thank you guys for all the great info!

I too find some of the devices which claim to be able to measure IR are not as accurate as one would hoped. I used Liit and zanflare, same battery the IR reported are about 50-100mOhm different, with zanflare always higher than Liit. I am not sure which one to trust.
 

Cherry67

Member
Joined
May 13, 2018
Messages
501
batteryq said:
Thank you guys for all the great info!

I too find some of the devices which claim to be able to measure IR are not as accurate as one would hoped. I used Liit and zanflare, same battery the IR reported are about 50-100mOhm different, with zanflare always higher than Liit. I am not sure which one to trust.

At minimum, do not trust them if it is not 4-wire.


https://en.wikipedia.org/wiki/Four-terminal_sensing

If they are, i can tell its higly dependant from temp, and method.

So, if it is stable, when you do 10 contactings, its most probably good.
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,653
Here is a spreadsheet with the IR and mAh results of 4brand new Samsung 25R cells tested with 6 different testers.
The commercial YR1030 is used as a reference that I trust.
The LiitoKala and the XTAR Dragon seem to be the closest.
Now I do notice that when I check recycled cells esp. the ones that have the remains of the nickel artifacts on the positive and negative terminals the readings are definitely different. Much higher in most cases on all the testers.
That being said the LiitoKala and XTAR Dragon (with its separate probes) seem to still be really close.

Wolf


image_cqkjsr.jpg
 

Cherry67

Member
Joined
May 13, 2018
Messages
501
Wolf said:
Here is a spreadsheet with the IR and mAh results of 4brand new Samsung 25R cells tested with 6 different testers.
The commercial TZ1030 is used as a reference that I trust.
The LiitoKala and the XTAR Dragon seem to be the closest.
Now I do notice that when I check recycled cells esp. the ones that have the remains of the nickel artifacts on the positive and negative terminals the readings are definitely different. Much higher in most cases on all the testers.
That being said the LiitoKala and XTAR Dragon (with its separate probes) seem to still be really close.

Wolf


image_cqkjsr.jpg

Wow, the IR of new cells ar so low..... never seen that. But, i have no new cells... :D

Re Nickel artifacts, I got Research info that the welding points alone can be about 0.8 mOhm, and another piece of nickel strip be valued as nearly 2 mOhm, so on both ends together this gets indeed iteresting when you try to be absolutely accurate.
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,653
Thats why I use the YR1030 as my true reference tool.
It has true 4 wire kelvin probes and you can bypass the nickel stalactites (or is it stalagmites)and artifacts and make contact with
the terminal properly.

image_pnqjcm.jpg



It is also my first tool in my work area lol.


image_yhqreu.jpg



Wolf
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,653
batteryq said:
Can't wait for your testing result!

The results you are waiting for.
After testing 522 cells and recording the results I can now definitely say that IR has a correlation with cell performance.
The results are herehttps://drive.google.com/drive/folders/1UOQUXa4Kwa99KoeuNDe7EV3iSECb8iIs
Spread sheet is calledHarvested Cell Analysis.xlsx and will be continually updated as cells get added.

Just look at the data and you will see how IR affects the capacity of the cell. Cells #1441 to #1487 was an experiment i did with around100ICR18650-22F cells. I set aside the 45 cellswith IR of 60m? or below and tested them. Take a look at the results. No really take a look at them!
I'm not here to convince anyone of the importance or IR measurement to save time and bin cells you know are going to be poor performers and maybe revisit them at a later date. But if the cell has > 1V and the IR is aceptable you will more than likely not have wasted your time IMHO.

As this spreadsheet grows to 1000 and up cells, the answer will become clearer and clearer and users will only have to sort by cell manufacturer or model number to see where their IR stands and the cutoff for performance is.
If you have MS Excel download the sheet and play with it sort by IR, sort by mAh rated and results,etc, build scatter charts and correlation graphs it fun to see how this data can be used.

Again I am not going to try to convince you that the first thing to do is check IR with a Proper Kelvin 4 wire IR measuring deviceand if you got a YR1030 tester it will give you voltage as a bonus.
But I know what my cell harvesting technique will be after I have completed this experiment and I have at least 3000 to 5000 cells in this database There will more than likely be irrefutable evidence how IR matters.


Wolf
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,653
600 cells later I have more graphson IR and capacity.
By now you all know I test all my cells IR and voltage with a YR1030 4 wire Kelvin meter. Just in case you are wondering.
Each graph depicts one cell type and manufacturer. Indicated on the chart title. Well just the cell model number but by now I just about have them memorized.
You can make your own decisions on how IR influences capacity. :p
It does appear that individual manufactures chemistry creates different results.
I can't wait to have 3000 cells in this database. but it's only me with 9 reliable testers. :(
So that may take some time. Winter in Maine keeps me in the house anyway so it may go quicker.
Onward we go. ;)


image_evcfnm.jpg

image_qcjwxe.jpg

image_dzcqhb.jpg

image_iwxyku.jpg

image_zsbykd.jpg

image_ujudfw.jpg



Wolf
 

Chablism

Member
Joined
Jan 27, 2018
Messages
177
Great Job!

We really need to consider how we link such info to the cell database

Maybe we don't need a full chart for every cell but at least the crossover point
 

Wolf

Active member
Joined
Sep 25, 2018
Messages
1,653
Chablis_m said:
Great Job!

We really need to consider how we link such info to the cell database

Maybe we don't need a full chart for every cell but at least the crossover point
Chablis_m,

That is what I am working on I have just recorded my 700th cell. 1634 to 1700 (634 to 700) are waiting for the Testers but IR and Voltage have been tested. Now we wait for the mAh results.
What I'm trying to do is be able to get some baselines that are consistent and the more cells the more accurate this will get.
The results so far are very interesting.
As my initial thought was for 1400 cells for my wall I have now doubled the anti. :D
I have a great supply of packs and will just slave away recording all the cells even though I know by the IR reading it won't perform to what I want. This info that I am gathering will hopefully be very useful to other powerwallDIYers.
Im looking for at least 80% of capacity for my wall cells not just a 2200mAh reading. Any cellwith less than 80%capacity remaining is really no good to me. But with apreliminary IR reading reading of X I should be able to determine if this cell is worth the effort or not.
So thank you for your support and when there is enough data in the sheet I will post these graphs for all the cells that I have tested.
I would like to see at least 200 cells of each variety to be able to create these charts. Obviously the oddball ones I won't have that much info an but the mainstream cells definitely. After that, what the admins do with that information is up to them but the data and charts will be published on my google drive.
I think I will put the URL into my signature so it will be easily found.

Wolf
 

Cherry67

Member
Joined
May 13, 2018
Messages
501
What a luck that you have been hopping in to my 4-wire IR introduction.
I do not have that supply of cells, and i never would have been able to supply that amount of specific and accurate data.

I make a bow for the work you do.
 
Top