High and low drain cells mixed

Wolf

Moderator
Joined
Sep 25, 2018
Messages
2,004
100kwh-hunter said:
@ Wolf,
The story of different ir makes sense.
But the low ir (high drain cells) can be easy used with high ir cells(low drain cells)
We use the cells at really low drain, so i think there is no issue/problem here.
The other way around, if i would use the average laptop cell in my power tools, things might get to go sparky.
I found cells in laptop batteries that where the same as in my hilti power tool batteries, yes the one you don't like....begins with a S.
So that can be done, high drain in laptop = low drain application; other way around, no go!
Our powerwalls are basically low drain applications.(mine is down to 150mah a cell and will get lower with extension).
Charging up is a different story, on this one you are correct.
Cloudy day, and then a fullsunon your 10kwh array....and no consumers used at that time, only charging!?!?

What do you think about those thoughts/hypotheses? about high and low drain cells mixed, suited for the application?
If i am wrong just say it, i am in the learning process also.

Thanks in advance.
Started new thread
My thoughts and theories on high VS low drain are still in an experimental stage. As I am also in the learning process.
That is why I built the 4 cell parallel charging and discharging analysis board.
I have run some test as shown in several threadshttps://secondlifestorage.com/t-Locating-Self-Discharging-Cell-s-in-Packs?pid=53594#pid53594 andhttps://secondlifestorage.com/t-Behaviour-of-cells-in-parallel?pid=53209#pid53209but here is one that is interesting.
I pulled 4 cells out of my tested pile that had very close mAh results somewhat different IR and close voltage. I let them sit for ~20min and let them get to know each other. As you can see there was some interaction between the cells. 2 of them stayed even and cell 1 discharging giving power to cell 3.
Now this is in the low milliamp stage I was doing nothing but just observing. Then when I hit the discharge button you can see a large difference in the milliamps the cells are giving up initially.

image_jshjsc.jpg

I will run further tests and investigations on dissimilar cells such as ICR INR etc at low amperage draws such as 250mA per cell which is my build goal and 150mA and 100mA per cell just to see what a low amperage draw does. I will of course also run a 1A draw and possibly a 2A draw per cell to see the difference.
Once done I should have some idea of what actually happens and whether or not a low amperage draw does or does not make much of a difference in a mixed high/low drain cell pack.

Wolf
 
Its only a problem if you dont have enough cells. Ie when you go above and voltage drop occours.

When I started RC for MANY years ago this was common to do. We had 1 small HIGH C rated battery in parallel with the high capacity ones. This to be able to cope with punches.
What happens is easy to already know since its math. ;)

On lets say a 10p pack and lets say 5A per cell its important that they are matched.
At 250mA and 80p pack its quite a difference. Yes they will vary like one cell at 220 and one at 280 but thats not an issue.
At 5A it can easy be 2A and 8A because of closer to its limit.

I personally would not mix tool cells with laptop cells in same normal packs but keep them in different strings. Doing it you can work with the difference by actually just have a thinner wire towards the tool pack :p

If you want some real result go high current. 2A or more. At 100mA and such the result you get can be caused by inaccurazy


What I would like to see is:
At what current does it actually become an Issue with mixing IR between regular cells.
What is regular cells? I would define it as a normal average IR among laptop cells to be that. So what current could we say is normal current on them compare to a Tool cell that most likely can cope with 10x the current.

Hard to explain but basically if we stay under 500mA per cell could we consider mixing a no problem for most of good tested cells? Or is 1A a good current?

I have an idea on where we should have here based on many calculations i have made and tests but lets see if yours match mine. I hope so :D

Note that we dont take about difference of 50mA between cells at 1A because thats not an issue. Im talking about a larger swing like 25% or more. Where the cells actually get damaged.
 
daromer said:
If you want some real result go high current. 2A or more. At 100mA and such the result you get can be caused by inaccurazy


What I would like to see is:
At what current does it actually become an Issue with mixing IR between regular cells.
What is regular cells? I would define it as a normal average IR among laptop cells to be that. So what current could we say is normal current on them compare to a Tool cell that most likely can cope with 10x the current.

Hard to explain but basically if we stay under 500mA per cell could we consider mixing a no problem for most of good tested cells? Or is 1A a good current?

I have an idea on where we should have here based on many calculations i have made and tests but lets see if yours match mine. I hope so :D

Note that we dont take about difference of 50mA between cells at 1A because thats not an issue. Im talking about a larger swing like 25% or more. Where the cells actually get damaged.
daromer,
Agreed.
I'm just thinking of a 100p pack with say 5 or10 high drain cells in it. I understand that my 4p board cannot duplicate that but it may give a reasonable representation of what may happen. I do think my next experiment should be a high amperage one though so we'll see.
I just charged 4 INR18650-25R cells up and have some very good UR18650FMs that tested at ~2500mAh. I will put 2 of each in the 4p holder and fire up the iCharger x6 and pull 6A from the pack. :)

Lets see what happens.

On a second note if you look at the graph that I showed initially you will see that when the discharge begins the difference between discharge rates is quite dramatic between the high mA discharge cell 756mA and the low mA discharge cell 515mA there is a288mA difference.

image_bxkqrf.jpg

Wolf
 
But is the initial related to where in the stack they are sitting?

For sure interesting numbers that they inititally do that BUT... There is a simple explanation.

Cells have different type of discharge curves. Even on same chemistry... This will cause it to happen near top charge. If you do the same at lets say 3.9v or even 3.7 the difference wont be the same ;)
 
daromer said:
But is the initial related to where in the stack they are sitting?

For sure interesting numbers that they inititally do that BUT... There is a simple explanation.

Cells have different type of discharge curves. Even on same chemistry... This will cause it to happen near top charge. If you do the same at lets say 3.9v or even 3.7 the difference wont be the same ;)
Oh yea I totally agree with you the difference certainly is not the same especially as they get close to DOD.

image_undbyf.jpg


image_laltdx.jpg


Wolf
 
Very interesting topic. I'm currently building an additional 'battery' of 14s120p. I have mostly 1,500 Panasonic NCR18650As with average IR (according to my OPUSs) in 60mOhm range. I have another 180 MOLI ICR-18650M(s) that have much higher IR (120mOhms) and dramatically different knee on discharge curve.
Both cells are in the 2900mah range.

My max/average charge/discharge are in the range of 600ma to 350ma on each cell.

To avoid spending another set of $$ - my plan is to distribute the 180 MOLIs among the 1,500 Panasonics.... I'm particularly interested in the long term affect - say after 1,000 charge/discharge at 40% DOD - when the cells are degrading.

Any thoughts on this would be welcome... i.e. should I just buy another 180 Panasonics so they are all the same? or just use MOILs and resist Obsessive Solar Compulsion Disorder.
 
Yeah I was under the impression as well that for larger packs it's not that big of a deal but for smallers you really wanna match chemistry. I've been segregating my high drain ones with the hopes that I'll have the time to make a DIY car battery jumper pack to keep in my trunk or to maybe test and use on some of my older power tools, see how that works out. I know Milwaukee has been using a lot of the higher end LG cells now in their new (very expensive) M18 batteries so I've been trying to collect and liberate those when I can. I think ultimately I'm gonna try and make an Ultron-like robot just so he can vacuum and take care of crap around the house. He should be able to lift two metric tons between his 8 arms so I think I'm gonna need quite a few of the INR/IMRs.
 
Like the last chart, very interesting.

The issue with mixing cells is not only with a constant current draw or charge it is the resudual effect of changing current where some cells will see more use as per the initial balancing period where cell 1 discharges to charge cell 3.

Within a pack it is more likely that pulses of discharge (kettle going on for 3 minutes) that are then followed by a recovery / rest period are the times to look at as well.

For testing an interesting plot would be to perform a discharge test that is run with say 3 minutes discharge and then 3 minutes no load all the way down to minimum voltage. Then see how each cell recovers at different voltages (balancing flows)


Annotated the chart, hopefully I have the cells right from the color coding in the first chart.

This is to cause a response...

image_mdmvxf.jpg
 
OffGridInTheCity said:
To avoid spending another set of $$ - my plan is to distribute the 180 MOLIs among the 1,500 Panasonics.... I'm particularly interested in the long term affect - say after 1,000 charge/discharge at 40% DOD - when the cells are degrading.

The previous post chart should give you an idea of what can happen mixing vastly different IR cells, buit the exact chemistry behaviour would need a Wolf test...
 
OK new test results of the 6A Discharge for the 4 pack with 2 UR18650FM and 2 INR18650-25R.
Another interesting chart.
Note for clarification:
The numbers in front of the cells in the annotation is my number for these cells as they are recorded in my spreadsheet.


Observation
Very little interaction between the cells at insertion.
Total DischargeAh for the pack according to my Icharger x6 9.411Ah

image_iaoqyb.jpg

Here is the discharge curve. The expression 2 peas in a pod comes to mind.

image_uirnmn.jpg


Next are 4 of the same brand. 2 INR18650-25R and 2 UR18650FM discharge at 1A for the pack
Sneak peek of the first hour. ;)


image_mxawzm.jpg


Wolf
 
Very cool. Would be interesting to see the Voltages overlayed with the Current.
 
Huh I would have expected a bigger difference in the Voltage drop when there was a difference in current. Interesting But still very cool
 
jdeadman said:
Huh I would have expected a bigger difference in the Voltage drop when there was a difference in current. Interesting But still very cool
jdeadman,
You have to remember the cells are all on a common buss so each cell influences the other and voltage is the common denominator.
The Amp draw is the same no matter what so each cell gives up what it can at that voltage.
It just shows that the 2 low IR cellsand the 2 high IR cells work together in unison
Again this is a parallel test not an individual cell test.

Note: You can always look at these tests as they are happening.http://wolftech.mynetgear.com:4562/d/gsj7aXDZk/cell-analysis?orgId=1&from=now-6h&to=now&refresh=5s

username guest password guestw
Wolf
 
Wolf said:
It just shows that the 2 low IR cellsand the 2 high IR cells work together in unison
Again this is a parallel test not an individual cell test.
This makes me think that IF you are going to mix IR in a pack - its better to do it 50/50 rather than 2/100. Because if you do 2/100 - then 2 of the cells are going to be really stressed (relative to the overwhelming 98) as they are a very small minority. In other words, in my case, distributing 13 MOLIs with 107 Panasonics per pack might cause the 13 MOLIs to stress out more rapidly as the 107 Panasonics weigh heavily in the pack?

Does this make any sense?
 
Thing is that the cells getting a "bot more" stressed will most likely wear out and in the end get closer to the other cells in IR and such. This will result in more even packs over time. This of course demands that you have randomized your packs and keept them within reasonable ranges.

Perhaps I should not have said this so early in this test :p
 
OffGridInTheCity said:
This makes me think that IF you are going to mix IR in a pack - its better to do it 50/50 rather than 2/100. Because if you do 2/100 - then 2 of the cells are going to be really stressed (relative to the overwhelming 98) as they are a very small minority. In other words, in my case, distributing 13 MOLIs with 107 Panasonics per pack might cause the 13 MOLIs to stress out more rapidly as the 107 Panasonics weigh heavily in the pack?

Does this make any sense?
Makes absolute sense to me. But more testing is required.
Man will I ever get to build my powerwall :p

The 1 Amp discharge per 4p is coming up to 6 hrs and the chart is looking interesting. I am very curious to see what happens at the end when the cells will show their final gasp at 3.0V.

Wolf

image_yfbksg.jpg
 
Now that's Neat. But only time will tell over many cycles to see of there is any effect to different cells. But Wolf, Big thanks for taking this on and testing.
 
jdeadman said:
Now that's Neat. But only time will tell over many cycles to see of there is any effect to different cells. But Wolf, Big thanks for taking this on and testing.

Yes unfortunately determining what happens with many cycles that is the problem. My 4 cell charge/discharge analyzer can really only give us a representation of what is happening with the interaction of these cells.
But it does give us a good representation of what happens when you mix high and low IR cells.

Didn't know that these cells are artists in their own right what an interesting and pretty picture they paint. Took them from 5:56 AM to 3:39 PM 9 hrs and 43 min to complete.Here is the graphof these 4 cells.

Cell#1 1 INR18650-25R 14.8 4.1721 2527
Cell#2 2 INR18650-25R 14.0 4.1778 2522
Cell#3 7130 UR18650FM 42.1 4.1770 2573
Cell#4 7132 UR18650FM 41.7 4.1798 2562
1 Amp Discharge for the pack
Wolf


image_hxptbp.jpg
 
Wow Crazy. Interesting back and forth depending on DoD Wild difference in current at 12:00
 
Back
Top