• Howdy! we're looking for donations to finish custom knowledgebase software for this forum. Please see our Funding drive thread

Top balancing Lithium cells with Nimh cells

M

mistercrash

10 kW
Joined
Mar 9, 2011
Messages
972
Location
Guelph, Ontario, Canada eh!
Someone explain this to me please.

This is a 45P module of 18650 cells top balanced by three Nimh cells. How does this work?
4465b.jpg
 
It actually should not work. They want 3 series NiMH to self discharge to 1.4V (3x1.4=4.2) and it actually happens, after the are full. Only problem is, until they are not full, they ignore V raise up to about 1.6V, when full, they suddenly start to heat and self discharge. I do not want my Lixx batteryes anywhere (1.6x3=4.8V) :D
 
Beats me what that nimh does for it. All in parallel, it's impossible for any of it to ever be out of balance. All cells will always be at the same voltage unless the cell is damaged. A wrecked cell could be in there with 0v, but you'd only detect it if you cut the connections, or saw it swelling and leaking.

I suppose it would be possible for the nimh to absorb some overcharge of the lico, but the nimh could never be at 4.5v while the parallel connected lico was at 4.2v. It still seems to me to have no real benefit.
 
Thanks for answering, I asked because even with my limited knowledge, it did not make much sense to me. So I thought I'd ask and maybe learn something new.

This guy from Australia made plenty of those modules for his Vectrix VX-1, he also says that he makes Lithium conversions for Vectrix' using this method? I found this on the V IS FOR VOLTAGE forums. He claims he's up to over 30 000 km running this Nimh top balanced Lithium battery he made. :|
 
He is using 3 ni-mh cells to 1 li-ion parallel bank. Never seen it done this way. Interesting concept, would have to be low discharge Ni-mh for me to even use such a setup. Even then I would still prefer a proper balance charger which would inform me of any balance issues. It reminds of those RC lipo battery shunt balancers that you plug into the balance taps to periodically balance your packs, but he is using the Ni-mh to absorb any over voltage conditions in a limited amount.
1.4v x 3 = 4.2v
 
Excuse me? Balancing a parallel pack? Am I that dumb?

Think about parallel, if the packs voltage is brought to say 4.4v, then ALL cells will be 4.4v. It's not going to bleed overcharge into the nimh, unless the whole pack remains overcharged. GET IT? It's PARALELLED. Just because the nimh is undercharged, does not mean it will suck an overcharge out of the lithium. All the cells will be the same voltage.

The nimh won't get overcharged, but the lithium sure as hell could.

If you had a board on there, that opened a connection to lower voltage batteries, THEN it would be a discharger, lowering the voltage of the main pack. But if it's paralleled it's ALL the same voltage. Nothing's going to rush into the nimh just because it could hold more at a higher voltage.
 
Went over there and read the second page at the V you linked to. Did he measure that current he's shunting? Or did he calculate what he fantasizes it's doing?

Oh yes, at some point current flows into the nimh. At the same rate it flows into the pack it's paralleled to. I see no possibility that it does anything.

HE even says, this only works with a set of cells that are all good. Yep it would, many of us bulk charge a set of all good cells with zero problems staying balanced. It works great, there is no reason for the pack to get unbalanced very much.

Works great for solving a non existent problem.
 
Thanks dogman, that was my very first impression seeing this thing all in parallel. I first thought: how can a single cell become unbalanced? But I wanted to ask here just to find out if I was missing something.
 
I think you guys misunderstood, of course that single cell can't become unbalanced within itself. This should actually work like single cell balancers, like any solo balance board.
A problem is that nimh cells start shunting too early, at 4.05 V, if his writing is true.
 
I posted in the thread asking him to explain in more details why he's convinced that his system works. I'm waiting to see if he answers, I know he mentioned before in the thread that he explained his system before in that forum, I searched but couldn't find much.
 
riba2233 said:
parabellum said:
He also mentions have produced new pack, but using BMS this time, I wonder why. :)
Click to expand...

Yeah, but it's sort of BMS, he said that it couldn't read individual cell voltages? Or did I read something wrong?

To bad that he wont actually tell how the new BMS works.
Click to expand...
Sat, 11/30/2013 - 23:27 #9
LithiumVectrix
Offline
Joined: 01/03/2012
Posts:
Points: 59
Re: Paul's 90Ah conversion
The batteries I used are Panasonic CGR18650CG which are rated at 2C continuous and 4C intermittent. These batteries have been used in many commercial electric vehicles. The most well known being the Tesla Roadster. Spec are available on the internet. I have measured the intermittent current consumption of the VX-1 at 230 amps at full acceleration. Panasonic make many different 18650 batteries rated from the 1.1 Amp / Hr for power tools to the long range 3.2 Amp / Hr used in the Tesla Model S. I have already explained the design of the Li-Ion - Ni-Mh battery I used in my first VX-1 conversion both on this site and EV album. It is simple and no secret. I don't ship overseas.
I am a committee member of the " ATA Melbourne EV Branch". Our guest speaker at last months meeting was an Mechanical Engineer who works for Toyota and previously worked for Tesla Motors.
This is our Facebook page. https://www.facebook.com/groups/ev.melb/
There are posts of interest to EV Enthusiasts.
We have a Youtube page that you can watch previous speakers.
http://www.youtube.com/user/ATAMelbourneEVGroup
We are currently organising an EV expo at Swinburne Hawthorn, Melbourne on Sunday 23rd February 2014.
https://www.facebook.com/events/742788465749120/
I have just listed my latest conversion on EV album. This bike has electronic balancing of the battery and does not use the characteristics of Ni-Mh in its implementation. This has saved 7 Kg of weight.
http://www.evalbum.com/preview.php?vid=4802
This is a further development and I won't give detais on this new design.
Click to expand...
Just weight saving I suppose. :lol:
 
Please explain to me how the nimh waits till 4.05v to start shunting, unless there is circuitry that prevents them from taking a charge till 4.05v.

Once you put a bms between the lithium and the nimh, then yes you can bleed into a battery. Most bms just bleed wh into heat.

But with a hard wire connection between the batteries in parallel, the ONLY thing that will happen is when the lithium is at 3.5v the nimh will be at 3.5v. When the lithium is at 3.8v the nimh will be at 3.8v, when the lithium is at 4.2v, the nimh will be at 4.2v.

And when the lithium is at 4.3v, the nimh will be at 4.3v.

It can NEVER prevent the lithium from overcharging. You'll just end up with overcharged lithium connected to undercharged nimh. It might be possible that the undercharged nimh will absorb more wh than the lithium at that point. The not flat discharge curve of nimh makes that extremely unlikely. I would have to see wattmeters on both chemistries to believe it's enough to make any difference.

But again, he says himself, that it works only on a pack that doesn't have any reason to get unbalanced. Hey, it's cheap to do, and it makes you feel good. Why not do anything cheap that makes you feel good?

It just gets expensive if you think it's going to save your pack when it does need balancing. :roll:
 
dogman said:
Please explain to me how the nimh waits till 4.05v to start shunting, unless there is circuitry that prevents them from taking a charge till 4.05v.

Once you put a bms between the lithium and the nimh, then yes you can bleed into a battery. Most bms just bleed wh into heat.
Click to expand...

Nimh is at the same voltage as lion, and it starts shunting because of its characteristics. It also bleeds wh into heat, because it's designed that way, just like lead starts to heat over some voltage and thus it self-balances. As nimh starts wasting wh into heat, it also takes that "shunting" current from lion, because they are in parallel.

I think it's something like that :lol:
 
I see it like dogman said it, a whole bunch of cells in parallel is a single cell, and one cell will not have two or more different voltages. I see what you try to explain as some kind of a fight between the two chemistries, like who's going to win in settling the voltage to their liking. I can't imagine any benefit of some chemical war going on in each cell of a battery pack. I still can't wrap my brain around this. My gut just tells me it's wrong :D
 
Agreed I just say it starts doing that at whatever voltage the charge starts at.

Can't see how heating up the nimh before you are at 4.2v or whatever your end of charge voltage is helps any.
 
riba2233 said:
dogman said:
Please explain to me how the nimh waits till 4.05v to start shunting, unless there is circuitry that prevents them from taking a charge till 4.05v.

Once you put a bms between the lithium and the nimh, then yes you can bleed into a battery. Most bms just bleed wh into heat.
Click to expand...

Nimh is at the same voltage as lion, and it starts shunting because of its characteristics. It also bleeds wh into heat, because it's designed that way, just like lead starts to heat over some voltage and thus it self-balances. As nimh starts wasting wh into heat, it also takes that "shunting" current from lion, because they are in parallel.

I think it's something like that :lol:
Click to expand...
You forgot to mention, this magical process does not start until they are fully charged. :lol: Under some conditions we could see 1.6V on NiMH slowly taking the charge and not converting anything to heat, then in some moment they start to heat badly trying to burn everything they have over 1.4V. That is the moment where NiMH charger will stop to not overheat and damage the bats. It does not mean you can not charge them fully at 1.41V, it just need to be done very,very slowly. :D

Theoretically, charging this huge pack with 1/8C of the NiMH string capacyty will never overcharge the pack until some NiMH in the string fails from continuous heating. Its just about 10+ days to full charge :mrgreen:
 
Right, the undercharged nimh won't be making that much heat yet.

I'd have to watch separate wattmeters while it charges to be convinced. One on the wire from the charger, then another on the wire from the lithium to the nimh.

Then you could see how much of the chargers load was absorbed entirely by the nimh.

But the nimh voltage could never increase beyond 4.2v, without the lithium being beyond 4.2v. I think it could be possible that the nimh might give a larger time window for the charger to stop. But I doubt it's a very big window. Quite possibly so small as to be pointless. At top of the lithium charge, the resistance of the nimh might be lower than the lithium. That would divert current to the nimh. Question is, is the nimh lower resistance at that voltage? Is it actually doing this? Nimh doesn't have that flat a discharge curve, so pretty quick it will rise in volts, and there you go past 4.2v.

And in any case, it's completely unable to save a pack with a bad cell in it. AS HE SAYS himself. A bms can do that.
 
I think he'd have a more reliable "balancer" using a giant zener diode than NiMH cells like that. At least those don't generally explode or burn because of internal chemical failures (or crystal growth puncturing cell liners/etc), and are pretty easy to heatsink. :) You'd just have to characterize the actual ones you use to be sure they really do shunt (enough) current at the point you want them to. (and possibly make something to disconnect them once charging is disconnected, so leakage doesnt' drain your pack).
 
You must log in or register to reply here.

Similar threads

65nw
Tesla Module BMS Balance: lowest cells first, jumping values
Replies
0
Views
310
65nw
65nw
B
Experience with EIG ePLB-FO14 Lithium Cells?
Replies
5
Views
299
amberwolf
amberwolf
E
Learning about BMS
Replies
4
Views
235
James340
J
S
Travel friendly 13s1p battery pack using 18500 cells, need cell and BMS recommendation.
Replies
1
Views
673
DaLanMan
DaLanMan
Lucifer
Help Mixing pouch cells
2 3 4 5
Replies
113
Views
4,031
Lucifer
Lucifer
Back
Top