Common pack design mistakes, how to avoid?

[Minshull]

Hey I am building a 14s6p pack (x2) and would like anybody’s opinion on my current design with 25mm2 pure copper bus bars, 5amp fuse wire and lgm29 cells (10a discharge) 2850mah

 

[Minshull]

https://imgshare.io/image/ED8Dp

 

[kdog]

Can't really see the image clearly enough to make out all the bits and pieces. Can you make it a bit easier to view. The Yellow is really hard to see.

 

[pwd]

Pwd, that layout has many problems
Complexity, very poor current sharing, insufficient area for series connects.

Then add as many series connects as possible, one for each cell is ideal.. In you previous version several groups had only 3 series connections ~47amps each! Yikes! Thats going to seriously hurt the2 joining P groups
This was just a quick doodle, there is probly improvements that can be made to that layout.
To pull the power out, Iike to take several smaller gauge wires soldered to tabs which are spotwelded on. Ideally one for each cell but as many as possible. Bring them all together to join the main lead out.

Thank you for the feedback and a reasonable solution. I can see how that layout has better current sharing between parallel groups.

 

[Dung]

Hi all

I'm gonna make my battery very soon. Before I do that, I need your help to verify that I'm doing it right. I want a 20s12p battery (13 amp per cell) and have to split it up in 3, 2 x 6s12p and 1 x 8s12p. See pic:

I'm thinking to use 16 awg cable for the interconnections with a length of each cable is about 25 cm and then solder it to 8 awg cable with XT90 connector (6p per cable). This is about 9 meter of cable for interconnections (12*3*25cm). My question is, can I use different length or use the same length from all cells.

Thanks

 

[earthsounds]

How about 8x 20s1 packs?
Would 8x 20s1 each with its own BMS be a good idea for e motorbike with 72v motor?
My thoughts are...
8 or more smaller packs vould be easy to distribute inside the odd shaped box.
If one pack fails I will still have the required voltage.
If one cell in the pack fails, it will be easy to find and repair, or replace the single pack.
The wires for each pack will only have to carry 20A.
Does my reasoning carry some substance, or is there better solution?

 

[KWS Seuren]

How about 8x 20s1 packs?
Would 8x 20s1 each with its own BMS be a good idea for e motorbike with 72v motor?
My thoughts are...
8 or more smaller packs vould be easy to distribute inside the odd shaped box.
If one pack fails I will still have the required voltage.
If one cell in the pack fails, it will be easy to find and repair, or replace the single pack.
The wires for each pack will only have to carry 20A.
Does my reasoning carry some substance, or is there better solution?

Lot's of BMS's aren't doing there job correctly, it's alot of extra electronics that could fail and lot's of extra soldering work... and the bms's need space too.

The main disadvantage is space and lots of extra soldering.

 

[earthsounds]

Thx for reply KWS,
I do not see the negative in your point,
1) there would be much less spot welding and much less strips to be welded, no cross welded strips, just a short and thin strips required from one battery to the next one
2) 'm not affraid of extra soldering at all. I'm actually looking forward to it,
3) extra electronics- every electric car has a lot of it, if this was to prevent worst I'm ok with that.(I think it can even extend the life ot the pack)
4) space for the BMS's- they are quite small, especially the small Amperage ones and I'm ok space wise, having smaller packs will allow me to allocate them within avilable space more easily.

5) the small current from each pack means the wires going to the controller can be much thiner
6) if the single battery fails, or the BMS or the single pack ,IT WOULD BE SO MUCH EASIER TO FIX OR REPLACE.

Am I wrong?

 

[madin88]

I want a 20s12p battery (13 amp per cell) and have to split it up in 3, 2 x 6s12p and 1 x 8s12p. See pic:

The problem of having two or more single packs put together in series, is the voltage drop over the interconnections which could make the BMS to cut off.
If we take as example two 6s packs put in series for 12s wired to a 12s BMS, then the voltage reading under load between cell 6 and 7 would be off.
Lets say your cells sag to 3,2V at WOT and there is 0,5V of voltage drop over the connection wire, then the BMS would see cell 6 to have 3,7V and cell 7 only 2,7V (depeding on where you have soldered the wire). This would make the BMS to cut off while your battery still would have some juice left. This would be also a problem during regen.

I once had two 10s battery put in series to 20s (with about 50-60cm 8AWG wire between), and below 30-40% SOC i got cut outs at WOT (120 amps).
You would need really big wires and lots of them to keep the voltage drop as low as possible if you want to do this.
Or you use a more advanced BMS which has configurable inputs (thats what i was doing).

How about 8x 20s1 packs?
Would 8x 20s1 each with its own BMS be a good idea for e motorbike with 72v motor?

I like the idea of having strings with just "one p", but with 18650 size and 8p i find it a bit much.
Do you know about 20s BMS's which would suit your needs? Because they are usually made for much much higher currents and are quite large therefore.

 

[Dung]

I want a 20s12p battery (13 amp per cell) and have to split it up in 3, 2 x 6s12p and 1 x 8s12p. See pic:

The problem of having two or more single packs put together in series, is the voltage drop over the interconnections which could make the BMS to cut off.
If we take as example two 6s packs put in series for 12s wired to a 12s BMS, then the voltage reading under load between cell 6 and 7 would be off.
Lets say your cells sag to 3,2V at WOT and there is 0,5V of voltage drop over the connection wire, then the BMS would see cell 6 to have 3,7V and cell 7 only 2,7V (depeding on where you have soldered the wire). This would make the BMS to cut off while your battery still would have some juice left. This would be also a problem during regen.

I once had two 10s battery put in series to 20s (with about 50-60cm 8AWG wire between), and below 30-40% SOC i got cut outs at WOT (120 amps).
You would need really big wires and lots of them to keep the voltage drop as low as possible if you want to do this.
Or you use a more advanced BMS which has configurable inputs (thats what i was doing).

Hi madin88

Thx for ur reply and ur real life experience. It helps alot. I was thinking to use equal length smaller wires (16 awg) bundled into a single one as same thickness as a 8 awg wire for interconnections. This is alot of work, now I just want to use 2x 8 awg wire, since it can carry all that current. The lenght of interconnections are about 2x 10 cm.

Do you think its enough? My BMS is https://www.aliexpress.com/item/32826469363.html?spm=a2g0s.9042311.0.0.27424c4doCiW9y

 

[madin88]

If you do it like pictured, the cell on top would see a bit higher current flow compared to the cell on bottom.
Otherwise, one 8AWG wire soldered between cell 3&4 should be OK if you beef up the connection to the other parallel cells, and if you keep it short.
But thats 6p, didn't you say 12p?

 

[Dung]

If you do it like pictured, the cell on top would see a bit higher current flow compared to the cell on bottom.
Otherwise, one 8AWG wire soldered between cell 3&4 should be OK if you beef up the connection to the other parallel cells, and if you keep it short.
But thats 6p, didn't you say 12p?

Sorry for not make it clear enough. I have 12p, but on the picture u only see the half (6p) side of it. The other half is on the opposite side.

 

[eMark]

Got a question on a previous diagram "Perfect Current Share For High Power" from way back when first by Allex ...
<img src="http://endless-sphere.com/forums/download/file.php?id=98636" alt="Image"/>
and then later on by spinningmagnets ... https://endless-sphere.com/forums/download/file.php?id=205250

Question: WHY wouldn't three Series red buss bars halfway between each of the four Series cells further improve energy flow efficiency? Over the long run would three instead of two red buss bars increase the charge/discharge cycle life of the pack (even if minimal)? It certainly shouldn't decrease cell balancing to use three red wires/buss bars in place of just two. So, the only downside is a little more expense and labor, but aren't we all exacting tinker's at heart ?

Do understand the concept that the Series buss bars need to transport more amperage than the Parallel buss bars. I like the current flow diagram, but i've never seen an actual DIY pack with the Series connections (red) located halfway between the Series cells on the Parallel buss bar (where there's a way there's a will). Or is the excuse that spot welding in the middle of a buss bar is a pain ?

What about diagram "Method 4" in this link provided by tsourorf ...
"Cells within a parallel group will inherently be balanced with each other.
(see "Method 4" diagram in this link) ... http://www.smartgauge.co.uk/batt_con.html
... makes sense, but is this more efficient wiring overkill ?

 

[Voltron]

Yes... More buss bars are better... Some just skip the strips and do full sheets. The only downsides are more materials and time costs, but for a boutique battery designed to be abused, worth it.

 

[eMark]

Yes... More buss bars are better... Some just skip the strips and do full sheets. The only downsides are more materials and time costs, but for a boutique battery designed to be abused, worth it.

Your reply is mind boggling (don't be offended it's my mind not yours)

First-off ~ More buss bars aren't necessary if the buss bar(s) is/are the correct thickness/width and the best conductor for the battery pack application :thumb:

Second-off ~ "a boutique battery designed to be abused" is what i'm talking about. For example using a fat tire ebike for backcountry endurance road racing over washboard gravel roads (just an Xtreme ebiking example).

QUESTION: Are one of the series cells on the outside perimeter of your battery pack photo more likely to fail in the negative and/or plus final outside series strip connecton (when under Xtreme duress) OR one of the series cells inside the pack that's farthest from the final negative and/or plus series connection providing power to the controller-motor (all other things equal in an abusive bench test scenario)?

So (again) in this diagram example ... https://endless-sphere.com/forums/download/file.php?id=205250 ... assuming everything (proper buss bars) is about as equally balanced as possible with no one series cell being more susceptible to failure under Xtreme use. However, in your pack photo the final + and - connections delivering power to the controller-motor are on the outer perimeter of the pack at the same end ... not the best design for equalizing current flow for a "battery designed to be abused" (JMO)

Also in your bottom photo aren't the final + and - connections delivering power to the controller-motor supppose to come from the opposite kitty corners of a battery pack (i.e. a pack designed to be abused) ? So going back to this photo diagram ... https://endless-sphere.com/forums/download/file.php?id=205250 ... buss bar series routing conductivity is an important consideration especially when designing a battery pack for optimum Xtreme performance for getting the most number of charge/discharge cycles

 

[litespeed]

This is my latest and greatest battery I built last year. 20s15p 30Q with 8 gauge battery wires and 8mm connectors.......it’s amazingly powerful and gives my bike 100 mile range at bicycle speeds. I’ve pulled 200 amps multiple times and voltage drop is almost nothing comparably speaking. It’s prettier than the one pictured above that a built a few years back with Sony VTC 3 cells.

Tom

 

[999zip999]

Parallel keeps the voltage the series supplies the amperage. That kinda simple. Ampage is heat. I'm going to get hit for thiis.

 

[Voltron]

You've got that backwards... Series provides the voltage, and more parallel lets you draw more amps.

 

[Voltron]

And didn't mean to not give photo credit on those nice packs Tom.

And I didn't say more buss bars are necessary. In a mass production pack, when saving a few cents on materials adds up over thousands of packs it might be a consideration, but I'll take overkill on conductivity any day... It sure isn't hurting anything by having more, aside from slightly higher costs, and weight I suppose.
And yes, in theory the positive and negative should come off opposite ends... But with the copper traces laid onto the nickel strips, it's so conductive it doesn't matter as much, and having both leads come out the same end is more convenient.

 

[john61ct]

Also note that the imbalance issues Gibbo lays out in his SmartGauge appnote are wrt paralleled **strings**, using 6S Pb units as an example.

That sort of layout I would keep the number of "sub-pack modules" paralleled to 3, maybe a maximum of 4, at least without special balancing gear.

**Cells** being paralleled, in 1S units at the lowest level of the layout, do not suffer the same problems.

 

[fechter]

Pack looks good. Very similar to the pack I made for my Sur-ron.

The bus 'bars' in my case are 8ga copper soldered to the nickel. Soldered before welding. If you do the math on the voltage drop across the bus bar at the max current it works out to a very small imbalance.

 

[eMark]

This is my latest and greatest battery I built last year. 20s15p 30Q with 8 gauge battery wires and 8mm connectors.......it’s amazingly powerful and gives my bike 100 mile range at bicycle speeds.
Tom

Could you please explain your awesome 20S15P 30Q cell balance charging procedure. For example cell balance charging arrangement, how often and what balance charger you use/prefer?

Agree that there isn't a traditional BMS that comes close to outperforming balance charging; even moreso on your powerful DIY 200A battery pack

 

[Balmorhea]

Agree that there isn't a traditional BMS that comes close to outperforming balance charging; even moreso on your powerful DIY 200A battery pack

The balancing function of a typical BMS isn't for shoving a goofed-up pack back into a balanced state. It's for keeping cell voltages from going astray to begin with. If you have a pack that needs amps, rather than milliamps, of balance charge, you have bigger issues than BMS vs. balance charger.

 

[eMark]

The balancing function of a typical BMS isn't for shoving a goofed-up pack back into a balanced state. It's for keeping cell voltages from going astray to begin with. If you have a pack that needs amps, rather than milliamps, of balance charge, you have bigger issues than BMS vs. balance charger.

You're assumption that a ES newcomer with minimal posts doesn't already know that and much more shows your echip shoulder bias mentality and prejudice as some kind of ridiculous pecking order as if you're trying to impress someone.

My Question was about that previous awesome battery pack posted photo ... https://endless-sphere.com/forums/download/file.php?id=261787... with its balance charging leads. You apparently didn't even notice them or possibly don't even understand why the DIY builder decided NOT to include a BMS in the photo, but rather balance leads for balance charging ... PAY ATTENTION instead of trying to impress yourself by putting another down to build yourself up.

 

[Balmorhea]

My point wasn't to put anybody down, but to dispute that balance chargers are superior for anything besides beating an abused or defective pack into operable condition.