2020/21 What's the best 100A Peak Pack?

I'm building a Bomber clone with a QS205 V3 50H motor. Initial controller 150A Sabvoton. Have been researching what battery to use for this build. Not terribly worried about range, but want minimum 100A from the battery at 72V nominal.

I've only really found 2 options:

1) Standard answer, a spot welded 18650 pack. Probably 6P or better, 20S.

2) Non standard answer Headway 40152S cells 24S.

My first build used Vruzend and Panasonic cells. I was satisfied. Bluetooth BMS monitoring of cells has shown the pack completely in sync. At rest and under load. 20S 3P delivered 50A intermittent. However, after finishing the build, I determined the time to disassemble and replace a faulty cell would take so much time, I might as well build a new pack. It's even worse, if you use a normal pack and don't do your own spot welding.

But that's a bit of a roll of the dice? Labor wise anyway.

So that got me looking at alternatives. Pouch cells too complicated to interconnect. Prismatic square cells a possibility. Previous experience in marine applications demonstrated at least one brand didn't like being set on their side. Not sure if this is universal, but the BMS manufacturer said they had heard of similar problems.

So that leaves cylindrical cells. In 15-17 Ah, that seems to leave the market to Headway?

I read through all I could find here and other places on Headway cells, including the DrBass thread from 2006-2008. I understand the LiFePo4 chemistry, requires more cells 24 vs 20 for 72V. And the power density is less, requiring more space...and weight tolerance for similar capacity packs. Hard to make work for a triangle pack on a "normal" bike frame. But the 15-17 Ah 40152 cells fit fine in my enduro frame. I'd like to run 2P with the 38120HP cells, but can't make that fit easily.

So...the questions:

1) Is there any other drawbacks to the Headway cells vs. the other choices I missed?

2) Are the 20 or 25 Ah prismatic cells being used anywhere? Realistic max current intermittent?

3) Is there some other option I've overlooked? The Headway solution is 10 years old now...

Thanks!

super low density, true for any LFP

relatively costly

not as good longevity as A123 (LithiumWerks)

BlueSeas said:

My first build used Vruzend and Panasonic cells. I was satisfied.
…
However, after finishing the build, I determined the time to disassemble and replace a faulty cell would take so much time, I might as well build a new pack. It's even worse, if you use a normal pack and don't do your own spot welding.

Click to expand...

I thought disassembly and reassembly with spotwelding was not even an option? too much damage to the cells?

Please explain a bit more what is so fiffecult with the Vruzend?

Have you checked out N.E.S.E?

Spotwelding tabs then soldering wire maybe?

john61ct said:

super low density, true for any LFP

relatively costly

not as good longevity as A123 (LithiumWerks)

Click to expand...

I found small capacity cylindrical cells and pouches available from A123. The small cells have +/- attributes, but still basically an 18650 due to capacity and construction. The pouch cells are priced really right, but if you are going to work them, the construction requires somewhat complicated connections and considerations for compression. Not impossible, but also not DIY friendly.

john61ct said:

BlueSeas said:

My first build used Vruzend and Panasonic cells. I was satisfied.
…
However, after finishing the build, I determined the time to disassemble and replace a faulty cell would take so much time, I might as well build a new pack. It's even worse, if you use a normal pack and don't do your own spot welding.

Click to expand...

I thought disassembly and reassembly with spotwelding was not even an option? too much damage to the cells?

Please explain a bit more what is so fiffecult with the Vruzend?

Have you checked out N.E.S.E?

Spotwelding tabs then soldering wire maybe?

Click to expand...

I think repair is possible with spot welding. There is a whole cottage industry salvaging, and then reselling cells from prior builds. So certainly you could cut out a cell or two and replace. But it requires time. Probably less time than it would with Vruzend, but more equipment.

Is NESE the other option similar to Vruzend from India I think? I have no experience, but assume it would be similar to repair.

The point is...what is an easier alternative to 18650 cells, that are easier to maintain and troubleshoot? Is it really a 10 year old Headway? And if so, can that compete? Or is there an option I haven't found yet.

On the opposite to regular Lithium cells like NMC, NCA, LiPO, LiCO etc, the LiFePo4 cells never realy improoved in energy density.

They keep theyr life cycle incredible and flat discahrge curve.

Personally io made many packs with LiFePo4 and mostly A123 pouch and round cells and i always prefered these for their performances over headway or Thundersky or other cheap.

However i should not forget the greast PING battery from Signalab which from now have to my familly's ebikes, more than 12 years of great service with mosylt BMS problem and rarely had to replace cells over the 7 packs i ordered fro Ping. Ping are regular 1 to 3C LiFePO4.

The headway are great but i still have some cells here that i never used on a projet other than test these for the community.

I have friend who used these and made an awsome built with these but the bike was HEAVY!! 24s2p of Headway!
problem with Headway is that the connections using screw and the extended conductive parts on both end of the cells make them even less energy dense.


the only advantage for headway cells are the easyned to make a pack and i would recommand them for a beginner, not that these are bad, but for ebike i woudl prefer buying savaged A123 packs instead.

Also about spotwelding, yes if you are not so experienced with that and dont have any tool for that i would not recommand that. The old great method i used for all my 18650 packs in the last years was to solder directly on the nickel that is already spotwelded to the cells, just like i shown in the makita pack dissassembly at the end video, it is a very safe, reliable and cost efefctive method for all kind of power: https://youtu.be/Oz6Uis05ewA?t=772


Good luck with your great project.

Doc

My point is that ease of dis/re-assembly is only one factor.

The Headways are great for that, but only that factor.

If fitting a lot less energy in a given space is OK for you, fine.

But A123 is a far superior cell, for all factors except the dis/re-assembly one, if you want LFP

which is inherently less dense than the higher voltage chemistries.

do check out N.E.S.E. most familiar with both say far superior to Vruzend, and can use any 18650.

For best of both worlds, cheap and compact dis/re-assembly

use spotwelding for TABS if you can't find the cells you want already sold that way

and use solder for the interconnects.

Ripping off spotwelds IMO is too damaging.

Same reason to only buy new cells.

You can easily grind off a spot with a dremel sanding wheel.

For pouch cells i got some EIB C020 from jimbob01 here on ES - these are nice and interconnects are super easy with the brackets he also sells.

ES user Mywpn got me to build a 15 Ah 28s LiPo pack for his bike and he loves it. It's not the greatest for range or energy density, but its sure got the amps. Technically OK for 500 A.

Wow...started reading these replies which sent me off researching all the stuff mentioned. 4 hours later it was lunch time. Thanks for the information.

That N.E.S.E system does look easier to build and maintain than Vruzend. Wish it was mass produced and generally available. Hadn't seen that one before.

The difficulty with Vruzend end is the number of steps it would take to get to a cell in the middle of the pack after it was built. Getting to a single cell involves the entire row and all its connections. Of course to find the problem child, you only know from the BMS which "paralleled group of cells" is underperforming. So you have delink the 6P (or whatever you have) to isolate those cells from each other to identify the cell or cells causing the problem.

Perhaps the issue is having 100 or more points of failure? Regardless of how they are interconnected. I bet a lot of packs get trashed with many more good cells than bad ones because of the effort it takes to even identify the good from bad.

Now take a 1P pack with a Bluetooth BMS that is experiencing problems. Charge to max voltage in the highest cell, then discharge to lowest voltage in the lowest cell. Compare all the the voltages on the string at both points, and most likely the problem cell or cells are obvious. Sure the cell might be $30 instead of $5, but 10 minutes later the job is done.

So maybe it's the multiplier effect. You could use the same plan with the N.E.S.E system, just replace all 6 cells in the group for $5 each, then test the 6 removed at your leisure if desired.

Your right, with Headway this comes at a distinct size disadvantage. More on that in another post.

I was planning to argue the intrinsic safety of LFP over other chemistries (except LTO). Earlier research had led me to believe there was almost no "lithium" risk to this chemistry. But looking around to confirm this statement, I found this study:

https://res.mdpi.com/d_attachment/applsci/applsci-07-01314/article_deploy/applsci-07-01314.pdf

In some ways, on the surface, it was an unfair comparison. Appears they charged NMC and LFP both to 5V. But regardless, they got the LFP to runaway. Safer is probably still true, but LFP is still not your fathers D cell battery either.

For grins I wanted to compare density. So I calculated the number of both weight in grams per Ah of Capacity, and then Ah of Capacity per Square Inch. Note I did not equalize for watts, needed due to the different operating voltages. But neither did I adjust for spacing between cells, obviously the small ones

LFP is indeed both a space and weight hog. But there wasn't much difference between the A123 cells and the Headway.

Cell/Weight in Grams/Amp Hours/Grams per Ah/Ah per Sq Inch

Headway 40152 /480 /17 /28.23/ 0.95
NCR18650GA /48 /3.45/13.91 /0.40
A123 26650 /76 /2.6 /29.23 /1.03
SPIM08 /300 /8 /37.5 /1.99
Eplbc 20Ah /410 /20 /20.5 /0.62

jonescg said:

ES user Mywpn got me to build a 15 Ah 28s LiPo pack for his bike and he loves it. It's not the greatest for range or energy density, but its sure got the amps. Technically OK for 500 A.

Click to expand...

That is truly a work of art!

Pouch cells seem to be getting most of their attention because there are/have been some super fire sales on Grade B used cells, making a terrific value proposition. Some came with holders, perhaps making construction easier.

That said, I could not find a single vendor offering a list price to the public for an interconnect and enclosure system for pouch cells. Suspect the places that spoke to the capability would want lots of $$$ for new product.

Anyway, at the end of the day....I'm going to build this one using the Headway 40152S cells, supposedly good for 17Ah now. They come with the holders, screws and bus bars to assemble the pack. Spinning the roulette wheel again ordering from China, but got quotes on Alibaba proper not AliExpress. For batteries at least, that seems like a better system.

I don't know how reliable Headway's capacity or C-rate "ratings" are, but be aware that they may have significant voltage sag under load if pushed near those ratings, which can greatly affect how much actual power you get out of the pack under high loads.

BlueSeas said:

That said, I could not find a single vendor offering a list price to the public for an interconnect and enclosure system for pouch cells.

Click to expand...

The only one I have experience with is the OEM ones that come with the EIG NMC C020 cells.

I don't know of non-OEM interconnect/etc system that's publicly available for sale (except to other OEMs). There's the stuff that automotive/etc packs are built with/as, that could in some cases be used as-is or repurposed, but that's not really what you're talking about.

Nissan Leaf cells are bolt-together "sardine cans" in 2s2p configuration, IIRC, so they're probably the easiest still-produced EV-pack cell to reuse, and are commonly available from scrapped cars. (probably reasonably cheap if you can find someone with a wrecked Leaf that's parting it out, and not too bad from some of the batterypack breakers/resellers).

BlueSeas said:

Now take a 1P pack with a Bluetooth BMS that is experiencing problems. Charge to max voltage in the highest cell, then discharge to lowest voltage in the lowest cell. Compare all the the voltages on the string at both points, and most likely the problem cell or cells are obvious. Sure the cell might be $30 instead of $5, but 10 minutes later the job is done.

Click to expand...

It is also less likely to *have* a failure than the many-multi-parallel-cell packs, simply because of the lack of interconnections, and a cell intended for the purpose to which it's being put (assuming EV-grade cells vs typical 18650s) so it isn't being stressed as hard as the smaller cells (especially when one of the smaller cells does begin to have a problem, or an interconnect fails putting excessive load on the remaining cells).

I use the EIG NMC C020 cells on my trike; they *are* larger and thus less dense because of the OEM cell holder/interconnect system, but they are MUCH easier to assemble and interconnect reliably.

I was planning to argue the intrinsic safety of LFP over other chemistries (except LTO). Earlier research had led me to believe there was almost no "lithium" risk to this chemistry. But looking around to confirm this statement, I found this study:

Click to expand...

Under the right circumstances, no energy-containment device is "safe". It's not even about the lithium...it's about the amount of energy stored inside, which under failure conditions will be released as heat, not only into itself, but into surrounding cells and materials, etc. I've had or witnessed many non-lithium chemistries fail in fire or explosion over the years--even just very large capacitors, and once an inductor under high voltage (that was *spectacular*). If you're interested, Liveforphysics did a thread some years back testing various cells under "adverse conditions"; I think they were all lithium rechargeables.

Amberwolf, thanks for commenting. Some really good points.

I am concerned about the voltage sag. I saw that some on test data, The standard bus bars look a little wimpy for 100+A. So I may drill some thicker alternatives. Beyond that, it is what it is. I am somewhat limiting the system to 100A on the battery side. That's the biggest circuit breaker I can get from Midnight Solar, rated for 150V DC. It won't trip on transients to 120-125A, which is what I will limit to using a Cycle Analyst.

To make pouch cells easier, it doesn't have to be OEM from the cell manufacturer, but could be generic. Just doesn't look like anyone has gone down that path yet.

I will look up the Liveforphysics thread, there is no doubt he was pushing the boundaries.

Overcharging enough to runaway a 4S 12V battery is really hard to do, even if you have only a good regulated charger that maxes out at the traditional 14.4V. But move into 20 or 24S territory, it's far easier, without some sort of cutoff based on high cell voltage. 4S your gonna likely win the gamble over time, but 24S, you must either be the protection system monitoring what's going on or have a BMS that can terminate the charge.

Regardless, like you said, the energy in any ebike battery, even at 36V demands respect.

BlueSeas said:

The standard bus bars look a little wimpy for 100+A. So I may drill some thicker alternatives.

Click to expand...

It's not the busbars, its' the cells themselves, in teh stuff I recall reading back when HW was "the bomb" for new cell types.

To make pouch cells easier, it doesn't have to be OEM from the cell manufacturer, but could be generic. Just doesn't look like anyone has gone down that path yet.

Click to expand...

There's not really a single system (for mounting and interconnection) that could work for many types of pouch cells, because there are many different dimensions (thickness, width, height, tab size and width, even whether they come out the same end or not).

There have been several DIY methods that could be adapted to any cell model, and a few cell-model-specific "kits" that have been sold here on ES over the years in the for-sale section, but nothing really mass-produced.

Doctorbass said:

On the opposite to regular Lithium cells like NMC, NCA, LiPO, LiCO etc, the LiFePo4 cells never realy improoved in energy density.

Click to expand...

Yes the energy density of LFP was major issue for a long time, but recently a significant progress was made and so newest automotive prismatic cells from BYD and others are now attacking 200 Wh/kg and 450 Wh/l values. But it will take some time when this technology will be widely available.

Again, if you do go LFP

Headway is just not even in the same ballpark as A123.

And forget density at the **cell** level

need to look at the whole finished pack.

And yes LFP is indeed **much much** less dangerous than the 3.6-3.7V chemistries.

To the point you almost have to work hard to make it burn.

But nothing is 100% foolproof, even lead banks explode sometimes.

john61ct said:

Headway is just not even in the same ballpark as A123.

Click to expand...

Agreed. But the best cell interconnects a DIY player can make with A123 cylindrical cells aren't in the same ballpark as the easy default method for Headways. So there's some tradeoff either way.

For reasonably cheap, easy to connect, high power density, high energy density, readily available cells, model airplane batteries are hard to beat. But they don't last very long and they have safety issues.

Automotive cells are power dense, cheap, easy to connect, relatively safe, and long lasting, but not energy dense.

Cylindrical flat ended cells can have a lot of different virtues, but easily connecting them into high current packs is never one of those virtues.

IMO choosing a poor value cell just because they offer an easier connection method

is tail wagging the dog.

Choose the best quality cell for your use case, that allows you to get the shape and density you need

then do what it takes to get a well constructed pack made from those cells.

Cylindrical cells can also be bought pre tabbed

You've mentioned pre-tabbed cyllindricals before, do you have any examples or sources worth using? Most of what I see pre-tabbed are either no-name cells of unknown quality, rated for extremely low amp draw, or significantly more expensive than the same cells without tabs.
Seems to me that the convenience is not worth the tradeoffs unless you're building a very small, low power pack. The idea of them being pre-tabbed is appealing though.

john61ct said:

Cylindrical cells can also be bought pre tabbed

Click to expand...

They're still nickel strips spot welded on, not copper bars bolted on. There's no comparison. It may be good enough for most use cases, but it's not a method that gets the best from high current cells.

PHEV2 cells kick ass, but they surely wouldn't if you had to spot weld them.

john61ct said:

And yes LFP is indeed **much much** less dangerous than the 3.6-3.7V chemistries.

To the point you almost have to work hard to make it burn.

But nothing is 100% foolproof, even lead banks explode sometimes.

Click to expand...

Did you read the article I linked to:

https://res.mdpi.com/d_attachment/appls ... -01314.pdf

I felt the same way before, but now I'm not so sure. The more cells in series, the bigger chance of overcharging a cell and starting a chain reaction. So at 12V, 4S is way different than 72V, 24S. Prior to this, I would have considered no BMS on LFP...but far less sure now.

john61ct said:

Headway is just not even in the same ballpark as A123.

And forget density at the **cell** level

Click to expand...

In regards to density, that's just not true. Both Headway 40mm and A123 26mm come in at 1 cubic inch per Ah at 3.2V. That's not in a pack, but you're going to need some cell spacing with any build. And there's a lot of cells in an A123 pack compared to Headway.

Do you have a place where you found any direct comparison between the 2? More people are using A123 today, but that's inconclusive. I would also hope, but far from knowing, that Headway has tweaked the product over the years of experience they have now.

Anyway...I'll have some data soon and will report what I find, good or bad. I have the capability to capacity test cells at multiple rates if I can find the time. It takes awhile to set up for that.