Is Wrongway right about V8S BMS?

[agranner]

Just watched the newest WrongWay video: tear down of a V8S. He states that there is no cell balancing at all, not even passive top balancing. Can this be right? How can a pack survive without at least a per series group over-voltage drain resistor? Has anyone done any exploration of these packs/BMS? 
The battery is what makes me most nervous about riding and relying on these machines. Of course I don’t want to burn my house/office/anyone else, but I also want to ensure I don’t have a battery failure cutout and injure myself. That’s why I’m very diligent in my charging routine to give the pack the best chance. Without top balancing, nothing matters! This is very discouraging! 

[Paradox]

Here is the time stamped video of his claim.

 

[supercurio]

Wait what? 😳
My V10F is open at the moment, I have to check that on its battery pack.

There's empty space on the left side of the wheel where the mainboard is (on the right) Adding a Smart BMS in there would be feasible.

Edited May 21, 2022 by supercurio

[agranner]

I don’t have a ton of electronics experience (I do have some) but I suppose you’d be looking for some kind of array of 20 transistors and resistors chunky enough for at least a little bit of power and maybe a couple of ICs (maybe one per series or one per many) to sense and control the bleed. This kind of circuit would be as dumb as possible as in not communicating it’s states but would still get the job done. 
I’m hoping what he meant was there being no “smart” balancing ie. Cell reporting, balancing at less then full voltage, or some other feature. I agree, this would be great. But we gotta have something in there. 

[supercurio]

Some research first, I don't think I'd open my V10F unless a problem arise, as it's doubly wrapped unlike what Adam found on the V8S.

I did measure the pack's capacity tho in March 2021 after riding it all winter and a bunch of fast charging (4.2A) and got:

• 870 Wh instead of 960 Wh spec
• 11.7 Ah instead of 12.8 Ah spec

That might be in line with expectations with degradation and pack not being drained all the way down to what the cells specs indicate, but I'm still cautious in case top balancing is indeed missing.

 

Here's a good look a the V10F pack and BMS. Can someone identify balancing resistors there? I do see a bunch of them which could be but I am not sure.

 

Edited May 21, 2022 by supercurio

[agranner]

4 minutes ago, supercurio said:

I did measure the pack's capacity tho in March 2021 after riding it all winter and a bunch of fast charging (4.2A) and got:

• 870 Wh instead of 960 Wh spec
• 11.7 Ah instead of 12.8 Ah spec

Just curious: how did you do a capacity test?
 

Decades ago I did RC car hobby. We used to have discharge rigs with a bunch of car signal bulbs for cycling and testing the NiCd 6s1p packs we used. Manual pack balancing was something people did also but I can’t remember how. 

[supercurio]

27 minutes ago, agranner said:

Just curious: how did you do a capacity test?

Using a DL24-P (upgraded with stronger MOSFET after some worst case scenario stress tests)
It's a cheap device, but very useful!

You could use it to drain cell groups down to a specific voltage to manually balance a pack as well.

I would also recommend to wire it with an inline fuse (not as shown in the picture)

Results:

2A test:

1A test:

Note: unfortunately I forgot at which low voltage the BMS cuts out (it does tho)

Edited May 21, 2022 by supercurio

[Chriull]

8 hours ago, Paradox said:

Here is the time stamped video of his claim.

 

In this video at the timestamp and ongoing is no evidence shown of passive balancing missing. Just the backside of the bms is shown, were nothing us to be seen or to be concluded. Wrongway just showed the wrongside of the pcb without the components. ... Or i did not watch long enough ... As i don't enjoy "wasting time on youtube" ...

7 hours ago, supercurio said:

 

I did measure the pack's capacity tho in March 2021 after riding it all winter and a bunch of fast charging (4.2A) and got:

• 870 Wh instead of 960 Wh spec
• 11.7 Ah instead of 12.8 Ah spec

Afaik this arises from cell manufactures specifing capacity available used downto 2.5V, but EUCs stop somewhere around 3.3V to 3.15V. So this is an old well known "discrepancy" since EUCs exist...

7 hours ago, supercurio said:

Here's a good look a the V10F pack and BMS. Can someone identify balancing resistors there? I do see a bunch of them which could be but I am not sure.

 

Here many components are seen on the back side of the pcb which wrongway did not show? Would imho make no sense to skip passive balancing with such circuitry for a company as inmotion?

From my jumping through this youtube video no "classical bleeding resistors array" is obviously and clearly visible. But i did not see real detailed views of all the pcb to analyze. So no real clue again...

As inmotion batteries are not reported to die notable earlier or more often as others i'd suspect this rumours as youtube hype?

But imho best to ask directly @Cecily Inmotion for a clarification/statement!

 

Edited May 21, 2022 by Chriull

[Chriull]

Just watched the whole wrongway teardown video ;( No reasoning why there is no passive balancing - he just stated that no inmotion wheel has passive balancing...

In the youtube comments he writes he got this information from inmontion directly.

As i cannot really believe this, i just searched through the internet end especially @EcoDrift's technoblog - but a member of the forum was faster and send me a pm with a link to ecodrifts v12 teardown https://ecodrift-ru.translate.goog/2021/10/30/inmotion-v12-second-batch/?_x_tr_sl=auto&_x_tr_tl=en&_x_tr_hl=en-US&_x_tr_pto=ajax,elem showing the BMS  in detail.

This looks very much like like a great BMS with passive cell balancing. A huge amount of Mosfets for ?overcurrent protection? in parallel so they have no chance to overheat - never seen so much in one place! I like this!

The cell monitoring seems to happen by the 5 ics just above the labels B9, B7, B7, B5 (left lower side). For 24 cell (groups) in series this would mean 5 cells are handled by one such ic, the fifth one handles the remaining 4.

Above these ics are many resistors (2x8, 2x10, 2x10, 2x10 and 2x10). So about 4 resistors per cells group. I'd say with some 9x% chance that these are the bleeding resistors for passive balancing - could not imagine anything else these resistors could be good for. And as learned from many other GW/KS/... bms this is how passive balancing circuitry looks like...

Of course no real details can be seen on this picture and no real conclusions can be drawn from it, but by seeing this i'd bet wrongway is wrong...

I personally prefer much more ecodrifts teardown niceley collected at his technoblog - a great source of detail information and knowledge!

[Chriull]

Ps.: Just read what ecodrift wrote in this teardown: "balancing was not detectedBalancing on BMS is not detected."?! Very strange, indeed...

Beside @Cecily Inmotion maybe @Jason McNeil or @WilliamG could clarify this "mistery"?

[mrelwood]

 

On 5/21/2022 at 3:22 PM, agranner said:

How can a pack survive without at least a per series group over-voltage drain resistor?

Exactly, it can’t. The rumor of no cell balancing on any Inmotion has been around for a while, and the arguments have even gotten slightly heated at one point. The ones who believes the rumor can never be convinced otherwise, since the balancing circuit doesn’t look identical to the traditional GW one.

At least some Inmotion BMSs have an IC that to my understanding could be used for low amp cell balancing even without the traditional RT network.

 

I measured the cell groups on my V11 after about 4000km. All groups were within 0.01V or closer (= as close as my meter can measure). That’s simply way too improbable without a balancing feature.

Other than that, some V10F units that have shown symptoms of imbalance, have been fixed by pumping the balancing by cycling the pack between 90-100% a few times. Without balancing it shouldn’t have helped.

 

On 5/22/2022 at 12:55 AM, Chriull said:

But imho best to ask directly @Cecily Inmotion for a clarification/statement!

A while back I asked Cecily to ask the Inmotion engineers to clarify whether the rumor is true for the V11. She said that it’s not true, and that V11 does have cell group balancing. I now wish I had asked about older models as well.

 One part of the rumor is that the Russians asked Inmotion about this as well, and that they said there is no balancing. My guess is that at a time when there might not have been very competent English speakers in the company, both the Chinese and the Russians trying to communicate in English is simply too open for misunderstandings and miscommunication. The situation is different now, since besides Cecily at least some of the engineers are capable in communicating in English as well. I let you be the judge of my own English skills.

 Why Ecodrift keeps the rumor alive though, I have no clue. I guess they are also expecting to see a circuit  identical to the GW’s.

[supercurio]

Okay, so I have suggested Inmotion to post a statement that clarifies which wheel does what. In short: it was all a misunderstanding, lost in translation like @mrelwood anticipated. Thanks @agranner for noticing!

Inmotion Engineers confirmed that the cells can be balanced during charging and riding on current wheels.
However, only the V13 (will) have Smart BMS capability which allow to balance even without charging or riding.

I suppose the main misunderstanding was on the meaning of the word "passive balancing", understanding it as "even when not riding" whereas in this context it means "resistive top balancing" instead of "active balancing" which can shuffle energy between cell groups.

Edited May 23, 2022 by supercurio

[Chriull]

2 hours ago, supercurio said:

However, only the V13 (will) have Smart BMS capability which allow to balance even without charging or riding.

Smart BMS could be a misleading wording again - my ks 16s was sold in 2017/8 already with a smart BMS.

But could be the first EUC with active balancing once it hits the market!

[Pingouin]

Hello,

So I sent an email to ECODRIFT for clarification. I asked them if the V12 (so not the V8S but it is the Inmotion flagship so I would guess other EUCs would not have balancing if the V12 doesn't) has active or passive balancing.
They confirmed to me that there is no balancing...
I also sent an email to EVE Energy CO who makes the V12 battery for confirmation but they haven't replied yet, as for Inmotion I have been unable to get in touch with them.

This is quite confusing because we got different answers.
 

Edited May 23, 2022 by Pingouin

[supercurio]

Salut @Pingouin!

Yes it's odd that ECODRIFT, usually an accurate authority would be mistaken.

Maybe they got the same reply from Inmotion as Adam and that would be enough to confirm their suspicion after not finding 20 or 24 bleeding resistors on the BMS PCB.
Following Adam's advice I asked Inmotion how the balancing is implemented. Since it's not via the same bleeding resistors design as we see on some other BMS. My best guess would be that it's handled by an IC which uses the MOS to dissipate the energy. But I'm just making this up.

[M640x]

Most lithium cells are matched when the packs are built and unless there's a faulty battery, they really don't stray too much as far as voltage is concerned.  A little battery wandering isn't a bad thing, in fact it's normal and won't cause a vehicle to explode as many believe.  A bms is not a requirement for function of an EV.  It's an add on safety device to catch things like wandering voltage, over/under tempurature, etc.  Ultimately helps add to longevity of the pack.

Specifically when talking about these balance type of devices, I like charge only bms's.  That's a bms wired in such a way that during operation of the vehicle it is electronically or mechanically disconnected from the power output and cannot affect it in any way.  Balance devices don't coast to a stop so a fault in the bms leads to a faceplant.

Edited June 18, 2022 by M640

[agranner]

I’ve decided not to worry about it. This has sent me down a rabbit hole of pack building, high current cells, individual cell charger/analyzer, rebuilding tool packs, etc. etc. I’ve learned many interesting (to me) things. But in the end, I’m probably going to smash this wheel trying to learn to do tricks or something. I am sold (as soon as I can scape up the funds) on the Molicel P42a version S18. So that will likely become my commuter and the V8S will just scoot around the neighborhood with my kids. If it breaks, I will open the damn thing and figure it out because that’s what I do with everything I own from my phone to my house.  

[rcgldr]

I bought a V8F last August 2021. I don't ride a lot, only 610 miles on it now, but the batteries still charge to a displayed 83.6 to 83.7 volts, the same as when I first got my V8F.

[lagger]

On 5/23/2022 at 11:37 AM, Chriull said:

Smart BMS could be a misleading wording again - my ks 16s was sold in 2017/8 already with a smart BMS.

But could be the first EUC with active balancing once it hits the market!

personally I hate this word - smart BMS.  I fell like Its a buzzword. Doesnt say nothing technical about how the BMS is designed. BMS has its functions and modes.
Regarding balancing function of BMS. Active is the one that distributes the energy between cells. Passive is the one that discharges the cells by resistor. Balancing can be done in different ways, like on top of charge or actively switching the specific cell group resistors by software algorithms. Then there are things like thermal runaway protection, or undervoltage / overvoltage cell protection and cell monitoring. 

So if I think about it, for me a smart BMS would be any BMS with a software so not the top of the charge balancing systems (done by hardware mostly). So pretty much any EUC has a Smart system right?

Regarding balancing of V8. It looks like the only way to finish the discussion is for someone to look where traces of each cell groups are coming to on the PCB. Find the type of IC and its datasheet.

Edited July 12, 2022 by lagger

[Planemo]

42 minutes ago, lagger said:

personally I hate this word - smart BMS.

I agree, it means absolutely nothing. It's usually used by the layman when a BMS is able to transmit the voltages/resistance of each string of cells but as you say theres a huge amount of variables between the crappiest and the best systems. It's not one or the other.

[Tawpie]

1 hour ago, lagger said:

smart BMS would be any BMS with a software

The marketing department loves you!

I am also in agreement, "smart" is totally a buzzword and it means nothing at all. It wouldn't be terribly difficult to implement a full featured BMS without the use of a microprocessor—but it would be more expensive simply because the parts count would be higher.

Everyone's personal experience and KS's S20 BMS software team can vouch for this gem:
              "Hardware will break. Software comes broken".

Edited July 12, 2022 by Tawpie

[alcatraz]

Where is that series of resistors on the V8 bms? Has anyone found it?

[Chriull]

4 hours ago, alcatraz said:

Where is that series of resistors on the V8 bms? Has anyone found it?

Seems quite everything can be found in the forum :

https://forum.electricunicycle.org/topic/9891-v8-battery-expansion/#elTopicActions_menu

Not too much too see in detail, but way too much components to not include passive balancing.

Really interesting how strong such rumours, born from some misunderstanding can become...

[lagger]

Without detail this is only my guess. Maybe each IC is for a group of 5 cell series. You can see probably 10 resistors bellow so it can be that they use 2 resistors on each series because of power rating.
Again, details and datasheet would solve this for all.

Edited July 13, 2022 by lagger

[alcatraz]

Bingo! 

Thank you