Jump to content
svenomous

EUC BMS design argument

Recommended Posts

I apologize if I'm about to rehash an old topic. Did a search but even the most focused search terms I could come up with gave me 176 pages of search results containing the words BMS, voltage, cutoff, etc. Feel free to just point me to a thread where this has already been gone over :).

A battery pack BMS performs a set of functions, which I won't list, but which boil down to two purposes: (a) protect the battery pack's health; and (b) protect the rider and environment from explosion/outgassing/fire. This is a good thing, and it's what makes LiIon chemistry generally safe to keep e.g. in my pocket (smartphone) despite energy density and a tendency under certain conditions to be...enthusiastic about energy release. When it comes to EUC, though, there's a also a concern for rider safety in play, and that's something the wheel's mainboard and firmware tries to deal with. The wheel must be kept balancing, and it must use beeps (or KS voice) and tiltbacks (accelerating ahead of the rider, basically) to warn of danger and prevent failure.

I'd like to argue that BMS purpose (a) I listed above is sometimes in conflict with the wheel's overall mission to keep the rider safe. If a wheel with a low battery voltage is moving along and encounters an unexpected situation requiring a sudden burst of power (rider suddenly leans hard, or a bump in the road), and if the motor and wiring and MOSFETs and board are all capable of handling this situation, but drawing the necessary amperage causes the battery pack voltage to sag below the BMS low-volt cutoff, the wheel will instantly die. It won't get to give a warning beep, it won't get to tilt back, it will just turn off, mainboard and all, with predictable outcome for the rider. It seems wrong to me that the BMS should be allowed to enforce the low-voltage cutoff on an EUC. Going below 60V (using 84V pack as an example) is bad for the cells, sure, but isn't rider safety at that moment more important? Shouldn't the voltage be allowed to sag as much as necessary to deal with that bump or lean, and shouldn't the mainboard be given the chance to react by immediately beeping, tilting back to force dismount, and refusing to continue the ride until the wheel is charged (for example)?  If cells are damaged, they're damaged, but I'd rather have that than a broken face or arm.

So I guess I'm saying that custom BMS built for EUC use should do all the things that keep things safe and non-explody/non-combusty, but it should specifically and intentionally omit low-volt cutoff, which is entirely a "protect the pack" function, not a safety function. What am I missing?

Share this post


Link to post
Share on other sites
11 hours ago, svenomous said:

What am I missing?

The fact that since about 2016 the low voltage cutoff has already been omitted from the BMS on all EUCs. So you are absolutely correct, just a bit late. ;)

Share this post


Link to post
Share on other sites
1 hour ago, mrelwood said:

The fact that since about 2016 the low voltage cutoff has already been omitted from the BMS on all EUCs. So you are absolutely correct, just a bit late. ;)

Oh! :huh:

I thought that a wheel cutting out due to low-voltage sag is a thing, even now.  Have never experienced anything like it myself, as I don't push my wheel that hard, but I was sure I'd read about it happening to some people.

Well, never mind, nothing to see here.

Share this post


Link to post
Share on other sites
58 minutes ago, svenomous said:

Oh! :huh:

I thought that a wheel cutting out due to low-voltage sag is a thing, even now. 
 

It’s a vocabulary thing, people call overlean a “cutoff”/“cutout“.

 

Share this post


Link to post
Share on other sites
On 5/1/2020 at 8:47 PM, svenomous said:

So I guess I'm saying that custom BMS built for EUC use should do all the things that keep things safe and non-explody/non-combusty, but it should specifically and intentionally omit low-volt cutoff, which is entirely a "protect the pack" function, not a safety function. What am I missing?

Afair about the times of the Ninebot one E+ low voltage cutoff was a serious thing. In these times @hobby16s BMS topic was hot - how to deactivate the ouptput protection circuit to disable low cell voltage cutoff and overcurrent protection too (as this was a bit too "sensitive")

Low cell voltage cell protection was cancelled completely but by ninebots, which have now (Sx, Zx) smart BMS that monitor singe cell voltages and warn in problematic cases. And refrain from using a wheel with "high differential pressure in battery"😋. Unfortionately they have "misdesigned" the communication between BMS and mainboard so there is constant current draw (vampire drain). Don't know if this is fixed (besde diy) by now?

They have (had?) Some other design issues like with regenerative braking (longer downhill rides) both batteries got out of sync (different voltages) which the wheel could handle - but if turned off in this state it could not be turned anymore...

Current "BMS state" should be summed up in my lately written https://forum.electricunicycle.org/topic/18317-battery-faq/

Input/questions/additional information is highly welcome!

On 5/2/2020 at 9:47 AM, svenomous said:

Oh! :huh:

I thought that a wheel cutting out due to low-voltage sag is a thing, even now.  Have never experienced anything like it myself, as I don't push my wheel that hard, but I was sure I'd read about it happening to some people.

Well, never mind, nothing to see here.

 

On 5/2/2020 at 10:47 AM, mrelwood said:

It’s a vocabulary thing, people call overlean a “cutoff”/“cutout“.

 

Spot on, @mrelwood, as with the prior comment!

This overlean "due to voltage sag" is nicely to be seen in @DjPanJans log https://forum.electricunicycle.org/topic/18325-kingsong-16x-what-current-are-safe/?do=findComment&comment=308083

Battery were already "low" and by the high currents voltage cam even mote down. KS report the max speed (throtteled by battery voltage). And report with the inverter_load (seemongly) quite accurate the wheels overall workload (100% == overlean).

My personal wishlist for BMS:

Must have: Cell Voltage monitoring! Number one values to diagnose a battery packs health! 

Nice to have: Choosable balancing volage threshold in combination with choosable charger max voltage - to be able to get either 80% or 100% charge with safely balanced cells. So one can choose between full range or increased life cycles.

 

Share this post


Link to post
Share on other sites

Per-cell voltage readings would be very useful, and apps (manufacturer app or EUCW/DarknessBot) could simplify for a user who doesn't want to understand the nuances, by just giving simplified "battery pack health" warnings based on thresholds.

Would it make any sense to have some kind of reserve power capability in a wheel, e.g. via a couple of large capacitors, so that when a battery pack is getting low, the system can momentarily draw on the reserve when power requirements exceed the battery's ability to supply the necessary wattage, with an alarm sounding as soon as this emergency reserve has to be tapped?  Would this reduce over-lean cut-outs that are more likely to happen when the battery is at the bottom of its range?

Share this post


Link to post
Share on other sites
2 minutes ago, svenomous said:

Would it make any sense to have some kind of reserve power capability in a wheel, e.g. via a couple of large capacitors, so that when a battery pack is getting low, the system can momentarily draw on the reserve when power requirements exceed the battery's ability to supply the necessary wattage, with an alarm sounding as soon as this emergency reserve has to be tapped? 

This needs space and introduces new failure sources/will age too... :(

2 minutes ago, svenomous said:

Would this reduce over-lean cut-outs that are more likely to happen when the battery is at the bottom of its range?

Imho like KS has introduced the "inverter_load" in firmware (100%== overlean) is the better approach. If this approach is continued with corresponding warnings (tiltback,...) This should make a wheel as overlean safe as can be.

99.99% solution is to take the most powerfull and fastest wheel and limit oneself to about 1/2 lift cut off speed. 

Share this post


Link to post
Share on other sites
5 hours ago, svenomous said:

Per-cell voltage readings would be very useful

Absolutely!

5 hours ago, svenomous said:

Would it make any sense to have some kind of reserve power capability in a wheel, e.g. via a couple of large capacitors, so that when a battery pack is getting low, the system can momentarily draw on the reserve when power requirements exceed the battery's ability to supply the necessary wattage

There already are two large capacitors in every mainboard to do exactly that, to provide power immediately when required, because the batteries can’t provide the current fast enough for a solid pedal feel. It’s a part of normal operation though, and hence not a basis for a warning.

I don’t think that an overlean is too fast for the battery current to rise, unless you ride into a deep pothole.

An overlean near the max speed wouldn’t benefit very much from the capacitors either, since the issue is not the lack of current but the lack of sufficient voltage differential. The battery voltage must exceed the motor’s back-EMF voltage for any acceleration to happen. The no-load speed is where the voltages meet, hence acceleration can no longer happen. If one rides close to the no-load speed, only very little acceleration can happen.

 

 Solving an issue of a user error is very difficult to do with technology alone, since if you increase the power by any means, it just pushes the threshold a bit higher. But it’s still there. Looking at how many times more power, speed and stability we now have on every (non-toy) EUC compared to just four years ago, the issue is definitely somewhere else than in the wheels’ power  delivery.

Share this post


Link to post
Share on other sites

The difference between euc specific bms and generic bms is that euc specific don't cut out. They report errors (ideally) and/or slow down the wheel, to a stop if necessary.

That's what should happen. That is rarely the case though. Mostly they're just working discretely without any feedback, until they malfunction and your only clue is odd pack behavior, often when cell damage has already occurred.

A lot can be improved on the bms side. If they were built like tanks we'd still today have 10-15 year old packs working normally at a modest reduced capacity of 70-80%. Probably not good for business. :D

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...