AI BMS chip

[slomobile]

https://www.electronicdesign.com/techno ... ts-ai-chip

[Burgerman]

Just what we need. More electronics and programming experts adding logic to batteries that they dont understand.

They can also tolerate thousands of charge-discharge cycles before breaking down.

In thecase of lifepo4 thats many 10s of 1000s. IF charged correctly and a typical BMS is not used...

While these power-handling properties give them the edge in electric vehicles (EVs) and other markets , Li-ion batteries can be unforgiving—and even dangerous. They must be closely monitored by a battery-management system (BMS) to safely and reliably wring out as much power as possible.

Wrong. I am currently building one WITH NO BMS that will outlive the chair and me with 10s of thousands of cycles and more range han I can use because of a better charge system and a correctly sized pack.

To measure the ins and outs of these batteries even more accurately in real-time, Eatron Technologies has rolled out a BMS-on-chip (SoC) co-developed with U.S.-based AI startup Syntiant for light mobility, industrial, and consumer devices.

god help us...

Eatron embedded its intelligent BMS software into Syntiant’s neural decision processor (NDP) so that it can continuously deliver accurate and robust cell-level assessments of remaining charge and other facets of the battery. Eatron said the NDP120 can run its latest AI models directly on the device, wringing out up to 10% more useful power without changing the chemistry of the battery cells or the construction of the larger pack.

10% MORE than a typical life shortened BMS? Thats very easy.
Not using one, at all. And chargeing correctly at a lower voltage and not discharging too deeply without the typical BMS charge issues already at least doubles the usable lifespan. Probably more. And eeking out 10% more "useful power" is exactly whay you DO NOT WANT TO DO!
What you DO want to do is fit a BIG pack and then use it lightly. And get at least 25 years of service from it. And you can. I just paid LESS for a 230Ahset of grade A certified cells, than the lead bricks they will replace cost. Thats 500% range gain, for 2/3rds the cost. Why fit a small pack and then try and eek out an extra 10% and kill it off fast? Because programmers, BMS makers, etc do not get batteries!

The AI SoC can also run predictive diagnostics, including lithium plating detection, on the edge. By gaining more insights into the battery’s internal state, it’s able to identify potential issues before they cause failures or permanently impair its performance.

The problems that a BMS and eeking out 10% more performance and BMS use generally actually caused presmably! If a pack is big enough no BMS in needed or wanted. It wastes space better used for capacity and is a fire hazard. Its the cause of the very thing it claims it improves!

According to Eatron, the BMS software can deliver prognostics such as overall health, aging trajectory, and remaining useful life prediction to increase the battery’s useful life by as much as 25%.

We allready know that using a BIG pack that results in low current draw per Ah, and low average depth of discharge (and made possible by not trying to make room for a BMS in a battery case) can increase service life by 3 to 10 times minimum. ALREADY. So if my new pack doesent give me at least 10,000 cycles, or 25 years use with minimal degradation I will be very surprised. It will be removed for better technology long before that ever happens.

So colour me unconvinced. Electronics engineers and programmers keep of doing things that dont make sense. Look at all the rediculous programmng options and settings on a powerchair for e.g. They steered better before they added all this nonsense in the 80s. Now you have to disable it all before you can make a chair go where its told.

So not exactly convinced its useful. And AI? Or just some fuzzy logic programming. My VW van had that. It "learned" how you drove. So if you drive like my gran, it changed gear early, didnt kick down when you accelerated etc. So when a gap in the traffic suddenly appeared and you wanted some real GO and wanted it now, it decided you wanted eco mode... And it accelerated like a slug. It was useless. 3 seconds after that overtake possibility has gone THEN it decided you wanted sport mode. And stayed that way for about 2 minutes. Instead of AI or FUZZY LOGIC it needed a switch. So I could decide! And thats what it got. I had to pay a man to remove this fuzzy logic code in the gearbox by fitting a kit that disabled it via a switch. THEN it worked properly. Programmers... All have lots of marvelous ideas sat staring at code on screen. On how to make stuff worse...

[Burgerman]

Years back a CC/CV/Float chip appeared on the market. Many chargers all now use this universal battery charge chip. The result is that all our mobility chargers are now useless clones that claim to be "universal"... They all overvolt for gel, stop too soon, so undercharge and then go to a too high float voltage pretty much. Because of the ubiquitous pre programmed cheap Lead battery universal IUU charge chip.

[ex-Gooserider]

Not arguing with why you don't like it if talking about LiFePO4 cells like we use on chairs, but those basically can't go up in flames.... However Li-ION often refers to lithium-Cobalt or other more flammable tech that can / will go boom if abused... But this is preferred by car makers and so on because of better power / weight, so they need to have some sort of BMS protection, and if that is what this thing is for, it MIGHT be better than the current BMS tech...

So not necessarily as bad as you are making it out to be...

ex-Gooserider

[Burgerman]

EV cars already use way better systems than any addon BMS.

The entire system is super complex, charges in the same way the PL8 does, with preprogrammed parameters that vary with all sorts of inputs inc usage and temperatures, cell resistance etc. As for all the protection stuff a BMS does then they control all that way better than a typical BMS does too. Including doing the max current, cell temp needed for max acceleration and efficiency, with water heat and cool as required, and all of this is already a part of the cars computerised power module. EV systems are very advanced and designed to maximise cell life and car performance balance. They never just chop off power when sme arbitrary value is reached.

A BMS of the sort that the chip refers to, as added to all youtubers packs, and drop in lead replacements, etc has so many issues that I am not sure how this chip could help. It doesent know ebough about the specific battery, or usage anyway. Its only there because the typical BMS is a last resort so you dont end up sending them all back under warranty. They are the wrong thing, in the wrong place, trying to protect a cell from extremes to stop numpties killing them off under warranty.
They do 2 things.
1. Charge balancing. There are much better ways to charge.
2. All the protection systems that chop off power are NOT REQUIRED if the pack is big enough for the job and if the controller is current limited, or even fully integrated like an EV.

Lithium is a part of a system. Where each part of it works correctly together. Its not a battery in need of protecton from some youtuber fitting a too small pack, into an unsuitable location with the wrong parameters and a dumb charger. Or it shouldnt be... When it is, a BMS is a last resort with extreme limits that "SAVES" it from disaster. It doesent fix all the ssues that its trying to save it from. It just chops off power when things go too far. If its programmable, then in order to use the vehicle/ebike/whatever, then you will (must) adjust all these cut off parameters so that it doesent interfere with your usage. At which point they may as well not exist because they now do nothing while you damage the battery...
If you want decent cycle life then you are better using a properly programmed cell balancing charger, like EVs do which are way more complex than any BMS, OR even like we do here.

At which point what is a universal "BMS" for? Youtube. The ignorant. The manufacturers of cells so that 90% dont come back under "warranty"... And people that think programmers and phone settings and monitors mean its all great! So youtube numpties and manufacturers not having to replace cells under warranty...

The charging thing is much easier on the non LiFe cells. So a dumb charger and a BMS can be used on those. But you still dont get adequate control of charging to do it properly like tesla for eg. And the same with cutoff. So in a laptop? Yes. In a bigger pack, like we use, or a tesla? Not likely! The cost/benefit of replacing a battery doesent add up.

[slomobile]

all of this is already a part of the cars computerised power module.
They never just chop off power when some arbitrary value is reached.
are better using a properly programmed cell balancing charger, like EVs do which are way more complex than any BMS, OR even like we do here.

cars computerised power module = BMS
All EVs chop off power at arbitrary values while charging, or discharging, or sitting idle. If you are driving at the time, they force the vehicle to slow to a stop first, but that is at a voltage somewhere above the very bottom. If it hits bottom or top and cant shut it off nicely, it blows the explosive fuse regardless.
Here is a Tesla onboard charger. I don't see any connectors for cell balancing. Just one output 50-430VDC @ 30A
https://www.ebay.com/itm/115853659444 Cheaper than a ZTE ZXD2400

[Burgerman]

Tesla chargers are controlled and throttled and charge in whatever way their complex algo and balance system tell them to do. Regardess of what you can see.
The system that does this is very complex too. That bit you are looking at is useless without the computer that controls it. Its not a BMS in the way that all the dumb ones that we are talking about for that chip. Tesla and panasonic who make the cells and the cars know a thing or two about their batteries.

Thats why it says 50 to 450A amps. The balancing is controlled by another module all linked by a bus cable that controls everything. The balancing can probably synch or move 50A.

Yes if you run a tesla low enough and ignore all the warnings and keep on going then at some point it will have to pull the plug. After slowing etc. And after limiting current, lowering acceleration levels, and a lot of other things that go way beyong what that chip does.

[Burgerman]

https://circuitdigest.com/article/tesla ... erspective

Have a read. The vehicles BMS is way more complex than the things that youtubers fit. It is connected to the vehicles computer via a bus cable that tells it exactly what to do and when to do it. And its also connected to the vehicles "charger" voa the same bus, which basically it can throttle and control. So that the BMS and the power supply together make up the charger in the same way that the PL8 is both BMS and charger and the cell balancing in that too can control the charger output.

But the difference here is that theres a computer with some complex stuff thats way beyond any of us that makes all the decisions. No ON/OFF BMS nonsense.

[Burgerman]

Incidentally. Heres a full discharge test of a 4S 2.2Ah 18650 Lithium Polymer cylindrical cells used in a small quadcopter.

This is starteing fully charged. Ending when the cells drop off a cliff at 3.50V.

Note that unike LiFePO4 the discharge curve FALLS in voltage with state of charge in a nice line. So that with these cells, Voltage = state of charge. So these can be balanced regardless of state of charge, and actual battery meters are very accurate. And the same with a BMS charging them. They are so easy to balance and control that here a traditional BMS works OK. As in most lithium ion battery packs. These LiPo and Lithium ion, both exhibit the same slope. But lithium ion go lower, down to 2.7V approx.

They dont have the flat discharge curve of LiFePO4. Unlike LiFePO4 the voltage falls in proportion to charge level in use. So teslas job is quite easy! They can balance at any voltage and so an 80 or 90% "full" charge is possible. And they do this. Although it varies with pack age, temperature, etc. To get a seriously long service like from a derated capacity. They could offer much more range already. But the battery wouldnt last long.

[slomobile]

I read. Only thing I didn't already know is the model number of the chip. https://www.ti.com/product/BQ76PL536A-Q1 "Device type Cell monitor and balancer" That is what I think of when I hear BMS. What does it mean to you? This is clearly a Tesla BMS on the pack. It has a separate on board charger with 1 output only that does not balance. BMS does the monitoring and balancing. Right? Am I missing something?

The BMS used by Tesla in Model-S is based around Texas Instrument’s bq76PL536A-Q1 3-to-6 Series -Cell Lithium-Ion Battery Monitor and Secondary Protection. The BMS is integrated into every module and monitors the battery life, temperature, and charge-discharge cycle of cells. It is a stackable Battery Monitoring System and uses a High-Speed Serial Peripheral Interface (SPI) for Data Communications. The image below shows a simplified system connection of the BMS.

[slomobile]

https://www.jkbms.com/
These have been recommended by Shirley, and I see them used in a lot of youtube videos. Are these the BMS you are saying we should not use? It seems very similar to the Tesla BMS you seem to admire. The complex computer control system is just an EVCC https://www.thunderstruck-ev.com/evcc-c ... oller.html

The important element seems to be communication between the BMS and the charger, so that the charger can adjust its output to the whole pack voltage which the BMS prefers.

That is available in ready made bricks now. RS485, or CAN, or Bluetooth

[Burgerman]

It does but it also throttles the charger (the power supply) or rather the computer that controls everythng does. As does the PL8. I think theres a very similar looking 10mm package inside the PL8...
Also all that chopping off power that a youtube BMS does doesent happen on a tesla. You dont need to adjust all the parameters until it stops tripping you up. Because here the BMS is mostly sed as a bunch of sensors that report to the computer. That tells the power module what to do. The BMS itself doent keep chopping off power. Same with the "charger"...

So you add high power DC or AC to the car. And the car does what the PL8 does. With added logic from a complex computer. And from tesla live over the air if they want. So for e.g. the battery was allowed to charge full, and discharge fully or at least more so for a few days when there was a tornado in the US. Its a more integrated and comtrolled system, in every way, that a dumb BMS that cant control the charger. Or the power module. Everything taks to each other and the computer too.

[Burgerman]

https://www.jkbms.com/
These have been recommended by Shirley, and I see them used in a lot of youtube videos. Are these the BMS you are saying we should not use? It seems very similar to the Tesla BMS you seem to admire.

No. The difference that you are not getting is that a dumb BMS like that cant control what the power module (your joystick) can do. And it cant control the charger output.

This is what you get when a dumb BMS cant control the charger... This is shirleys BMS.

Here it is just CHOPPING OFF power every time the BMS sees a cell go OVER VOLTAGE and then it waits until the voltage gets pulled down again, then it reconnects... Then the thing overshoots over and over... Because it cannot throttle the charger to below or same as balance current... It keeps on reconnecting the charger at full power. Which instantly causes a huge jup above the safe voltage. Rinse and repeat. For how long? We never know!

Its the wrong thing, in the wrong place doing something you dont want! Teslas do not do this! They do the same thing as the PL8. Below.

[Burgerman]

Heres what you get when the balancer IN THE PL8 can throttle the charger... The current is reduced so that no cell goes above the set voltage.

Then when balance current falls to zero or close to it the charger knows that it can stop.
It NEVER knows when to stop with a BMS!!! Because it doesent even know if the cells are balanced yet. This can take minutes, to days. How is it supposed to guess?

Then, you say BALANCE AND CHARGE means BMS. NO!!! Thats a charger. And as you can dsee a bad one. UNLESS it can control the charger or power supply output proportionally.

A BMS has 101 safeety nazi cutouts. All designed to stop a numpty from wrecking the cells through ignorance. Only required if the pack is too small. Or you run it with too high current etc. These things stop you dead because they cannot talk to the controller that drives the chair. So instead of limiting current, or whatever, it stops you dead as you try to get off a road up a slope.
And many other things.
If the pack and controller are selected correctly, then you only need one protection. A battery MONITOR. $3



So no A BMS that cannot talk to the power supply (charger) and control it, and cannot talk to the controoler for the vehicle is useless!
The reason EVs have the charger in the car (and not on the wall or the street) is exactly the same one that means the balancer and the charger are together in the PL8. Power output control.

And all the normal unwanted BMS safety cutouts etc are controlled properly by the computer not a BMS in the tesla.

[slomobile]

Several of those JK BMS and others do communicate via Bluetooth, CAN, or RS485. Outputting cell conditions, warnings, etc. Several chargers also have Bluetooth, CAN, or RS485, for turning on/off or adjusting output voltage and current limits. It isn't rocket surgery to read the BMS data about the cells, determine what the charger should do based on the logic you've given here in the forum, then send it the command to do so. Does the PL8 do that? Communicating to the ZXD2400? Could it?

The cutting on and off, I thought you were talking about the power supply before, but from your graph it looks like individual cells. Are we looking at the BMS balancing the pack there? It isn't being switched on and off, because it doesnt drop to zero. Looks like a reverse sawtooth. Quick rise, more gradual drop, no plateaus. Is that shorting a resistor across a cell periodically?

Frankly I don't understand how the balancing is done. I'd like to learn. It must have to involve some switching control, otherwise you would produce a ton of heat operating transistors or FETs in linear mode rather than switched.

[slomobile]

https://www.jkbms.com/products/active-balancer-bms/ says

The above active balancing BMS with active balancing current cover from 0.6A to 5A series, and support max continuous discharge up to 350A. Widely used for LTO battery packs 20S to 24S, LiFePO4 battery packs 16s to 24s, Lithium-ion battery packs 14S to 24S. Feel free to contact us for assemble manul or custom design BMS or technical support is needed by email info@jkbms.com

So how bout we see if we can design a BM BMS without all the safety cutouts, just the ones we want. Rather than say "that computer stuff is too hard" we look at your advice on how to properly treat the cells, translate it into programming for a BMS, made as you like. As you did to get the Roboteq response right with Lenny's script. We aren't in a hurry. I'm basically working that problem already from the EV angle.

I think I know how to do it with https://www.thunderstruck-ev.com/bms/ TSM2500 power supply. But we want to use the ZXD2400 right?
Shirley or anyone modding those, is there an API or some kind of guide for controlling modified ZXD2400 via CAN, RS485, or bluetooth? Either of those can be trivially converted to UART or SPI or I2C if that is easier. I've tried to read that thread, but it is just too huge.

Edit: What is the time scale on that spikey graph. It looks like a variation of 0.0504 over 10 minutes if the scale is hours like the other graph. Then it smooths out once balance is reached. That doesn't look terrible to me.

[Burgerman]

Several of those JK BMS and others do communicate via Bluetooth, CAN, or RS485. Outputting cell conditions, warnings, etc. Several chargers also have Bluetooth, CAN, or RS485, for turning on/off or adjusting output voltage and current limits. It isn't rocket surgery to read the BMS data about the cells, determine what the charger should do based on the logic you've given here in the forum, then send it the command to do so. Does the PL8 do that? Communicating to the ZXD2400? Could it?

Non of the BMS on youtube do any of the things I mentioned above. They just talk to your phone...
They all do the same nonsense.
The PL8 is self contained. Does the tesla have to communicate to the wall supply it is connected to? No. Its just the power source. Same reason the PL8 and all hobby chargers do not need to communicate with the ZXD, the car battery they are runnig from or whatever. They DO communicate internally, the balance system controls the charger output, tells it when its balanced, tells it when the correct termination point is reached etc. All in one seld contained unit. Thats what a BMS whn used for charging cannot do.

The cutting on and off, I thought you were talking about the power supply before, but from your graph it looks like individual cells. Are we looking at the BMS balancing the pack there? It isn't being switched on and off, because it doesnt drop to zero. Looks like a reverse sawtooth. Quick rise, more gradual drop, no plateaus. Is that shorting a resistor across a cell periodically?

The BMS CHOPS OFF the power supply. It then doesent drop to zero because you are looking at a BATTERY! Each cell has its voltage. The CHARGER or POWER SUPPLY pushes this up. So all the BMS can do is pull the highest cells down. Then reconnect the charger or power supply, and then up it goes again. It cannot throttle the output. It cannot talk to the power supply/charger. Its in the wrong place.

Balancing is simple.
When a cell sees itself reach the choosen voltage, the BMS tries to pull that down By applying a load. Typically a load of almost bugger all. Meanwhile the charger keeps trying to achieve the 29.something volts. At high Amps... While it does this IN SERIES across all cells, the highest cell just ignores the feeble balance current and then reaches the OH MY GOD point at which time the BMS CHOPS OFF THE CHARGER TO TRY AND PREVENT MORE DAMAGE. Then it pulls that cell down again over x amount of time. When it has, it reconnects the high amp charger and off it goes again into orbit. So it cuts off power all over again.

What it SHOULD do is throttle the charge current to whatever is required to smoothly hold the highest cell at the voltage that we want. If the balancer is 1A then the charge current may need to be dropped from say 40A to 1A. So you get that cell voltage bouncing up and down for hours to days.

Thas why BMS are crap for charging.
Its why an intergrated system like tesla and hobby style chargers work properly and accurately.

[Burgerman]

So how bout we see if we can design a BM BMS without all the safety cutouts, just the ones we want.

We dont want ANY!!!

Rather than say "that computer stuff is too hard" we look at your advice on how to properly treat the cells, translate it into programming for a BMS, made as you like.

You are not getting it.

The only way to do that is to let the BMS for CHARGING be integrated with the charger. So that it can control the power output, see the cells are balanced, watch the termination current etc and then terminate when all is well. As per the PL8. Or tesla. Rather than being in some BMS on the battery that is connected by 2 wires. You could fit the PL8 on board and that would do the job fine. But it takes a lot of space and adds complicaton so why? Once you have more range than you can use the charger (and balancer) is best at home.

And all the safety stuff it does we DO NOT WANT.

Non of it is helpful in a correctly sized pack in a powerchair with a current controlling system. When would you want it to stop you? I cant think of ANY reason why I would want a BMS to pull the plug ever. Unless to save the battery from extreme over discharge. But a $3 monitor would prevent this if the battery was too small and give you time to get of the railroad tracks... Or road.

So as I keep on saying. A BMS is the wrong thing, doing a lot of stuff we dont want and it cant control the charger so it cant even do that part well.
So whats it for?

I might use one on a solar system. But not in a powerchair.

[slomobile]

Damn you're thick sometimes. Faith in wrong ideas so strong its almost religious.

The only way to do that is to let the BMS for CHARGING be integrated with the charger.

That is exactly what I've been talking about. Doing exactly that. Maybe not integrated in the same box, but connected and cooperating to the point of being the same thing. It is what EVs do. Wheelchairs are EVs. Its about time we do it too.

You have not seen all the BMS builds on YouTube, just the ones simple enough to trigger the algorithm to recommend them. Most people would say there are no prone wheelchairs on YouTube because none have been recommended to them. Even if I search for my name and the exact title of my prone wheelchair video on the very computer I uploaded it from, it does not come up in the list of results. But it exists. I put it there. With the link, I can find it.

You do want some minimal safety. At some very high voltage, the charger is obviously malfunctioning and leaving it connected is just damaging the battery for no reason. You may want to disconnect the charger. It is possible to just disconnect charger, and can be automatically reconnected after a timeout. Or reconnected when Rnet is power cycled.

At some low voltage the wheels can no longer be turning. After say 12 hours you are certain the chair was put away left turned on with the joystick off center. You don't even need to jump right to switching off the load. Once you recognize the problem, you can trigger an Rnet error that wont stop the chair but tell you there is a problem. If condition persists, override rnet joystick commands as Luci does, to slow the chair to a stop.

What makes you think that a discontinued hobby charger is the only thing that will ever be capable of properly charging a wheelchair battery?
https://www.orionbms.com/

[slomobile]

https://www.orionbms.com/category/charger-integration/
https://youtu.be/XUrpciP-6t4
https://www.fellten.com/

[Burgerman]

That is exactly what I've been talking about. Doing exactly that. Maybe not integrated in the same box, but connected and cooperating to the point of being the same thing. It is what EVs do. Wheelchairs are EVs. Its about time we do it too.

I looked at your links. They dont do the same thing. But they attempt to do it by trying to make a BMS talk to a charger and you can make that work if you get everything right but why?
And NO EVs do not do that!
They have everythig integrated on board in the same way as a hobby cell balancing charger does but purpose designed for their specific system. The only part thats not in the vehicle is the AC or DC power.
I could probably separate al the bits of a hobby type charger and have some bus cable to connect the two bits. But they are already working and do a great job. So why do that?

You do want some minimal safety. At some very high voltage, the charger is obviously malfunctioning and leaving it connected is just damaging the battery for no reason. You may want to disconnect the charger. It is possible to just disconnect charger, and can be automatically reconnected after a timeout.

Tats EXACTLY what a BMS does. It does it for an awful lot of reasons and cuts off power. And I never want that to happen. I cant walk, cant get home, may be stuck where no people, or in a dangerous spot. Once you adjust all the limits so that it never does any of this stuff, then it does nothing and so wht is it for? Added complexity, added fire risk (failed mosfets and memted balance wires etc have been shown to be the cause of most LiFePO4 fires. Not the cells. So no I wouldnt want this thing on any chair I am sat in or pared next to my bed thank you!

Or reconnected when Rnet is power cycled.

But they dont. And by then I may be under a truck.

At some low voltage the wheels can no longer be turning. After say 12 hours you are certain the chair was put away left turned on with the joystick off center. You don't even need to jump right to switching off the load. Once you recognize the problem, you can trigger an Rnet error that wont stop the chair but tell you there is a problem. If condition persists, override rnet joystick commands as Luci does, to slow the chair to a stop.

Dont understand that.

What makes you think that a discontinued hobby charger is the only thing that will ever be capable of properly charging a wheelchair battery?
https://www.orionbms.com/

I dont. Theres many things that can PROPERLY charge a lithium pack. But 99% of what a traditional BMS does I absolutely do not want or require. Including the fire hazard. As for charging, they are crap.
Your orion BMS just tries to do what a hobby charger, does. Badly. With a load of extra complication caused by trying to split it into two places.
Its bulky, expensive, and meant for full sized cARS. yOU CAN DO EXACTLY THE SAME THING, With a hobby charger and one connector without trying to add a mountain of complication onto a chair.

But feel free to do it your way! I just dont understand what advantage there is. But I see many disadvantages.

[slomobile]

There is this too. https://openinverter.org/wiki/Battery_Charging

[Burgerman]

Do you understand how big, bulky and heavy a full sized EV system like the ones you are linkng to are?
Completely impractical, and 4x the cost of the entire battery! And it doesent do anything any better than a hobby charger for our purposes. It is fact tries to do the same.

But theres no place to fit all that unless you completely remove the battery. Wait...

Seriously what meds are you on?

To charge a wheelchair we want exactly what a hobby charger does. I use the PL8 because I like the capability and PC control, logging, and user setting and adjustability for all the other batteries and hobby stuff I charger. Some use iChargers. But this is the simple way. Leaving no complication, no modules, no anything on the chair. SPACE for battery power. Thats how we get 230Ah in place of a GRP24. Or on a permobil a 160Ah pack. Where are you going to put all that unessasary nonsense on a char?
And what possible advantage do you think it gives you?
I can name a lot of disadvantages!

There is this too. https://openinverter.org/wiki/Battery_Charging

You realise that the PL8 and all hobby charger are also buck/boost chargers right? And that they CORRECTLY balance cells, and throttle charger as required, and allow you to set about 50 or so charge parameters to your choice?
ALL IN A SMPLE BOX not needed to be wasting space on a chair. ONE connector.

[slomobile]

This aint big.
90.00€
https://openinverter.org/shop/index.php ... duct_id=74
This BMS has up to 16 cell voltage inputs, one current sensor input and two temperature sensor inputs.

It can be cascaded for larger battery packs.

Continuous transmission of aggregated cell data (min/max/avg voltage and temperature)
Continuous transmission of pack data (charge/discharge limit, SoC, current, pack voltage)
Single cell voltage obtainable via CANopen SDO
Automatic module addressing scheme, no jumpers needed
Configuration via command line tool: https://openinverter.org/forum/viewtopic.php?t=2907
SoC estimation by configurable SoC lookup table
Continuous SoC calculation via current sensor integration
Bidirectional balancing, low cells can be charged individually, high cells discharged

[Burgerman]

It and all its connections take up space, cause fires. And I still cant see why I would want it? Read pages of that link, no idea what that all means. Sounds like a physics project to charge a battery. Why?

And has all the same cell charging bounce up an down stuff all the rest do. As far as I can see. Where is its specs?

And again, whats it do other than waste space I would be better off using for cell volume again?
One reason that the drop in lead bricks have half the capacity of seperate cells is that a BMS that does a load of things we nether need or want takes up a lot of the battery space.
I dont see what its for? As in, apart from being crap for charging, what else does it do?

Lets say for the sake of argument, it does the same as the PL8 I use now. And I can monitor, test, configure without a damned command line, what does it do thats better than I am doing already? As in where is the advantage?

Dont let me stop you. If you think you know enough to do it and dont mind all the complication and extra wires and wasted space,then try it.
I just dont get the point of it?

[shirley_hkg]

The cheapest ORION BMS @$875.00
seems doing not much better than what a $5 aliexpress BMS could do.

We are users, and now a set capable of 5X range, outlast our lives , at just 2/3 of MK Gel … , why bother something we don't need.

BM isn't smart at all, IMO. He lingers too long till his EVE-240 cells emerged, while many members here have been enjoying such a decade back.

Enjoy

[Burgerman]

I have been trying to wear out the headway cells! And given up!

And waiting for GOOD cells, Real grade A rather than aliexpress Grade a at a rediculous price. To be fair I have only been doing half days due to sores for years. So it didnt matter. But I think I got that beat!

Seriously, that super expensive thing can only do 200mA balance? Thats five times worse than the PL8.
If a 230Ah cell is just 1% out of balance, then thats a day!

Worse, I think its an active balancer. And so its only 200mA when its miles out of balance. If its close it might be far less!
I dont get why he wants to add a mass of complication, to charge worse, and waste space where we have non. Even ignoring the silly price. I think he is still on the pills.

When fitting a 230Ah pack, there isnt even 5mm (1/4 inch) to spare. Where is it going to go? And what does it achieve that any number of hobby type cell balancing chargers dont already do better? Baffles me. But its up to him. If he wants to do that its his choice!

[shirley_hkg]

Seriously, that super expensive thing can only do 200mA balance? !

Here. Hope I am wrong.

[Burgerman]

It appears not. Thats rediculous. Its probably because it is intended for Lithium ion and so it balances all the time regardless of the voltage. Another active balancer that drains one to fill another all the time 24/7. That wont work on LiFe cells. So it gets away with it.

[slomobile]

The balancing can be switched off. It is bidirectional. Can take away 200mA from the highest cell and charge the lowest at 50mA simultaneously. It still isn't a lot, but its faster. The Orion is overpriced and I know it. It isn't oversized beyond what is required to meet automotive grade cooling requirements. The Orion BMS Jr. 2 is for 16 cells and all that is needed. $675 including 200A dual coil current sensor. It measures and internally logs internal resistance of all cells as they go through various temperature and load points while in use. Eventually building up an exponential graph of impedance vs temperature, which it uses to modify its charge curve from the initial estimate formed from user entered impedance data. They offer a service where you send them a sample cell and they characterize its impedance curve every 10C and provide you with the graphs and a custom charge profile based on it.

I'm not saying its the best, or cheapest, or one I will use. Just that I knew it was possible for BMS to control the output of the power supply and that Orion was definitely one example that could do that. I assumed there were several other brands that also did that, but I haven't found them yet. Perhaps that is why they charge so much.

Regardless, with any BMS that simply acts as a monitor and transmits cell data, we could use a microcontroller to read that cell data, and with our own decision making logic programmed in, we can send CAN outputs to raise or lower the voltage of the charger. If the charger is capable of responding in that manner. That is the reason I'm looking into automotive onboard chargers which I know can do it. I'd be happy to try the ZXD2400 instead if anyone can tell me if it is capable. I've talked about that before but didnt get an answer.

Not all BMS are even capable of interrupting the load. That seems to be the major issue BM always comes back to, wrongly assuming they all can shut you down and leave you stranded. Therefore they are all bad, no matter what. BM instantly saying that ANY BMS, including proven automotive grade are junk just because they bear the label BMS is nonsense. He's lost credibility with me on this topic by taking the Trump political tactic of criticize everything instantly for any reason and when that fails criticize the person.

BM keeps asking why I would want anything other than the PL8? Because I don't have and probably can never get a PL8. If I could, and it ever has problems, there aren't any replacements available. Plus it is a complicated interface from what I understand, having never used it myself. I want something smart enough that I can just plug it in and don't have to think about it. It does the thinking for me once I program it once. If I were to get any other hobby charger, you all know it wouldn't matter which one, BM would say ALL THE TIME it isn't as good as the PL8. So I don't give a shit about any hobby charger. BM has poisoned that well.

One feature I'd like to implement is a long term storage setting. How many of us have bought a used chair, they say "It worked when I put it away". But batteries are trashed. With LiFePO4 ability to outlast us, it would be nice to pass on a bit of that extra capacity. Not fun to think about, but if we go into hospital for a while, or travel without our chair, it would be nice to still have good batteries when we get back without hiring someone to come around once a week to plug them in (Ok, thats a lead thing, not lithium, I know). I read that Lithium battery impedance is highest when fully charged or discharged and lowest around 50%. If that is true, there is a way to stay charged enough and discharged enough with a long term battery maintainer program built in to the BMS. It isn't a glamorous feature. Most of you won't care. But I have bought a LOT of batteries. And nearly all of them have gone bad without much use because I needed them in remote locations to do things I couldnt physically. Water pumps, hoists, winches, fans... I didn't use them enough to keep them maintained, and didnt want to lug them back to the barn all the time. So they sit out, dying slowly. With a small solar panel and maintenance mode BMS I think I could leave them unattended indefinitely.

It might not work. I'm speculating and exploring options here. Sharing what I learn as I go. I appreciate the relevant feedback.

I'm exploring things that may not fit well in a typical wheelchair because I cannot use a typical wheelchair. Those size constraints dont apply to me. I'm lying face down. I take up a larger wheelbase. I have extra room for stuff. I already have my basic mobility needs solved, I'm looking to build higher speed, rougher terrain, and longer range into a new thing. More a small EV than a wheelchair. I'm mostly doing it for the fun of thinking up the thing, not because I need it. So yeah, sometimes what I come up with will be impractical for you and not make sense to you. We are in different circumstances. That said, the EV stuff isn't as big as you think. The ones in the video are large because they are sized for a full sized car, and most of the size is in the Tesla battery modules. The ones I would use (Jr.) are sized for 48v golf carts. Much smaller. Even smaller than the stock permobil lead charger. 3.3kw or 6.6kw if I can find a battery.BMS combo that accepts charge at 120A.