"Simple" solution for Lithium charging

[rune]

JK bms settings

Thanks Shirley

I've programmed it. Now just to hook up a PSU... I currently have one that's 16A at 48v that should do ok for now.

[emilevirus]

I am using the JK-PB2A16S-20P

It's 200A, has the screen and extra interface bits and power on/off button. The Salsa M2 battery box I am fitting it in has space for it beside the battery without issue. The battery fits with it's longest sides orientated front to rear. I think these inverter BMS's are their latest spec models. Docan were able to supply this with the cells, so I didn't need to pay any extra shipping from China.

Yes, the 16s20p configuration is overkill, but it's what Docan had and going for the this means I get 2A balance and 200A rating. It also has two programmable relay outputs that can be triggered on various conditions, that I might use. I am thinking of connecting one to an inhibit line so that it's not possible to drive the chair when it has the charger plugged in, for example.

You could technically use can2rnet to inject the real SOC from BMS.

[Burgerman]

I ordered the one with the extras, cost me £37 delivered. Because after reading many things, thats the easiest way. Even if theres room alongside the battery on a salsa, I am not sure I like the idea of one on board. Mosfets tend to fail shorted which can end up with fires. The EV forums are litered with lithium fires and with LiFePO4 packs too. Almost always traced to the BMS deciding to throw a wobbler...

But that can connect directly to the windows software via USB which means I can do that via BT as I did with the PL8s anyway. It isnt possible to do this with he built in BT by all accounts.

So while £37 delivered is more expensive its still pretty cheap. Does 8S and 2A balance, although it isnt what it seems. Thats only 2A if its way out of balance. The PL8 is only 1A but it balances at 1A even if its just 2mV spread so its way faster and more accurate.
And theres another issue. Its only accurate to 5mV per cell. Compared to 1 on the PL8. Does that matter? Not as long as you only balance above 3.450 to 3.500V. If you try and balance below that level the 5mV inaccuracy will severely unbalance the pack in time.
So if I dont like what it does with charging after careful configuration then I can dump it and go back to using the PL8 regardless since theres 4 of them here!

So its an experiment to see how it really works!

[emilevirus]

I tried looking online and never seen a JK BMS or JBD BMS causing a fire.. Most fires are caused by human errors or cheap crappy BMS..

[shirley_hkg]

Done. Simple for carer to do charging.

[shirley_hkg]

Does 8S and 2A balance, although it isnt what it seems. Thats only 2A if its way out of balance. The PL8 is only 1A but it balances at 1A even if its just 2mV spread so its way faster and more accurate.
!

It does active balance by dumping energy to a capacitor, which then charges the low cells. You get ½ the rate actually. It also does that ONE cell at a time. PL8 does it 7 cells together, so it is far more efficient.

However, JK bms drains the high cell, that helps to avoid cooking the high cells. Just allow enough time to it.

[Raro]

Rune, what type of cells did you buy from Docam? Can you post a picture of the cells installed with the BMS?

[emilevirus]

On my MAGIC Extreme X8 I just put the BMS where R-Net PM is. No isssue fitting it!

[rune]

Rune, what type of cells did you buy from Docam? Can you post a picture of the cells installed with the BMS?

The cells and price were:

$46×8pcs EVE 230Ah lifepo4 cell with welded Single M6 studs=US$368

I'll look at sending pictures tomorrow if I can. A bit late to be taking pictures and stuff right now, off to bed soon :-)

To install, I have made a pack using the wiring diagram which is shown above. It has the insulating sheets that came with the cells in between and is wrapped in clear tape. Over the tape I have thin sheet plastic and then another tape layer.

I lined the battery box with adhesive-backed thin foam, of the type that is often used for vibration/sound deadening. and then placed the battery on that. The BMS will sit alongside and I think I will just fix to a vertical plastic sheet beside the battery with heavy-duty velcro.

I've also made a wiring harness with anderson connector for battery, load and external power(charge).
I'm going to mount it like this so it's secure but can also be easily be removed and swapped out, disconnected, etc.

I am also going to mount the display that came with the BMS in a small project box mounted somewhere at the rear, so it can be plugged in and used but also removed easily if required,

I'll mount the power on/off switch for the BMS on the rear, near the P&G controller.

I think I am also going to use one of the DRY contacts on the BMS for an inhibit, so it can't be driven off when the external PSU is plugged in.

[rune]

Those of you using the JKBMS, what are you using for a power supply? The data sheet and instructions suggest you can use a dumb power supply of anything from 20-70v However, is anyone using one at 48v or even higher for a 24v pack?

What I am unclear of is should the PSU be a few volts over the pack max voltage and no higher or is it, a few volts over the pack and anything up to the max which is 70v on my BMS. >Clearly higher voltage has an advantage here.

So, is anyone charging their 24v pack with, say, a 48v or 60v PSU with the JKBMS?

[Burgerman]

I dont use a BMS but ordered one to evaluate as the PL8 does a better job. But those are no longer available and so many want to use a BMS. I WILL use the following:
Charge at 8x 3.550V ideally. 28.40V. 40A. Or 50...

Set top charge voltage to 3.550V per cell
Set balance to NEVER start unless above 3.45 to 3.50V.
Set the CHARGER to cut off charge when the BMS has had time to fully balance the pack... Unfortunately thats guesswork. So will need the tail current monitoring to evaluate and test after a few charges.
Why? Because even a active balancer is not 100% efficient and its very slow. So if the charger ends before balance does then it will not fully balance the pack. As voltage will fall below that 3.45V point. Hence the buy and test thing.

Use a ZXD set as lithium.
These are cheap. Built like tanks electrically for 24/7 telecom use, adapted to be a power supply by a group, ask shirley... I got 2.

Can be a simple user adjustable supply.
Lead Acid 3 stage charger that YOU can configure.
Lithium charger that YOU can configure correctly...

From 0 to 50A and from 0 to 60V officially... (70V max if you mess about with the bios settings)
Auto off, auto stand by, switch off or end charge after current that you set + x time etc. 3000 watts output capable all day long.

REAR
https://www.wheelchairdriver.com/board/ ... &mode=view

FRONT
https://www.wheelchairdriver.com/board/ ... &mode=view

More
https://www.wheelchairdriver.com/board/ ... start=2010

[emilevirus]

Charge at 8x 3.550V ideally. 28.40V. 40A. Or 50...

50? Shirley told me to run a bit under 50 like 45 max to avoid being constantly on its limit.

[Burgerman]

Well its a 50A supply. It does it fine! Yes 40 is also fine.

I charge lithium at whatever rate I have to in order to replace whatever I need to replace in the time available. Currently using 120Ah one!

So if I have all night, 20A may be adequate. Do simple math in head! You learn to estimate how much will need to go back in. Not been out the house? 25Ah approx to replace. Dont bother charging...

After 4 days of that you have half a pack. Theres no need to charge daily like lead, the battery prefers to be half charged. Hates empty or full. So dont! No need. Also charging fully is the only way to balance. So you do have to do so occasionally at least!

Its all swings and roundabouts. They are so cheap today, and have so much range (capacity), that we cannot do anything other than baby them. Thats a function of huge Ah. Its no longer as important to worry about lifespan so much.

[Burgerman]

Only do big currents with the rear Anderson!

12 GAUGE Silicone charge wires? 40A max, or they get a bit warm. Safe at 50 but get quite hot. So many different things to consider!

The front connectors are good for about 20A on the banana 4mm ones and the yellow XT 90 can do 40A without getting very hot. But those dont like repeated use much and then they melt. I use those on hobby batteries. So use REAR Anderson!!! Those can do 100A plus. and hot swap 50A.

[snaker]

Those of you using the JKBMS, what are you using for a power supply? The data sheet and instructions suggest you can use a dumb power supply of anything from 20-70v However, is anyone using one at 48v or even higher for a 24v pack?

What I am unclear of is should the PSU be a few volts over the pack max voltage and no higher or is it, a few volts over the pack and anything up to the max which is 70v on my BMS. >Clearly higher voltage has an advantage here.

So, is anyone charging their 24v pack with, say, a 48v or 60v PSU with the JKBMS?

Be careful when choosing a PSU. It must be able to throttle the oputput current at a specific level e.g 50A. Or able to adjust the output current. A generic PSU e.g a server PSU does not have this function. I do not know if JKBMS can handle 48V when charging 8s. But in practice, to charge 8s, 29.2v (8 x 3.650V) is already too much.

In VN, I see people call a PSU that can adjust the output current as "a power supply". And as "a PSU" if it cannot throttle the current. Is it correct in English?

[emilevirus]

Only do big currents with the rear Anderson!

What if we have the cheap version with XT90 only?

[shirley_hkg]

When CV is low enough, it won't trip OverVoltProtect and residual current continues.
BM might find a profile to config ZXD to shut down at a certain termination amp.

[Burgerman]

Correct. And that flexibility on the ZXD matters!

You also want the power supply to be able to LIMIT current. The BMS cant. Any old power supply will try and charge at whatever power it is capable of. Ohms law says that it will go bang as it tries to pull a dead battery to the supply voltage. The BMS cant control that. Other than by chopping off the power...
All it will do is chop off the power if it exceeds some preset point. You do not want it doing that!
In fact you dont want it chopping off power from the charger OR the chair at any point. All that protection stuff is not wanted!!!

Such as when charging.
You want the CHARGER or supply to control the charge current, AND voltage. And you do not want it too high! The BMS can only CHOP OFF the charging completely on/off as it exceeds some pre deternmined Voltage point. If thats set to say 3.550v PER CELL, or 3.555, then you should be charging at 8x 3.550V. 28.40V. The actual voltages you need will depend on the calibration of the power supply and the BMS. You need to be charging at slightly less than the OVP point.
With balance to start at 3.45 to (prefered 3.500V) and above.

Now while practically anything and any settings will work, in a fashion they wont work properly and will torture the battery and the BMS mosfets.
You want the charger to be able to continue charging without chopping on off at reduced current smoothly so that the balancer can work fully and then as the CV current eventually tails of to a pre determined value To END the chage at the correct point as soon as possible.

that should be:
a. AFTER cell balance completes (and the charger cannot know this which is 1 reason the PL8 is better)
b. AFTER the residual CV current falls to around 0.75A? (To be determined by testing) so that the CV stage lasts around 15 mins. Provided a. is complete.

[shirley_hkg]

max voltage and no higher or is it, a few volts over the pack and anything up to the max which is 70v on my BMS. >Clearly higher voltage has an advantage here.
?

I got big spark when the voltage difference is high. Connectors were welded at the contact.

Beware of your 24v wheelchair electronics are connected too. Not sure of ripples and noise by doing that .

[rune]

Thanks Shirley. Either the PSU I have is faulty or it cannot handle an initial surge of high current and just shuts down instead of regulating the current as when I plug it in, it just shuts off the output anyway. Not great :-(

I am interested in these PSUs that BM said you could supply. Is this something you could help me with? I'd be interested to know the cost and how to order.

Thanks!

[Burgerman]

What is the supply you have?

It must be output current regulated. Or the iniial surge will always overload it. You are connecting a low impedance battery at x volts and a supply at xx volts... Huge currents will try to flow. Possible 1000s of amps. The BMS will see that and will chop off power but its too slow. Electritcity travel at light speed. The BMS spends an age in comparison figuring out it should turn off. So the power supply either dies or shuts down depending on design.

Re: "Simple" solution for Lithium charging

Its quite amazing how complex the "simple" solution is.
And once working then theres 101 things to configure and set correctly and try to match to even approach the charge capability of the olde PL8 can do.

[rune]

Yes, true BM. However, it depends how we define "simple". I wanted simple at point of use (i.e. for my daughter) I am not so worried if the setup is complex as I can do that myself.

In fact this is almost always the issue, the simpler the solution needs to be at point of use, the more complex the design, engineering and configuration is because, as they say, "when something is made idiot proof, they just make better idiots"!

Regards the JKBMS and how it manages the charging etc, it seems from what I have learnt online, PB series "inverter" BMS's are more advanced than the old ones. In fact I have read that they do provide full charge control whereas the older models do not. There are also some differences in the configuration options. For example, the new JKBMS no longer just monitors the current at the end of the charge, but implements a timer, as you referred to above.

It would be useful to confirm if internet sources, like offgridgarage, are indeed correct that the new JKBMS is a fully-featured charger as well.

[Burgerman]

PB series "inverter" BMS's are more advanced than the old ones. In fact I have read that they do provide full charge control whereas the older models do not

No.
The problem a BMS has is that it is not part of the charger. It sits in between the charger and the battery.
The problem that this causes is this:

The PL8 and other hobby chargers can and do balance the cells at a specific voltage. In order that the cells dont exceed this the charger watches the highest cell voltage of say cell 3 and reduces current during this phase to allow the balance circuit to remome the full 1A from this top cell. Once this cell is actually full, that then limits the carge current to 1A max while still providing the correct CV voltage that we choose say 3.550V.

A BMS cant control that charger proportionally. So its got 2 choices. 1 allow overvoltage while the charger or supply keeps trying to achieve its set CV volatge at say 40A on high cells. That then rapidly results in over voltage OVP and so now it chops off the power... It then tries to balance while voltage decreases NOT at 2A as claimed, but 1 cell at a time (so 1/8th of the rated current spread across 8 cells because time) and it does so at its rated Max amps only where theres a huge difference in cell voltages. So its real balance current is way less than expected going by the 2A figure.

Its doing this, while the battery top cell gets pulled down... Once it reaches the RECONNECT voltage the charger starts its 40A again and rinse and repeat.

All BMS work this way. Whats really needed is a BMS that IS the charger/supply and then it will work exactly like a hobby charger in fully proportional fashion.
Thats why a hobby charger is better.
********

The NEXT issue is this:

The BMS doesent really determine when charge ends. The charger / supply does.
On the PL8 for e.g I set 3 parameters that determine this accurately.

I want ALL cells to be at the choosen CV voltage of 3.550V. Then if curret is low, balance current is also low, STOP!

I want all cells to have had as little ime at CV as possible because batts dont like to be held high/full any longer than neccassary. So here we need rapid constant voltage balancing to happen. BMS is way slower than PL8 that can balance at 8 seperate continuous 1A loads. While cell is at fixed set voltage.
I want the termination current to be at a figure I tell it. And I want the termination TIME to be:
ALL balanced and balance current at or below my figure + 15 mins at this state of the shortest CV cell time. And or my choosen termination current minimum. Whichever occurs first.

On a PL8 I can set these 3 vaues and the required logic and it does this at a fixed accurate CV voltage and ends when it sees all balanced plus 15 mins and my termination current whichever happens first.

A BMS cant do this as a power supply or lithium charger is on or off, (ovp/reconnect voltage) it also has no idea what the state of balance actually is, and so no clue how long the cells have each been at CV or what the individual cell voltages are. And the BMS cant control the power supply output to reduce as required to allow the balance current to hold a cell at a steady voltage. Just on/off.

[Burgerman]

With a GOOD BMS and most are crap, and a very adjustable charger for lithium like the ZXD then its is possible to get it to end up quite close to what we need. But it needs a lot of fiddling and fudging and fine tuning and testing.

While practically anything will work and charge the batteries in some fashion. Getting it properly right isnt easy!
All depends if you care really!

Also once we relax the huge number of safety nazi stuff that we just do not want to ever happen on a chair or that are irrelivant to our use then that side of the BMS is really pretty pontless too. More nusance that helpful. After you edge all hose parameters to settings that will never interfere in your daily use then what were they meant to do?

I can tell you what they ARE for. Preventing some bunch of retards using a too small lithium pack in some device/bike/scooter/whatever. They keep the incorrectly choosen cells from being damaged while cutting off power.

So its only really useful for 2 things.
Monitoring. (no need with 244Ah! Thats rediculous already)
Charging. So it can allow a simple 2 wire charger like the ZXD to charge a battery safely. If set up right. But you can do this by having the BMS external with the ZXD. At which point you look and decide to use the PL8 as it works even better!
Why externally? Because finding space for a BMS can be very hard. And because mosfets fail at some pont. Usually shorted. And thats where BMS fires you see on EV forums and EV bike forums were caused. The wires get hot the insulation burns and the car/bike/shed follows. So externally is safer as we cant get off and run away.

So theres a lot of things to consider.

[snaker]

"Inverter" JKBMS is designed for solar systems, it's a lot bigger than normal JKBMS. It can work for e-vehicles but almost the same as a normal JKBMS. Its advanced features are only when using with a solar inverter.

Just buy a dumb 8s charger with an Amps rating you want (e.g 30A). Or better a power supply that can adjust the output voltage and current (more expensive). All should be fine. It's not too complicated.

[shirley_hkg]

Thanks Shirley. Either the PSU I have is faulty or it cannot handle an initial surge of high current and just shuts down instead of regulating the current as when I plug it in, it just shuts off the output anyway. Not great :-(

I am interested in these PSUs that BM said you could supply. Is this something you could help me with? I'd be interested to know the cost and how to order.

Thanks!

You've got message.

[emilevirus]

"Inverter" JKBMS is designed for solar systems, it's a lot bigger than normal JKBMS. It can work for e-vehicles but almost the same as a normal JKBMS. Its advanced features are only when using with a solar inverter.

Just buy a dumb 8s charger with an Amps rating you want (e.g 30A). Or better a power supply that can adjust the output voltage and current (more expensive). All should be fine. It's not too complicated.

The "inverter" version has a larger heatsink so it can run at its max current 24/7. "Normal" one is made for vehicles. It'll overheat quicker.

[rune]

In case this is helpful for anyone:

If you are running your chair power through the BMS, make sure you set the charge minimum temp setting really low, like -20 or something. Why? Well because if the batteries are below the set value and hence the BMS turns off the charge MOSFET, current can only flow in one direction; battery to power module. When the the motors are in "overrun" (is the term overrun valid with electric motors?) , i.e. chair driving motors not motors driving chair, the generated current is not controlled as it doesn't have the battery acting as a sink because the discharge MOSFET is acting as a diode and only allowing current to flow from battery to BMS and not from BMS to battery. Because this causes the voltage to be higher than the preset limit in the power module, it trips the overvoltage protection and you get an error on the pod. In my case, 10 flashing LEDs. Once the temperature measured from the batteries climbs above the low temp protection recovery value, the chair behaves normally again.

Guess how I know this

The chair was left in the van overnight and it was about 3 deg C when my daughter tried to go into town for the first time since I finished working on it. Confusion ensued, util I suddenly connected the dots with the fact it was an over voltage situation and a couple of threads I'd been reading on here last night.

So, when the chair is being used, at all times you need current to be able to flow both to and from the battery.

[shirley_hkg]

That's why I connect chair to battery DIRECT, and via BMS for recharge only.

Burgerman warned to get around all those unnecessary features, that would strand us in mid course. That could be fatal if that happened at extreme cases.

[Burgerman]

In case this is helpful for anyone:

If you are running your chair power through the BMS, make sure you set the charge minimum temp setting really low, like -20 or something. Why? Well because if the batteries are below the set value and hence the BMS turns off the charge MOSFET, current can only flow in one direction; battery to power module. When the the motors are in "overrun" (is the term overrun valid with electric motors?) , i.e. chair driving motors not motors driving chair, the generated current is not controlled as it doesn't have the battery acting as a sink because the discharge MOSFET is acting as a diode and only allowing current to flow from battery to BMS and not from BMS to battery. Because this causes the voltage to be higher than the preset limit in the power module, it trips the overvoltage protection and you get an error on the pod. In my case, 10 flashing LEDs. Once the temperature measured from the batteries climbs above the low temp protection recovery value, the chair behaves normally again.

Guess how I know this

The chair was left in the van overnight and it was about 3 deg C when my daughter tried to go into town for the first time since I finished working on it. Confusion ensued, util I suddenly connected the dots with the fact it was an over voltage situation and a couple of threads I'd been reading on here last night.

So, when the chair is being used, at all times you need current to be able to flow both to and from the battery.

Its far worse.
Any time the BMS cuts off power for any reason at all then thats a huge problem. That overun voltage as you slow with your momentum, spikes and can destroy the mosfets and other electronics inside the controller too. Its inertial energy from the moving chair has no place to go and so voltage spikes goes super high.