Why BMS for lithiums are hopeless!

[Burgerman]

http://www.wheelchairdriver.com/BMS-lithium.htm

WHY BMS sucks... Draft. Needs finishing. Because I am forever having to explain the same thing year after year month after month!

The graph is for 4.2v charge LiPo but the same applies to LiFePo4 cells.

[bushytails]

Sounds more like the problem is cheap BMSes, not BMS in general. A properly designed BMS adds an extra level of protection for the cells, say if the charger fails, a load not controlled by the motor controller is left on, etc, and improves pack life.

Time to play devil's advocate...

The BMS's overvoltage is supposed to serve as a safety in the event of a failed charger, not to terminate a normal charge. Let's say the feedback loop in your charger fails (bad opto, whatever), and tries putting out full current non-stop... the BMS is to kill the charge and prevent catastrophic failure. If you're using your BMS to terminate a charge routinely, there's something else wrong! I would much rather a charger failure results in shortened cell life, not "venting with flame".

The BMS doesn't need to know the max amp draw for a low voltage cutoff, just the current amp draw, which it does know. A properly designed BMS is probably better at determining the shutoff point than the motor controller, as it also takes into account lighting or other loads. And a BMS monitors every cell, while a motor controller doesn't, and will happily run a weak cell down to where it fails.

Obviously a 30A BMS is mismatched to a 300A application. I wouldn't use #18 wire if I planned on pulling 300A, and I wouldn't use a 30A BMS either!

Balancing during use is a good thing, lets you get more runtime out of a pack, and prevents the problem of weak cells getting weaker faster than the others. A pack is useless as soon as the weakest cell goes undervoltage. Maybe most of your cells still have 10% usable charge left - but you can't use any of it if just one cell is low. Trying to use it further will destroy the weak cell. (And, as said above, the BMS knows when one cell is low, while the motor controller only sees the entire pack voltage, and will happily destroy the cell.) Even if you (smartly) don't run your pack all the way down, the weakest cell is still cycled deeper than all the others, making it even weaker still with every use, until eventually it fails before the rest of the pack. With flying capacitor balancing, the weakest cell receives current from the other cells, keeping its voltage and charge up longer, using more of the pack's total storage, and reducing the extra cycle depth of the weaker cells. You DO want it doing this. Of course, 100ma is pretty useless for this with larger packs. See above about properly matching bms to application.

.1 mohms is very, very low. At a 300A load, this is only 0.03 volts lost. This is far less than the voltage drop in your wiring, and a non-issue, assuming the number isn't fake - which it probably is.

I'm too tired to write more, and at a glance, the rest of it is pretty accurate. Bed time!

--Bushytails

[Burgerman]

The BMS's overvoltage is supposed to serve as a safety in the event of a failed charger, not to terminate a normal charge.

Wrong. That is the ONLY mechanism for controlling the charge voltage when using a charger of 29 to 31 volts as suggested and commonly used with a BMS and sold as a LiFePo4 charger. And since these are all CC/CV chargers they ALWAYS try to charge at max amps. The BMS is the device that keeps the cells from frying by cutting on/off while its ridiculously small balance circuits try to stop the highest cells from going over about 4 volts. Which happens all the time.

Let's say the feedback loop in your charger fails (bad opto, whatever), and tries putting out full current non-stop... the BMS is to kill the charge and prevent catastrophic failure.

No that isn't what it does. Its in use as soon as the battery reaches a almost charged state in cyclic fashion. Because SOME CELLS are high and above cut off point, and the rest are low. Even when the total voltage is OK.

If you're using your BMS to terminate a charge routinely, there's something else wrong!

Well there IS something else wrong! The whole principle of charging in this way. It allows the PACK to total 30v. Lets say charger is 29.4v (they are often higher) Already far too high, and damaging, But it allows a max cell voltage of the highest cell to reach 3.75volts, before CUTTING OFF CHARGE instantly to ALL CELLS. As you charge one or two cells will go sky high at the end, well before the rest. This always happens. And will continue happening right up to the point that the rest catch up. So as there is just a tiny balance current, the only thing the BMS can do is cut the power. This happens cyclically for hours to days depending on balance.

And if the stupid BMS has tried to balance the pack in the middle or bottom of the charge as you used it the day before, it may now also be several Ah out of balance at the top. Making everything much worse on subsequent recharge. Its miserable 100ma has had about 16 hours to screw up pack balance at the not full point. So it will add a massive amount of time to the rebalance at the top. I will post a charge graph so you can understand this. They really do shoot up sky high at 100 percent full. And the rest stay flat.

In other words if all cells were charged correctly to FULL at 3.600v, they may be at a mix of voltages at 50 percent, To try and balance at 50 percent means a massive Ah difference as each cells running voltage AT THE SAME AH POINT may be 3 to 7 Ah different.

[Burgerman]

Cant be bothered to re go through all the rest. Its endless.

EG cut off voltage. IT VARIES depending on load. A well designed computerised setup that KNOWS all the parameters can do this correctly. A dumb fixed value BMS that doesn't have all this info cannot know the loads, the cell characteristics, the recovery time, etc. The PROPERLY PROGRAMMED motor controller can know all of this much like the complex lead battery gauge they have built in, MUCH MORE ACCURATELY than the dumb BMS.

As for amps, lets see you find that 300 amp capable one? And its NOT REQUIRED if you fit a breaker/fuse/ or use an amp limited programmable motor controller rather than a simple e-bike speed controller with no limiting.

And you may not think 1 mOhm resistance matters, but that's what is limiting the power output to 30 amps. It heats up the mosfet and it dies... And it need/should not exist with a correctly designed pack/controller/motor system.

[Lord Chatterley]

http://www.wheelchairdriver.com/BMS-lithium.htm

WHY BMS sucks... Draft. Needs finishing. Because I am forever having to explain the same thing year after year month after month!

The graph is for 4.2v charge LiPo but the same applies to LiFePo4 cells.

In case you would like to put up some screen grabs from the manufacturers and suppliers-

3.65v...

http://www.ev-power.eu/CALB-40Ah-400Ah/SE100AHA-Lithium-Cell-LiFePO4-3-2V-100Ah.html

4.2v!!...

http://gwl-power.tumblr.com/post/576978080/faq-initial-charging-information-the-above-is

They also recommend charging thereafter at 3.65v as do the manufacturers:

http://en.calb.cn/product/show/?id[/url]
LC

[Lord Chatterley]

Last link screwed - ran out of time to EDIT!!! Grrrrrrrrr...time limit for editing is too short!

http://en.calb.cn/product/show/?id-600

LC

[Burgerman]

Heres why you CANNOT balance them in the middle and at the bottom as well as at the top at charge end. These cells were all at EXACTLY 3.600V +/- 0.002.

As you discharge them by in this case 17Ah each, the voltage spreads. It does NOT mean they are unbalanced, and each has the same capacity plus or minus say 2 percent. What it does mean is that during the day, you will tale power from the higher voltage cells (incorrectly) and put it into the lower voltage cells. This UNBALANCES the pack to the tune of 100ma (balance capability limit) over a full day. And so on subsequent charge will take a FULL DAY to correct!

Look at the highest cell on the graph. And the lowest. The highest one is still higher than the lowest 17Ah later!!! But they *ARE* BALANCED AND WILL RECHARGE TO 3.600V FAST. As long as the useless BMS doesent screw with it.

[Burgerman]

Remember all these cells start off 100 percent full, exact same voltage. And have exactly the same Ah capacity. And all drop like a rock at the same point INITIALLY, and when empty. But during the long flat part of the curve their voltage is NOT exactly the same. It has a spread of 0.013 volts. They are IN BALANCE.

If I balance them at say 17Ah of discharge, to get rid of that 0.013v spread, as shown, they will take some 20 hours to balance, and again on recharge another 20 hours to rebalance. And if a BMS is doing this it will allow one or more cells to repeatedly go over voltage as it cuts in and out for 20 hours! I have WATCHED this happen...

[bushytails]

As I said, you're talking about crap chinese fake BMSes, not BMSes in general. I wouldn't consider anything without some brains to be usable - and a proper BMS is much better at figuring out what the batteries are doing than any motor controller ever made.

I'm not sure why you don't want it balancing during use - this is a good thing. Unless you build a pack of perfectly matched cells, some will always have a lower capacity than others, just due to manufacturing tolerances. If during charge you balance these to all the same voltage, during discharge, these weaker cells will drop faster. At any given point during discharge of the pack, these cells are at a deeper %DOD, and use up more of their lifespan with every cycle than the stronger cells, making the problem worse with every use. And, since the pack is useless when any single cell goes undervoltage, you don't get to use the full capacity of your pack. Balancing during use prevents this. As I said before, 100ma is useless, and suggests a greatly undersized BMS.

A motor controller _does not know_ individual cell voltages. It can and will destroy the weakest cell in a pack every time, as it only sees the average cell voltage. For extreme performance, people build packs with matched cells - for everyone else, you might see a 10% variation in cell capacity - or a lot more if you buy chinese cells. Panasonic, generally considered one of the premier li-ion manufacturers, specs +/-5%. You _need_ to monitor individual cell voltages. A motor controller doesn't know the difference between (I'll use li-co numbers, since that's what I'm most familiar with) 8 cells at 3.0v, and 7 at 3.2v and 1 at 1.6v. Discharging without monitoring cell voltages is, if I were to hazard a guess, the #1 cause of early pack failure. Unless your motor controller is monitoring every cell (which I've never seen, as that's the BMS's job!), it will ruin packs if you run the batteries flat.

Many systems also have loads that do not run through any centralized computer - do you hook a headlight, 12v converter, etc up to your motor controller? The pack needs to be protected from these loads as well.

Why do you think a BMS doesn't figure in the internal resistance of the cells, and use that as part of its undervoltage cutout? It'll also be monitoring the cell temperature too. The BMS knows the discharge current, and will know the cells and li* characteristics better than a motor controller could or should. Chances are your BMS, assuming you didn't buy chinese crap, also can give you an good estimate of the remaining Ah capacity of the pack based on integrating the discharge current regardless of voltage (virtually all of them do this, even those without balancing), and for better ones, the capacity of every individual cell, and other failure prediction metrics.

If your BMS recommends just applying a random voltage to the batteries and waiting for the BMS to cut off on overvoltage, toss it in the garbage, and never buy anything from that company again. A BMS is _not_ a charger. The charger still needs to provide a precisely regulated voltage to the pack regardless of the BMS used, and the BMS should not be deciding when to terminate the charge with its overvoltage cutoff. A good BMS/charger combo will communicate charge status information to each other, but it's still the charger's job to make sure it never applies a voltage greater than the charge cutoff at average cell voltage to the pack.

What you are calling a BMS, we'd call a "protection board", and wouldn't use for any serious application. A BMS has a fair bit of brains, usually some form of interface to the rest of the system being designed, and will do things a motor controller could (and should) never do.

Off to work... another 13 hour day today. yay.

--Bushytails

[Burgerman]

I normally steer WELL clear of anything close to 3.600v as the capacity gain over 3.500v per cell is less than 0.2Ah. And lithiums hate being charged to high voltages, it has a drastic affect on lifespan. Especially above 3.6v.

The reason that the manufacturers recommend 3.65v or even higher, especially for initial charge on a new pack, is to BALANCE the cells correctly. An overcharge doesent add a thing to the already full cells, and in fact does *some* harm. But it pulls up the low ones. They do this because it helps the shitty bms to balance the cells on a fresh pack or you would be there all week! And they don't want "faulty" (unbalanced) cells back...

[Burgerman]

With a decent charge and balance system there's no need to go anywhere near the upper limits, and this makes it acceptable to charge to say 3.500v as here, or 3.550, or 3.600v. No need to go higher, and absolutely no benefit in doing so.

Quite the opposite in fact. Unless your charge system is crap, and it is needed to balance cells properly... BMS anybody?

Look here. This is a laptop 4.2v type cell, but all lithium cells are the same. See what going above its recommended 4.20v charge limit does...

[Burgerman]

THIS is the result of balancing at about 70 percent charged point. Shown here wasting 4 hours RE BALANCING after charging...

Remember that with a tiny 100ma instead of the hyperions 300ma balance this would have taken 12 hours by bms... Or longer as every time one cell goes above their too high 3.75 or more volts the cut off the charger and wait! On off on off on off... Still their too high volts helps the natural balance to assist the inadequate 100ma. The flaw is not the actual numbers they use, but the method. Its common and widely practised. But wrong. Just like religion.

[LROBBINS]

Bushytails,

Could you provide us a link to an example of what you consider a quality BMS for a large pack and high current use? ("Large" and "high current" in the WC context, not a 5 px car.)

Ciao,
Lenny

[Burgerman]

There are custom systems in use that combine the correct cells/C rate/Controller settings, other safeguards, and on board charge systems that work like the Hyperion and other integrated chargers. These are used by everything from laptops to full sized electric motorcycles and the auto industries cars.

The BMS that worries me are the dumb ones that are UNIVERSAL stand alone add ons, or in built to a universal lead brick replacement battery.

And there are some battery chemistries that don't need additional balance circuits and these are just balanced by a controlled overcharge like a lead battery.

[bushytails]

I'm not sure where you think I'm advocating overcharging, or saying the BMS you're using as an example isn't crap - I'm saying BMS, in general, is a good thing. I would not use any li* system without a BMS monitoring individual cell voltages. At some point, you will overdischarge, either the weakest cell or the entire pack, and then you've gone and ruined some very expensive batteries. I'd say an appropriate title would be "Why you shouldn't buy cheap chinese BMS units", not "Why BMS is hopeless". BMS is required to get the full life out of your pack. (without a good charger, you'll ruin your pack too)


As to links, here's some random search results. I haven't examined these for details, just glanced at them to check the rough size range and that they had a reasonable feature list. There's a lot more options for full-size EV systems, but they'd be overkill. (Although I'm sure burgerman could build a chair needing 2000A continuous from 130 cells....) And then there's the industrial solar storage systems, with 60,000 3000Ah cell modules...

http://www.irf.com/product-info/bms/
http://www.lithiumbalance.com/battery-m ... stems/blog
http://www.123electric.eu/
http://www.rec-bms.com/
http://www.claytonpower.com/products/bms/
http://pacificev.com/t_battery_management_system.html
http://www.reapsystems.co.uk/products
http://elitepowersolutions.com/bms-overview.html
http://hybridpropulsion.com/cellcard (Note: Explicitly lists bottom-of-range balancing as a feature)
http://www.evaira.com/Energy%20Management.html
http://elithion.com/battery_management_systems.php

Ok, I'm bored now.


Why do I always get "The submitted form was invalid. Try submitting again." when I post?


--Bushytails

[Burgerman]

OK. The reason I say they are crap is because I HAVE examined many more than you listed of the sort of cost and size used by e-bike or small car conversion enthusiasts, and a few that are the correct sort of size to use on a chair.

I used that particular BMS as an example because all of the ones I have seen, work exactly like that, and the values are typical.

There ARE complex external purpose programmed SYSTEMS that use some parts of a the BMS to monitor and control things that work. But non are affordable, or simple enough or are STAND ALONE that bolts to a pack and uses a dumb 2 wire simple charger.

As for always needing a BMS that's rubbish. Running them too low, can be a simple bit of code added to the motor controller that monitors the total voltage, that KNOWS the load, and the cell resistance etc and can safely advise when to stop. Or force you to stop. This happens on a well built lithium pack at almost exactly the same time on all cells, and well before an individual cell goes too low. They all stay pretty damned solid and almost the same voltage from 98 percent full to about 95 pecent used up. Then they start to fall off a cliff. So its easy to see this end point BEFORE any cell goes anywhere near the 2v that the BMS typically looks for... Which is way too late anyway.

And a motor controller can also monitor cell voltage individually. As my model heli does.

You only need a BMS if you are using a badly designed system, or a crap no cell balancing charger, or cells that do not have adequate C rate for the purpose, or a controller that has no current limiting and a motor with too low impedance. Design the system correctly and non is needed.

Tell me, why do I need one on my Roboteq BM3 chair? Answers on a postcard. Or here if you can think of a reason.

[Burgerman]

ONE thing the BMS does that is bad, is control CHARGE. Its not the ONL;Y thing, but we don't need those either.

Remember that UNLIKE the charger here that controls the AMPS fed to the battery in conjunction with the balance circuits and has very fine control, the BMS can only do the following:

Watch the volts go over the top, and TURN OFF the charger supply. Then when the voltage falls back, turn it ON again... On off on off . Over volt under volt, rinse and repeat until the high cells are pulled down by the feeble balance circuit to a (already too high) 3.75 volts...

Now look at this graph. This is what SHOULD happen...
I charged this tiny 4Ah LIPO battery from an almost flat 3.750v per cell (balanced at the BOTTOM on purpose which is subsequently then wrong at the top) to 4.200v per cell.
I did so at a huge 20 Amps. Flat to full AND BALANCED in 30 mins. That's a 4Ah battery! Or solid 5C charge rate. Only the balance slowed it down. If this was a 100Ah battery that would be an impossible 500Amp charger!!!

Now as you can see its fine at this 20Amps, until the highest cell reaches 4.200v - at which point the charger progressively 'throttles back' till the 300ma balance circuit can hold the cell down at its exact 4.200 volts MAX.
It ends up charging at just 300ma...
Eventually this battery is completely full, charge current zero, all cells at 4.200v (=/- 0.003v) and perfectly balanced with ZERO overvoltage in the minimum time.

A BMS cannot do this - not because its not accurate (although it generally isn't) but because the charge METHOD is flawed. It cannot control the chargers output. Its not possible to do it that way properly. The METHOD is wrong. A dumb CC/CV charger cannot correctly charge a bunch of unbalanced cells.

[butchec]

Hi,

Since the debate is about BMS, Lithium and Charging.

My questions is... Compared to the Hyperion charging method; Is there a Controller(BMS or otherwise)/Charger that is currently being manufactured and availiable in the UK?

Also, If you are using the standard Power Chair controller/motors etc and only replacing the batteries to Lithium, What is the best option / best practice for keeping the batteries in peak performance?

Cheers.

Chris.

[Burgerman]

Is there a Controller(BMS or otherwise)/Charger that is currently being manufactured and availiable in the UK?

That makes no sense??? I read it many times.

ANY controller that works with lead 24v batteries will WORK with 8 cell (or multiples) of LiFePO4 cells and charged to 28.800 volts. Or 3.600v per cell.
Some odd lithium chemistries recommend higher voltages. These may or may not trigger over voltage errors.

None will show BATTERY CHARGE STATE correctly. They will all say 100 percent all day long, and as soon as it drops a fraction, then the lithium is all but empty... That COULD be addressed by firmware. And they COULD be made to shut down at this point so no need for BMS to stop you at the wrong point. These things would only require a firmware fix.

As for charging, any 8 cell LiFe capable hobby charger that does enough amps with balance connectors, can charge the chair correctly and accurately.

And these things, along with the CORRECT CELL CHOICE and capacity choice to be sure you go nowhere near the cell max cont C rate are all you need to throw the BMS away.

As we cant do the firmware thing, ourselves, so you only need an external method of monitoring LOW voltage when OFF LOAD to find the empty point. But this difference is very marked and obvious so your backside can tell you this anyway.

[Burgerman]

Also, If you are using the standard Power Chair controller/motors etc and only replacing the batteries to Lithium, What is the best option / best practice for keeping the batteries in peak performance?

1. NEVER use a dumb charger and a BMS as that will screw them up fast.

2. Charge with a hobby type balancing charger.

3. use a battery monitor, or voltmeter, or your backside to tell you when they are depleted! (To check you are correct, read Ah BACK IN as you charge with 2. above.)

4. make sure your PEAK amp draw is less (with a good margin) than the packs continuous rate in Amps.

You are done!

[LROBBINS]

I looked at each of the sites Bushytails linked to, but must confess that I didn't have the patience to work past the claims to try to figure out what each of these, each one of which claims it's better than the others, actually does, nor to try to get an idea of pricing - which seems pretty well hidden. I did notice that one of them pairs their BMS with the same Chinese charger that I'm using (programmed for Pb); so much for the blanket condemnation of Chinese stuff - some of it is very well engineered indeed.

It seems to me that balancing cell discharge rates, carefully balancing them before first use, using the Hyperion to charge and top balance does most or all of what would be needed to keep the cells healthy. I do wonder, however, whether something could be done, simply and cheaply, to have a bit more reassurance as one approaches discharge. I am not too concerned about protecting the battery when the whole battery is deeply discharged; keeping track of voltage under no, or low, loading should reveal the start of the knee. That could even be automated; set up the motor controller to read voltage when battery current is e.g. less than 10 Amps, and reduce max speed if the voltage starts to sag. That should protect the batter and, as opposed to the "cut off" imposed by some of these "sophisticated" BMS and all of the cheap ones, should let you get home and not cut out on the first steep ramp you hit.

What I'm wondering about is knowing if one series string has, with age, become weaker than the others. Burgerman will probably see this already because he always checks the charge curves he gets from the Hyperion, but I think that it wouldn't be hard to have real time, automated, monitoring as well. I wouldn't do this by looking at absolute voltage; to avoid false positives that voltage cutoff would have to be damagingly low. Instead I'd compare the voltage of one string to the average of all of the others. Given that the balance leads are already there, this would just mean reading each balance lead with an ADC and doing some simple calculation. For 8 strings (24V), you could even do all this with just an Arduino Nano as it has 8 analog pins; for more than 8 strings you could either go to a fancier processor or just multiplex read the balance wires. Then feed a digital output off that microprocessor to the motor controller if any string is more than 1 or 1.5 standard deviations low compared to its brothers. In effect, this would also detect loose connectors as well as iffy (strings of) cells.

Ciao,
Lenny

[Burgerman]

Quite.

I look at the individual cell resistance on the Hyperion, it shows a loose connector, bad cell etc early.
I also look at voltages after use, and so on. They are remarkably identical, and all hit the discharged point at damned near the same time. So there's really only a need for the controller to look at complete pack voltage - as long as you know what's happening. But I doubt that even this is important. LiFe cells are very tolerant on discharge compared to say LiPo. I have allowed plenty of hobby ones to run until dead by leaving stuff switched on. They all still work fine. But its obviously not a great thing to do to them. Its overcharge, or over C rate they really don't much seem to like.

[ex-Gooserider]

My take on that question is that he was asking if there was a charge controller / BMS made and with UK availability that would let you do with a dumb (typical mobility?) charger what you do with an Hyperion... It would seem like it MIGHT be possible from a technical standpoint if you had a fairly high current balance circuit (that possibly only worked when a charger was connected) and a way of providing feedback to the charger that would make it go to an appropriate CV value and keep the current at a point where it wasn't overpowering the balance circuit...

No idea if something like this couid actually be built, but it seems like it might work...

ex-Gooserider

Is there a Controller(BMS or otherwise)/Charger that is currently being manufactured and availiable in the UK?

That makes no sense??? I read it many times.

ANY controller that works with lead 24v batteries will WORK with 8 cell (or multiples) of LiFePO4 cells and charged to 28.800 volts. Or 3.600v per cell.
Some odd lithium chemistries recommend higher voltages. These may or may not trigger over voltage errors.

None will show BATTERY CHARGE STATE correctly. They will all say 100 percent all day long, and as soon as it drops a fraction, then the lithium is all but empty... That COULD be addressed by firmware. And they COULD be made to shut down at this point so no need for BMS to stop you at the wrong point. These things would only require a firmware fix.

As for charging, any 8 cell LiFe capable hobby charger that does enough amps with balance connectors, can charge the chair correctly and accurately.

And these things, along with the CORRECT CELL CHOICE and capacity choice to be sure you go nowhere near the cell max cont C rate are all you need to throw the BMS away.

As we cant do the firmware thing, ourselves, so you only need an external method of monitoring LOW voltage when OFF LOAD to find the empty point. But this difference is very marked and obvious so your backside can tell you this anyway.

[Burgerman]

It would seem like it MIGHT be possible from a technical standpoint if you had a fairly high current balance circuit (that possibly only worked when a charger was connected) and a way of providing feedback to the charger that would make it go to an appropriate CV value and keep the current at a point where it wasn't overpowering the balance circuit...

No idea if something like this couid actually be built, but it seems like it might work...

It is possible. That's EXACTLY what the hobby type chargers are doing... Very accurately.

[butchec]

It would seem like it MIGHT be possible from a technical standpoint if you had a fairly high current balance circuit (that possibly only worked when a charger was connected) and a way of providing feedback to the charger that would make it go to an appropriate CV value and keep the current at a point where it wasn't overpowering the balance circuit...

No idea if something like this couid actually be built, but it seems like it might work...

It is possible. That's EXACTLY what the hobby type chargers are doing... Very accurately.

Thank you GR for clarifying what I was trying to ask for.

Althogh the Hyperion is a wonderful/perfect charger, it is not going to be used by me. Also the Hyerion is not the same as a builtin BMS. It is a separate charger and BMS. Plus it is not available in this country without paying import costs.

What is am after is a unit that has all if the "Special" features the Hyperion has, but it needs to be builtin and currently for sale in the UK to ensure warranty and service is available.

BM Is there anything out there that offers this that matches the Hyperion? Except the Hyperion

I will not be the end user as I work long days and occasionally away. It needs to be as simple to use as a standard single connection charger with an on/off switch.

Cheers. Chris

[Burgerman]

Every single hobby charger charges lithium's this way. There are thousands of different ones available that charge just lithium polymer, or life as well, or Nickel variants, lead too, in 1 to 14 cell and 1 to 200 Ah, or 1 to 40 Amps. The Hyperion is available in the UK anyway, and I have 5 bought here. But that's not the point.

The problem is that if you cannot control charger output, as with a typical generic BMS, you cant charge the lithium's properly. And that the Hyperion is really not intended for powerchair charging and it does way more than 99.9 percent of wheelchair users ever need or understand. But since I helped develop it, I know how involved it all is, and that it works well and how bad most of the rest are!

What exactly are you looking for?

[LROBBINS]

I think that what a lot of people are looking for is a system that, once set up properly, can just be turned on and will do its thing. I think that the Hyperion could be made to work that way, perhaps needing to add a little logic to replace the several button pushes that it needs now, housing it along with a power supply in a reasonably portable case, and fixing the bullet connectors in place. Alternatively, the Hyperion could be mounted on the chair so that the balance leads could always stay plugged in and one would plug and unplug the DC supply (if that's safe for the Hyperion - I haven't re-read the manual of late). That's also why I suggest automating a bit on the discharge end as well:

Quite.

I look at the individual cell resistance on the Hyperion, it shows a loose connector, bad cell etc early.
I also look at voltages after use, and so on. They are remarkably identical, and all hit the discharged point at damned near the same time. So there's really only a need for the controller to look at complete pack voltage - as long as you know what's happening.

You "look at the individual cell resistance on the Hyperion", but how many non-technerd users want to be doing that?
"They are remarkably identical, and all hit the discharged point at damned near the same time." That's true now when the cells are fairly new and haven't ever been abused, but will it be true after a year's heavy, and inattentive, use?

Ciao,
Lenny

[Sully]

Lenny,

From my standpoint you have hit upon all the questions and their answers, in this one post. In other words you get it.

No average power wheel chair user has the expetise that the Hyperion charger and I must say LiFePO4 batteries require. Many do not do any of the maintenance things, like battery charging, themselves. Turn it on turn it off, look at a light or two to see if it is charged, that's all.

[Burgerman]

Quite. So if you can understand these things its easy to change to lithium.

If not you will need to wait until the retarded powerchair manufacturers catch up, and fit properly designed embedded systems like the non mobility scooter, e-bike, full sized automotive, camera, laptop and phone market has done many years ago.

The simple "wait till they make LEAD BRICK replacements" that the dummy wheelchair manufacturers seem to be waiting for will never happen or at least if it does it will always be a 2nd best setup, that doesent take proper advantage of the technology.

It needs several things.
1. One either more volts, or more use making of the higher C rate possible. (A 50percent GAIN in either battery range, or speed, or smaller battery is possible here alone)
2. An EMBEDDED lithium protection system, as far as over discharge, battery remaining etc.
2. A safe charge system suited to the batteries used. Likely ON BOARD and cell balancing, like an on board "dumb£ version of a Hyperion.

This is what the rest of the world do. To a better or worse degree.

Lead brick swap? No.
Dedicated user friendly embedded system? YES!

Until they do this, the Hyperion is really the best way to go. Or any suitable hobby type charger. And some other means of clocking Ah used. And cells that can cope with your peak Amp demand.

[ex-Gooserider]

This harks back to the same discussion we had a few months back where Nandol was looking for a charger that his caregivers could use - where the skill levels were challenged by just plugging in a typical mobility charger...

Certainly it would give up the flexibility and power of the Hyperion, but a BMS like system that moved the "brains" of the system from the charger to the battery box, so that the charging technique would be about the same as with todays mobility units would have a lot going for it, and (I think) is pretty much what butchec is looking for.

While it might also save and keep the charging info available, I think it would also need to have some level of ability to self monitor and flag if it sees some signs of problems (i.e. chronic imbalance, or more than a certain amount "out" and so on) rather than relying on the user to give it the attention that you do.

ex-Gooserider