Bafang and battery disagree.

[Everso]

Hi Peeps - new here and learning all sorts of interesting things about my new electric trike.


I have fitted a 26inch rear wheel with 500W BAFANG G.020.500.DC motor - 22A KT Controller and KT-LCD8H display. All is going well and it makes my 11 mile commute to work quite pleasant. I have however noticed a strange disagreement between my display and the battery. Towards the end of my commute the display is showing 1 or maybe 0 bars left (did cut out one day but lucky - I carry a spare battery) but when I press the battery check button on the battery it's self it is showing 3 out of 4 lights.

Given that the BMS on the battery should know the battery better than a generic controller I am more inclined to believe it is correct. That has got me wondering - am I leaving "juice in the tank" when the controller cuts out for low voltage? From the full manual a kind user here posted I see that I can set parameter C12 to go down to a lower voltage. So I guess two questions:

Would this increase my range?
If it does try to take the voltage too low would the built in BMS on the battery shut down and prevent damage?

 

[amberwolf]

Adjusting the controller LVC down might give you more range, but it depends on why this is happening:

Did it used to work differently, and if so, was the change sudden or gradual?

When it cuts out, does the display go blank or does the motor just stop providing assist?

If the former, it means the battery's BMS is shutting off to protect against low cell(s). The latter means the controller is turning off the motor at it's LVC to protect the pack against overdischarge.

Normally if there are multiple bars left on the battery meter it should mean there's a fair bit of capacity left...but if the current draw of the system is high compared to the battery's capability (not uncommon for cheap batteries, but happens even with better ones if aged, damaged, or sized too small for the system), the battery voltage may sag under load so much that at that point in it's capacity it drops to "empty" voltage, and when the load is removed the voltage bounces back to what you see when stopped.

(that can also happen if the battery is unbalanced, so that some cells are lower in voltage than others (which happens because the cells are different from each other and so don't perform the same or provide the same capacity))

Do you have a link to the sale page for the battery?

If it's possible to safely see the battery meter while riding, you can check for this problem. (the display's meter probably averages over significant time, and if so won't show you sudden voltage sag).

The battery's meter might also be the wrong one for that battery, and so show a higher level than it should. This you can check with a voltmeter and measure the battery's voltage at full charge, and when the controller shows "empty". The voltage you see should help figure out what's happening.

 

[Everso]

Hi Amberwolf,

First off - thank you for taking the time to write such a detailed and useful reply… much helpful information to think about.

I think a little history would help here. I bought this trike with a 1K motor 48v and a 9ah battery. I needed more range so bought a 14ah 48v battery to go with it (I have them in a rack mounted box and carry both). This seems like a big expense/hassle but my 12 mile commute involves a toll tunnel that is free for bikes so I am saving £8 each day I commute by trike. That’s going to add up fast - not to mention the fun/exercise elements of my trip. Time wise it’s only 10 minutes longer so it's a win all the way. After a while it turns out the controller/motor/something is not at all happy - the controller starts overheating and crapping out. So I bite the bullet and get the 500W motor - new controller/display.

Given that the batteries seemed happy enough to drive the 1K motor I think that takes out one concern you listed of current draw being too high. Both batteries seem just fine - are able to drive the lower powered motor with a sustained load and are not getting hot.

This setup is still new (maybe a 100 -200 miles on it so far). I have only “sucked it dry” once and that was the controller saying it was empty - the display was still on but showing 0 bars and not engaging. This was new to me so I just swapped batteries and didn’t think to check the built-in lights.

Moving on - both batteries are showing this behaviour - low reading on the controller but 2 or 3 lights on the battery so I am very much leaning towards the controller being overly protective (or maybe just reading wrong).

So - and please correct me if I am wrong here - I can safely change the controller in 0.5V increments. When I get to a point where the controller turns off rather than saying “battery empty” I then back off a step as that means I have reached the point where the battery is shutting off to protect itself.

My main take away - again please correct me if wrong - is that the battery will protect itself and not allow the controller to damage it.

Once again - thank you for your help - it does look like there is much more range available. This was my main hope in switching from a 1K motor to a 500W. I am not bothered about the top speed (this was about 30MPH but is now 20-25ish so still plenty fast) but in being able to climb hills and get into work without needing a shower. This very much seems to be working.

 

[marvin78]

Edit: don't use this table it's not the correct one.

Measuring says more as what either your display or the lights on your pack say. My display doesn't show volts, so I check percentage with a multimeter before a long ride and after.

Controller information afaik isn't just a rolling average it also stores values on shutdown and checks those at startup. If I had an empty battery when I got back, charged up the battery outside the bike, the moment I turn everything on it first displays the charge I came home with the day before, then after a few moments it jumps to the current estimate.

So - and please correct me if I am wrong here - I can safely change the controller in 0.5V increments. When I get to a point where the controller turns off rather than saying “battery empty” I then back off a step as that means I have reached the point where the battery is shutting off to protect itself.

Don't trust either indicator, measure first. Find out what is happening, the correlation that two batteries both indicate something with those lights and your screen something else still leaves in the middle which one is right.

edit: just to make sure; you can change lvc and have it drain to the point where your battery has only the low warning indicator. But you said it worked before. So what if you change lvc now, to correct whatever is going wrong, so the controller actually thinks it draining them more when in fact it's not.

Now that 'something you haven't figured out, which caused all that to start with' suddenly changes back. Maybe you had a loose connection / bad joint somewhere on your controller pcb which jolted loose on a shock... and then it it makes good contact again after another bump at just the right speed and angle..

Then your controller will suddenly try to drain your battery to the exact lvc you set which will be damaging for your batteries.

I know it's far fetched perhaps, but batteries are expensive. More expensive as controllers. I rather wouldn't risk damaging the more expensive part instead of checking why the less expensive part is causing issues, it doesn't make much economical sense.

 

[Everso]

edit: just to make sure; you can change lvc and have it drain to the point where your battery has only the low warning indicator. But you said it worked before. So what if you change lvc now, to correct whatever is going wrong, so the controller actually thinks it draining them more when in fact it's not.

Sorry if I caused any confusion. When I said worked before this is in relation to the batteries being able to supply the old 1K motor. The 500 is new and I only have 100-200 miles on it so it's all newish.

That chart is however very interesting. Acording to the Bafang manual it is sett to the default cutout is set to 40V - well below what that chart would suggest as empty - and at that point I still have several bars on the battery. I am starting to think the measured voltage on the controller is just not very accurate.

 

[marvin78]

Wait, sorry I posted the 48v chart judging from your comment?

Nvm really, the important part is

I am starting to think the measured voltage on the controller is just not very accurate.

Do you have a multimeter? Measuring is knowing.

To get the correct table(s) you could try a site like this -> Li-Ion Ebike Battery Charge Charts

There are many other places, I'm pretty certain they are posted around here as well.

That chart is however very interesting. Acording to the Bafang manual it is sett to the default cutout is set to 40V - well below what that chart would suggest as empty - and at that point I still have several bars on the battery.

I will edit that post as I think that table is wrong, use the link I posted above. 40v ~ 6% or 7%. You don't want to think of those last % as left on the table, as you shouldn't discharge your cells that low.

Either way, all academic. Measure your battery voltage the moment your controller shuts off. See if the issue is in the controller. Theoretically I think it shouldn't matter since the battery bms will cut off current if it detects the cells are getting to low.

edit: I try to keep my batteries above 42v, and not dip into the last 20% capacity. Been told I will be payed with a much longer battery life. So I wouldn't worry about a 40v offset being to high. Which is why trying to correct that by just setting an even lower offset in your controller is NOT a good approach, imho.

But if someone has another opinion, please step in

 

[Everso]

Been told I will be payed with a much longer battery life. So I wouldn't worry about a 40v offset being to high. Which is why trying to correct that by just setting an even lower offset in your controller is a good approach, imho.

Yeha. I'm still new to all this (electric not trikes). Also - I guess - this is a different approach to a lot of use cases. I want to be able to commute with a minimum of fuss in as many conditions as I can (spending a lot of effort on waterproofing things). This is to be used as a utility work horse that will save me £8 per day. If part of that convenience comes at the expense of shortened battery life so be it (as long as it's not too bad). The money saved will pay for a new battery. The maths I am doing is £8 per trip - so if the battery lasts 1000 cycles it's saved me £8000 - with a replacement cost of £400 or so. If riding it hard reduces that - even by say 20% it's still going to save me £6,400. I regard that as an absolute deal. I know there will be other costs - tires - chains etc but I still think I will be better off by a decent amount and that's not taking account of the exercise/fun/no traffic.

I have a manual trice classic as a warm weather pleasure craft and that will be used for fun weekend trips and pure exercise.

Theoretically I think it shouldn't matter since the battery bms will cut off current if it detects the cells are getting to low.

That is what I was hoping to hear. Now that a couple of people who know more than me have confirmed this I feel confident it a slow and safe process of dropping it a step at a time will not kill any thing.

 

[marvin78]

I had to edit my post as I obviously missed the word 'NOT' in that quote. I do not think it's smart to set lvc to even lower to correct something which might not be measured well by something.

Get one of these

Then measure your voltage at the battery prongs when your controller says they are empty, but your battery indicators think not. Look for source of the problem, not for the workaround. Don't trust you diagnosis on the indicator light of your battery

That is what I was hoping to hear. Now that a couple of people who know more than me have confirmed this I feel confident it a slow and safe process of dropping it a step at a time will not kill any thing.

PLEASE NO. Wait for others, I am not an authority on battery management My first reply was mainly to point out you should really measure your voltage.. which still seems to be main point I am trying to make but keeps getting lost in translation.

Your BMS 'should' protect your cells. But you're already very low on discharge state, and you do not want to milk that last mile out of those batteries unless you plan on replacing them often. I strongly advice AGAINST trying to fix a problem which we don't know the actual cause of with doing something which would damage your battery and cost you money

edit: If you're going from 100% to as low a state of charge as can milk, don't expect 800 cycles. I was told that you over provision by 40%, and prefer to keep your soc between 80% and 20% as this should last me 3 years with decent quality cells. Now I just been told I shouldn't use the bottom 30%.. and I guess it's all depending on how much you value range vs longevity.

 

[Slowchris11]

I had a commercial build new 36v battery, which would suddenly, when 75% capacity where reached on the e-bike display, jump to 5% and then cut off.
When I stopped and checked on the battery's own indicator, it still had just one bar missing =75%.
I could always restart the bike and drive like normal for a short while, until it cut out again. This was an absolutely repeatable behavior.
Turned out some welds in the battery pack where faulty, so basically of 4 cell's connected in parallel, only one was discharged. While driving the contacts heated up and failed, just stopping for a short while made the connecting stripes cool down and reconnect.
It worked like it was planned that way. A perfectly reliable fault.
Got a new battery and all was well.

 

[amberwolf]

I think a little history would help here. I bought this trike with a 1K motor 48v and a 9ah battery. I needed more range so bought a 14ah 48v battery to go with it (I have them in a rack mounted box and carry both).

If you have both, you can make sure they are both always charged to the same voltage, and parallel them. Then it is easier on both batteries, as they share the load at all times, and you may even get marginally more range because of less voltage sag and less wasted power as heat inside the batteries.

I recommend only doing this with batteries that have BMSes with common charge/discharge ports, rather than separate C/D ports, so that they can block input or output as needed in various situations where one pack has a problem but the other is working. It also means that you can charge them at the same time from the same charger without risk, and without disconnecting anything.

If you parallel two batteries with separate C/D ports, you should disconnect the D ports when charging (even if paralleling the C ports), so that the BMS can safely disconnect the cells whenever it needs to (if the D ports are paralelled, it can't). When discharging, the C ports need to be disconnected from each other.

If you do not use regen (and disable it in the controller if possible to prevent accidental voltage spikes), you can use "battery combiner" boxes / splitters, that have diodes or ideal diodes, etc., to prevent backflow from one battery into another. But these waste a little power, so if you don't have to use one, it's more efficient.

This seems like a big expense/hassle but my 12 mile commute involves a toll tunnel that is free for bikes so I am saving £8 each day I commute by trike. That’s going to add up fast - not to mention the fun/exercise elements of my trip. Time wise it’s only 10 minutes longer so it's a win all the way.

I commute, shop, etc, by bike (trike nowadays) only, no car needed, so to me it's not a hassle. (see my SB Cruiser trike thread, and my CrazyBike2 thread, for the two most recent long-time-usage heavy-cargo-hauler builds).

Given that the batteries seemed happy enough to drive the 1K motor I think that takes out one concern you listed of current draw being too high. Both batteries seem just fine - are able to drive the lower powered motor with a sustained load and are not getting hot.

Until you measure, you don't really know. Best is to measure what the old system drew, and then measure what the new system draws, but if the old system isn't installed and isn't working right, that's hard to do. Measuring the new system's actual load, and comparing to what the battery is "rated" to do, would tell you if there are current peaks or sustained times of higher current than the battery is meant for.

Without a current measurement, then measuring voltage sag under various loads on a ride will tell you something (sometimes quite a lot). (sag being the drop in voltage while loaded vs while not loaded).


Watt ratings for systems don't necessarily mean much--mostly for motors, but some systems just aren't rated correctly because the manufacturers and especially sellers can say (and label) anything they want on things, and most buyers have no way to tell (even if they cared). For your situation I recommend looking at the current limits of the controllers involved, and any settings they have in their displays that affect those. This way you are comparing things in the controllers that you can also see in the battery specs, and are easy enough to measure to find out if they are correct and/or causing any of the problems.

Also, some controllers don't have good limiting of their current, and may allow much higher current than their stated limits, at least for short periods.

Moving on - both batteries are showing this behaviour - low reading on the controller but 2 or 3 lights on the battery so I am very much leaning towards the controller being overly protective (or maybe just reading wrong).

Could be. You'll have to measure the voltage you actually have at the controller vs at the battery while under load to reveal connection problems, (difference between the two indicates problem, identical eliminates that)

Knowing the actual voltage at time of controller shutdown, while the load is still there, and after you've stopped, will also tell you things. If the voltage doesn't rise between the two, the battery isnt' sagging much. If it rises a lot, it's probably too loaded down.

So - and please correct me if I am wrong here - I can safely change the controller in 0.5V increments. When I get to a point where the controller turns off rather than saying “battery empty” I then back off a step as that means I have reached the point where the battery is shutting off to protect itself.

That is one way to eke out the last usable energy from a battery...but not knowing what the battery is really doing could mean being harder on it than before, and aging it faster (which can mean shorter and shorter range).

My main take away - again please correct me if wrong - is that the battery will protect itself and not allow the controller to damage it.

It *should*. That's what it the BMS is there for...but:

The BMS is meant as an emergency protection system, so that if the cells reach some limit, it will turn off to protect the cells. Some of them are not well-designed and may fail to protect in some situations, such as turning off under load (which is what can happen with a controller LVC set very close to the BMS LVC and a high enough load relative to battery capability). In these situations, the FETs on poorly designed BMSs can fail stuck on, so the BMS can no longer protect against anything. (there's a thread by Methods about a version of a Daly BMS that did this; don't recall the situation that caused it). The more often the BMS has to protect the cells, the more likely a failure is on one that isn't well-designed (or cheaped out on it's parts, etc). Until it happens it's hard to know if any particular BMS is prone to this.

 

[Everso]

If you have both, you can make sure they are both always charged to the same voltage, and parallel them. Then it is easier on both batteries, as they share the load at all times, and you may even get marginally more range because of less voltage sag and less wasted power as heat inside the batteries.

I would love to do this BUT ... one battery is a 9ah and the other is a 14ah. Just to make it more fun the 9ah is a year or two older compared to the new 14ah. From what I found when looking into this it seemed there were issues with two identical batteries - let alone two mis-matched.

I figure in the few years it takes before these are no good a new formulation will be all the rage and I'll just buy a 100ah battery that is only 500g and costs less than £100 - a guy can dream

 

[Everso]

I have a multi-meter and will take readings before I make setting changes. Just wondering if one of these would be worth it for £10 - wire it in all the time and get real time data.

https://www.amazon.co.uk/JZK-Motorcycle-Voltmeter-Waterproof-Display/dp/B09V2SHF67/ref=sr_1_5?crid=2XI5DRRL56V5M&keywords=waterproof+voltage+meter&qid=1691313430&sprefix=waterproof+voltage+meter%2Caps%2C98&sr=8-5

 

[marvin78]

Double Battery Discharge Converter For E-bike 20A/30A/40A Dual Battery Pack Switch Balancer 24-72VDC

Double Battery Discharge Converter For E-bike 20A/30A/40A Dual Battery Pack Switch Balancer 24-72VDC

www.pswpower.com

This will drain two same voltage but different Ah ratings without risk of damaging the lower capacity one. It's not to be used for charging.

I have a multi-meter and will take readings before I make setting changes. Just wondering if one of these would be worth it for £10 - wire it in all the time and get real time data.

https://www.amazon.co.uk/JZK-Motorcycle-Voltmeter-Waterproof-Display/dp/B09V2SHF67/ref=sr_1_5?crid=2XI5DRRL56V5M&keywords=waterproof+voltage+meter&qid=1691313430&sprefix=waterproof+voltage+meter%2Caps%2C98&sr=8-5

• This 12v voltage meter is suitable for detecting DC motor vehicle 12V battery voltage. Voltage test range: DC8V ~ 16V) . When the voltage is lower than 11.5V or higher than 19.8V, the number of the voltmeter will keep flashing and alarm.

I don't think you're gonna get much use out of that.

https://www.amazon.co.uk/Battery-Setting-Capacity-Indicator-Monitors/dp/B093KTM1PS/ref=sr_1_3?crid=NHCOPM3JLEFX&keywords=waterproof%2Bvoltage%2Bmeter%2B60v&qid=1691313907&sprefix=waterproof%2Bvoltage%2Bmeter%2B60v%2Caps%2C84&sr=8-3&th=1

Seems like a better option

 

[Everso]

I don't think you're gonna get much use out of that.

Annnnnnd ... that's what happens when you are not paying attention. Thanks for catching that. I have canceled the order and placed one for the meter you kindly found me.

 

[Everso]

This will drain two same voltage but different Ah ratings without risk of damaging the lower capacity one. It's not to be used for charging

Ohhhh - this is interesting. Both battery packs have dedicated recharge sockets so the "not to be used for re-charging" does not sound like it will be an issue. It would be nice to plug both in and not worry about swapping. I think I will get the voltage meter I just ordered and get some date to see what is going on in there. When I have got this all sorted I will be very keen on getting one of these in use.

Are these things a good known quantity and reliable?

 

[marvin78]

Are these things a good known quantity and reliable?

Ask me in a few years.. No seriously I can't tell you,. I have on coming in but don't own one.

Also, there are different 'versions' floating around. I read about switchable one's, and one which is 'smart' enough to first drain a more full battery if the charges aren't equal, and only start sharing load when both are the same voltages.

Afaik the shop I linked has an excellent reputation ( they do with me ). I don't expect them to carry 'crap'

 

[Everso]

OK Gents - we have our first data point. I had to go into work today to pick up my keys I left there on Friday - and as it happened I also snagged a multimeter. Now see if any of this makes sense…

25 mile ride - 14ah battery mostly on throttle so basically riding it hard both ways.

When I got home the controller was saying 45v. Out with the multi and - dang if it didn’t agree exactly - 45v. What didn’t make sense is the capacity on the controller was showing 0 bars - just the battery outline. This is pretty much “shutdown imminent”. According to the manual when C12 is default - “4” it’s shut down level is at 40V - what happened to my 5V? There isn’t even a setting that would shut down at 45V. I would have doubted my memory but knowing how I forget things I took a picture before I turned it off and that’s what it's showing - empty on 45V.

I will keep on checking after rides and see if this is consistent. If it is then according to the sheet with the voltages - 45V is 39% charge left. It does kinda imply I have more range available if needed/wanted (and I can find a way to convince the controller to take it). As it goes, even using the motor a lot I can still make it in and back on one charge so I don’t “need” the range. If my maths (and the chart) is correct my absolute range should be around 40 miles with another 25 on the 9ah. I might even stop taking in the 9AH battery as it’s just dead weight … get a few more trips in to build confidence first though.

Thanks for hanging in with me on the wild and confusing ride.

 

[amberwolf]

I would love to do this BUT ... one battery is a 9ah and the other is a 14ah. Just to make it more fun the 9ah is a year or two older compared to the new 14ah. From what I found when looking into this it seemed there were issues with two identical batteries - let alone two mis-matched.

Paralleling them isn't a problem, it helps them both--the better battery "holds up" the weaker one as needed. As long as both are the same voltage range and are the same voltage when connected together.

*Seriesing* them would only allow the lowest total Ah and lowest maximum A, but in parallel they add up.

 

[marvin78]

Paralleling them isn't a problem, it helps them both--the better battery "holds up" the weaker one as needed. As long as both are the same voltage range and are the same voltage when connected together.

*Seriesing* them would only allow the lowest total Ah and lowest maximum A, but in parallel they add up.

I thought it mattered a lot that both batteries should be of equal characteristics ( kind of cells / age / load cycles -> aka same chemical behavior )? Load cycles with same voltage but different characteristics batteries would negatively affect the better battery while it could cause even faster degradation of the already weaker pack since it now has to keep up?

I thought the discharge converter acted like a buffer, monitoring discharge curves and adjusting load on the batteries.

What happens when you parallel a 1500mah with a 3500mah cell? Isn't that exactly the same with two battery packs? I'm just a bit confused.

*Seriesing* them would only allow the lowest total Ah and lowest maximum A, but in parallel they add up.

I translate that in my head as follows: series adds voltage, but not capacity and discharge current, while parallel gives capacity and discharge current but not voltage.

Question though, if I mix different discharge current rate cells how does this work exactly? 10A 10A 8A = 28A? Does it matter in which p pack the lower discharge cell would be? I'm wondering, if you know you're mixing discharge rate's, and you had the option, would you then not pick 8Ah 8Ah 8Ah = 24A discharge current? Or would the lower discharge not heat up more if trying to keep up with the higher discharge cell?

I'm pretty certain I can find the answers if I look around and I'm being lazy for asking, it just came to mind due to my confusion.

 

[amberwolf]

I thought it mattered a lot that both batteries should be of equal characteristics ( kind of cells / age / load cycles -> aka same chemical behavior )? Load cycles with same voltage but different characteristics batteries would negatively affect the better battery while it could cause even faster degradation of the already weaker pack since it now has to keep up?

I'm pretty tired right now, but that sounds backwards.

A weaker battery has higher internal resistance, so it "automatically" supplies less of the load than the better more capable one, as it's voltage will drop under higher current. Since voltages on two paralleled sources will always be identical (assuming minimal / no resistance between them), the better battery then has to supply more current than the other to keep this true.

If you were paralleling individual cells to build a pack from, then for best pack performance you should use identical cells, so that they all remain in balance between the series groups.

But for two (or more) whole battery packs in parallel, they act just like different capacity / property cells in parallel within the same parallel group--the more capable ones supply more of the current. Usually the ones with less capacity also have higher resistance (so supply less amps), so this works out so that they all "run out" at about the same time, rather than the better ones running out and leaving the weaker ones holding the bag, so to speak.


There's a number of discussions about this, with varying opinions, but there are also testing threads if you poke around for them (will probably take a lot of looking, though; there's literally at least hundreds of battery-pack-paralleling threads; I don't know a good way to find the testing-specific ones).

I thought the discharge converter acted like a buffer, monitoring discharge curves and adjusting load on the batteries.

To actually adjust the load on a battery, it would have to be a full DC-DC converter, one for each battery, that would be basically the same size or bigger than the motor controller or battery charger (same concept inside, slightly different operation).

If you're talking about a battery combiner, all most of them have is a pair of diodes connected at their cathodes, with the anodes at the battery positives, and controller positive at the cathode join.

Some use "ideal diodes" which is a specific circuit using a FET to operate as a kind of one-way switch...but same very simple operation--it's on or it's off.

It *could* use an MCU to monitor battery voltage and/or current over time to switch one battery on or off based on some set of parameters (possibly even user-programmable), but it's still just turning things on or off, not adjusting any loading on the batteries. (unless you count taking the load completely away, or completely putting the load on the battery, as adjusting ). But this is pretty complicated and costs more (and is more failure prone), so is less likely to be how they would actually be built (vs just using a couple of fat schottky diodes).

What happens when you parallel a 1500mah with a 3500mah cell? Isn't that exactly the same with two battery packs? I'm just a bit confused.

Yes. The capacities add up. So you get 5000mAh total, and each cell supplies what it has available.

I translate that in my head as follows: series adds voltage, but not capacity and discharge current, while parallel gives capacity and discharge current but not voltage.

Correct.

Question though, if I mix different discharge current rate cells how does this work exactly? 10A 10A 8A = 28A? Does it matter in which p pack the lower discharge cell would be? I'm wondering, if you know you're mixing discharge rate's, and you had the option, would you then not pick 8Ah 8Ah 8Ah = 24A discharge current? Or would the lower discharge not heat up more if trying to keep up with the higher discharge cell?

If you're building a battery out of different cells, then as long as every parallel group has the same total resistance and the same total capacity, so every group is identical to eveyr other seriesed group of parallel cells, then the only thing that matters is how you build each group.

If you need 28A discharge and you only have enough 10A discharge cells to use two per group, but you have some 8A discharge cells, then you can certainly build a pack that does that and has the 10A + 10A + 8A grouping in every parallel group of three cells. You'd be better off to build it with three 10A cells because it would give you more capability than you need, so it will still do what you want as it ages and is not as hard on the cells normally, but it would "work" the other way until the cells age enough to not be able to supply it all.

If you have the option to match all cells in all groups so they are identical properties (resistance, capacity, etc), then you should do this, and also make sure there is significant excess capacity and capability so it still works as it ages, and is less stressed before then so it ages more slowly, too.

but if you don't have that option, you mix cells as needed to make sure you still have excess capacity and capability *in total* for each set of parallelled cells (group), and that every group matches every other group as closely as possible (to keep balance problems to a minimum).

The lower disharge rate cells wont' discharge more than they can because their resistance is higher, when paralleled with higher rate cells, because their resistance is lower and so supply more of the current (that differenc ein resistance is *why* they are different rate cells).

 

[amberwolf]

*ideally* you want to always use the same stuff and the same properites for the whole thing...but when you don't ahve that availble, it will still "work" becaus of hte differneces in properties of the cells that make them what they are, also makethem able to do what they do.

if that makes any sense...I think i am too tired to word it correctly.

 

[marvin78]

if that makes any sense...I think i am too tired to word it correctly.

Don't worry I got food for thought, thank you very much

I'm a bit sad i might have bought into a bit of a hoax myself then, I knew there were different quality 'combiners' to be had, and I am aware I can buy a combiner with buck converter build in ( so you can also match two different voltages btw ).

I will open up the one I bought from pspower, now that I'm curious how it actually works ( since I bought it assuming it actually balanced ... well based on what I already said ).

 

[amberwolf]

I'm a bit sad i might have bought into a bit of a hoax myself then,

it's probably not so much a hoax more likely just sellers having no clue to what they acutally sell and no understanding how they work, and tehcnical terms bieng really hard to properlty translate without that knowledge and understanding of what's going on to pick the right words to dsecribe it.

I knew there were different quality 'combiners' to be had, and I am aware I can buy a combiner with buck converter build in ( so you can also match two different voltages btw ).

simple diodes do that, there's no need for converters. the diodes just keep the higher one from backfeeding the lower one, and the lower one just doesn't contribute any current until the higher one has run down far enough to equal the lower one.

if they do have converters then the device will be as big as your charger or controller, or larger as it would have to have switching dc-dc converters capable of the entire power output needed by the controller. they can't use a simple (linear) buck converter or they will gnerate huge amounts of waste heat and drag down the higher voltage pack as fast as it can waste that power to make it the same voltage, and give you a lot less range than you would get otherwise.

 

[marvin78]

I thought worst case a boost converter would cost you about 10% ( as in, they are capable of 90% efficiency? ) And they are much much smaller as a linear

Also, a diode is enough to equalize any differences in same rating packs, but it won't let you combine a 36v and 48v battery to drive a 48v motor. For that a boost converter needs to step up the 36v to 48v ( loosing about 10-20% of capacity in excess heat / conversion losses from that 36v battery ). Or well, again, that was my theoretical understanding anyway.

Don't understand how the boost converter on the 36v circuit would drain the 48v, it wouldn't since it doesn't interact with it at all. The converter should step up the 36v to the level of the 48v SOC ( as measured by the combined discharging unit ), this would only incur losses in the converter itself not in the 48v battery pack? So yes, you won't get the Wh from the 36v battery, but the Wh minus conversion losses, added to the Wh of your 48v battery, or am I wrong?

ed: sorry for late edit, wasn't very clear myself and to get ahead of any creeping confusion on my side

 

[amberwolf]

I am not really sure if there is a language barrier, but I don't think what you're posting matches what I am. :?

I thought worst case a boost converter would cost you about 10% ( as in, they are capable of 90% efficiency? ) And they are much much smaller as a linear

Hmm. You were asking about buck converters, which is what I was answering (not boost). But:

Boost or buck, there are different kinds of converters. SMPS is what most chargers are, and the common controller (plus the motor) make what amounts to an SMPS. How efficient they are depends on what you're converting to and from, and the design itself. They might be *capable* of high efficiency if done right in the right situation, but they aren't all going to give you that. You can test the efficiency of a system by measuring the input power and the output power, and the difference is the inefficiency's power loss as waste heat.

For the same power conversion, the linear converter is likely to be larger and less efficient, and is just a buck converter, as it is usually wasting power resistively to drop the voltage.

Either way I have not seen a combiner that does either buck or boost...they just keep two batteries paralled to a controller while preventing either from backflowing into the other (using diodes or ideal diodes made from FETs).

Some might have an MCU that monitors the batteries to shut one or the other down based on various programmed limits; I haven't seen one that actually "proves" it does this, or has a detailed manual indicating how that works in it. I've seen block diagrams posted for one that claimed this but the diagram didn't make sense even within it's own claims, so I don't think it was actually what it said it was.

If you know of ones that actually do more than just connect two packs in parallel, please post links to them; I'd like to figure out what they are actually doing inside.

Also, a diode is enough to equalize any differences in same rating packs, but it won't let you combine a 36v and 48v battery to drive a 48v motor.

Diodes (as a battery combiner) wont' equalize anything. They only prevent backflow from a higher voltage pack into a lower voltage one.

The latter part is true, in that a 36v battery, by itself can't operate a 48v *controller* (but could certainly drive a 48v motor...it would just run it at 3/4 the speed you would expect at 48v).

For that a boost converter needs to step up the 36v to 48v ( loosing about 10-20% of capacity in excess heat / conversion losses from that 36v battery ). Or well, again, that was my theoretical understanding anyway.

Yes, but the boost converter (including it's cooling and/or heatsinking, casing, etc) is going to be about the same size as the controller or the system charger, if it can support the full power that the controller is able to draw while doing the conversion. Same thing for a buck converter. And that's assuming an SMPS. A linear version is simpler but has more waste heat and ends up being large too.

Don't understand how the boost converter on the 36v circuit would drain the 48v, it wouldn't since it doesn't interact with it at all. The converter should step up the 36v to the level of the 48v SOC ( as measured by the combined discharging unit ), this would only incur losses in the converter itself not in the 48v battery pack? So yes, you won't get the Wh from the 36v battery, but the Wh minus conversion losses, added to the Wh of your 48v battery, or am I wrong?

I am not sure where any of that comes from or is in reference to, if it is in response to my post?

I looked back at my posts and there isn't anyting about a lower voltage pack draining a higher one? (except maybe that this is something you don't want to happen and hence the diodes in the battery combiners?)

I can answer the part above in detail later though. First, just some info on the concept in general:



None of these combiners that I have yet seen are meant to use something as great a voltage difference as a 36v and 48v pack together. They are really intended to just take two similar but non identical packs (like two 48v packs at different states of charge and thus different voltages) and let you use them "at the same time" without switching wires around, or unplugging one and plugging in the other, etc.

They dont' really let you use both at the same time the whole time, though, if they are just diode (or ideal-diode/FET) based. The pack at a lower voltage isn't contributing current, only the higher voltage pack. When the higher voltage pack has been used enough to be the same voltage as the lower voltage pack, then they both flow current based on their internal resistances, down to the point that one pack's BMS turns it's output off and then only the other pack (whichever that happens to be) supplies the whole load.

That behavior isn't caused by the actual diodes, ideal diodes, FETs, or other combiner properties, its' just the nature of voltage sources in a system.


To make a combiner work by actually sourcing power from both batteries at the same time, regardless of their voltages, it would require a DC-DC converter on each one, that each by itself is capable of supplying the full power of the entire system (so when one shuts down because it's battery is empty, the other doesn't blow up or shutdown from the extra load), and can convert voltage from the input across the entire range of voltages of all the packs that could possibly be used with it, to the entire votlage range that might be expected on it's output. The wider the input range, and the wider the output range, the bigger / less efficient the DC-DC is going to be. And you need one for every connectable pack, built into the device. It's not going to be a small device, and it's going to make a lot of heat, and it is going to waste a lot of battery capacity.

If you just have random batteries you have to use to get enough range, such a device could be useful and worth having...but normally you'd be using batteries that are very similar or identical in voltage range to start with, and just ahve slightly different states of charge, so simple diode paralleling devices are sufficient.