Flying Capacitors Balancing Module

[@EVJ-World]

With good preparation this video can be shorten to ~ 5 minutes.

[@muddy11111]

They actually work very well for top balancing if you switch them on at the required voltage, easily done with an Arduino and relay. For occasional top balancing a manual switch is fine as they balance so quickly. A lot better than messing about with resistors. I would not leave them constantly on due to the high current draw, and they destroy the top balance at lower voltages. If you can run MQTT, have a look at the Shelly 1 switches to control them (flashed with Tasmota software) . They will run on anything from 12 to 60v and give you a single pair of 0v contacts.

[@solventtrapdotcom6676]

I chose these for my 400v project.

[@followthetrawler]

I cant be the only one who thought "Flying Capacitors? Fun with BigClive" :)

[@lezbriddon]

Flying Capacitor has one capacitor(bank) not a bank per cell pair(ripple balance) and 2xCell(n) mosfets to connect it to the highest voltage cell and then dump that into the lowest charged cell, so it can pull excess from cell1 and dump into cell5 directly, but you need a micro controller and other supporting cost, as you say they can be built with super capacitors for higher balancing current so faster balancing, jkbms sell one like that i have built it into my e-moped. They also need very little voltage difference between cells. inductor based ones can also be built in ripple but i've not found a flying inductor model for sale. i'm trying to design my own but i think i can save up and buy a dozen faster than i can design one....

[@MiniLuv-1984]

Excellent! Best explanation and demonstration of an active charge balancer I have seen. I have one too and was in two minds whether to connect it to my 16S pack, but you have given me that answer. Thanks Julian.

[@JulianIlett]

Thanks Benny - I think the only downside of this balancer is the relatively high self consumption. However, if your battery is charged each day from solar panels, I can't see it being a problem.

[@quandiy5164]

@@JulianIlett that can simply be solved by using the enable pads. Add a switch as you've done and turn it on only when the battery pack is in use or needs balancing 😁

[@JulianIlett]

True - switching the balancer from 'run' mode to 'standby' does reduce its current consumption - but not to zero.

[@Sparky400]

@@quandiy5164 but what if when you turn it on and the pack is more than 1v out of balance. Wouldn't that cause it to draw more than 5A and cause it to fail or worse? Also a worry when it still uses a bit of power when it's disabled.

[@quandiy5164]

@@Sparky400 yes, but you have bigger problems if your pack is way out of balance like that. A well functioning pack should not differ much even without balancing.

[@quandiy5164]

I have used these 4s,8s,16-17s and the previous ones you've shown which used inductors and I like these better as they keep the cell voltages much closer together especially the ones far apart in the series string like cell1 and cell 16 etc. The inductor types have a 30mv min difference in adjacent cells so the farther apart they get, the higher voltage difference is. As for operation, I have not tested it but according to the way things are wired, the first row of caps are in parallel with the cells. The second row of caps are first wired in parallel to their respective caps in the first row. Then the mosfets switch over to wiring the second row of caps all in parallel to the last cap. This occurs at a few khz as I heard it whine when I inadvertently overloaded one connection. Current is limited simply by the voltage difference divided by the sum of capacitor esr and mosfet on resistances. The pcb is 4 layer so it is a little tricky to trace out.

[@JulianIlett]

Interesting. There are newer versions of the flying capacitor modules which are more expensive. I wonder how they differ.

[@quandiy5164]

@@JulianIlett I have only seen pictures but never bought one as the older models are already pretty expensive and still worked well for my uses. The only difference I see are that the newer ones have more capacitors (possibly in parallel) and has a connector.

[@PioNunes]

There are LTO Active equalisers (for lithium titanate cells?) that will work at voltages between 2.0 to 2.8 volts.

[@station240]

This sort of module is the sort of thing a time lapse would be good for. Do you have a Gopro or something you could setup to do this ?

[@Sylvan_dB]

It may be more "eco friendly" in concept, but that heat and the operating current indicates losses are significant. Do those losses outweigh the "green" aspects for correctly balanced packs in good condition? It seems it might...

[@MuhammadHanif-bx4pb]

I think the heat come out from the capacitor. since those sot23 switching ic tend to have high frequency and the capacitor ESL starting to take effect.

[@james10739]

Have you seen the balancers with just one or 2 large Inductors I assume that mean it can only pull from 1 or 2 cells at a time but that's probably enough right

[@chrishartley1210]

You could have difficulty measuring the current. I was watching Offline Garage trying this earlier, the added impedance of an inline meter means you can't measure the current that way. And from what I understood from his attempts, the rapid switching between capacitors means that clamp meters simply cannot react quickly enough.

[@wesleymays1931]

> clamp meters can't react quickly enough There's clamp current probes for oscilloscopes, and those can usually work for at least a few tens of kHz

[@marbesky]

I'm using a few of these for about 2 years and they are brilliant 0.05mv balance between all cells

[@JulianIlett]

I think so too. They'll work a treat on my battery :)

[@BlackWolf42-]

I've got the Heltec one for LTO/NCM-LFP. I've always wondered how they worked as I could barely see any heat on the chips with the FLIR.

[@nicholashamblin3]

What is the easiest way to verify that a flying capacitor board is functioning? Could it be as simple that the light is on? Would the light not turn on if any part is not properly functioning?

[@gregorythomas333]

I just placed my order for 4 16s-18s @ 5A boards...they should be here by the end of the week.

[@JulianIlett]

I also believe they can be overlapped and paralleled for more current.

[@gregorythomas333]

@@JulianIlett I hope so...my systems are pretty massive and need a lot of amps to be moved around

[@RobinRastle]

J you dont mean a polyfuse is a regular current limiter as you said. PPTC devices. They are for very intermittant use and do not return to original low IR for hours - years. Repeated used wears out the polymer/carbon matrix and it goes to high IR. Its more likely those are plain low Resistors for current limiting

[@HelloNotMe9999]

Where did you get that torch in the upper left corner of you mat here? I much prefer tailcap switches that sit proud like that and wouldn't mind trying that one out...

[@raymundhofmann7661]

I don't get how it is eco friendly, as half of the charging/discharging power to balance dissipates in the parasitic resistances of all involved components, because it is basically a RC circuit where half of the power dissipates in R while the other half ends up in C, but only for adjacent cells. For non adjacent cells efficiency of the balancing power is 0.5**distance with distance 1 being the first adjacent cell. So if the main imbalance is between the first and the fourth cell the balancing efficiency is 0.5**3 or 12.5%. Looks to me like the effort isn't worth it. Putting a inductor in series to the transfer capacitors doesn't help, as when the switching frequency is uncorrelated and dissimilar to the resonance of this RLC circuit, half the power is still lost in all parasitic resistances.

[@donepearce]

Have you measured the switching rate here? I imagine it is not too fast as the capacitors need time to charge and discharge. But I'm also thinking that the abrupt current dumps must result in some noise on the total power bank voltage. Any chance you could pop a scope on it and show what this does to the quality of the supply?

[@JulianIlett]

Not yet, but using [Energy=0.5CVV] and [Energy=VIs], for 1V cell imbalance, 2200uF capacitor and 5A current, I get number of seconds = 0.0002 - that's a frequency of 5kHz. Very approximate I know. Once my system is built (and earning crypto) I'll go back and explore all the components in more depth.

[@klasop]

Could you please do a review of an inductive active balancer as well? Thank you!

[@fuba44]

Awesome video, really enjoyed it. I could not find a link to the device tho?

[@JulianIlett]

Thanks :) I've put a couple of links in the description.

[@fuba44]

@@JulianIlett Thank you, i bought the 8s version. The 4s would be more useful to me but it was out of stock. Hope that is an affiliate link :-)

[@rainzer0]

Hey have you looked into testing the AL8100 series IC for super capacitor balancing? The way it works is interesting and it looks like they can make one for lithium batteries with the same design

[@micropower8]

If the balancer is going of if the battery voltage is 7.5V, then the cell voltage is 1.875V and the four cells are damaged. 😞 But it looks like we can use this switch, replace it with a transistor or optocoupler and activate the balancer if the voltage of each cell is over 2.5V.

[@frankhobsonKJ4CDM]

When I read flying capacitor... Was expecting one to hit ceiling 😂

[@JulianIlett]

That might still happen :)

[@zaidudinzaidudin1852]

This support li-ion 18650 3s

[@phester28]

i would think the Fet drivers just dont have enough voltage and CCT can no longer operate as opposed to it being a design feature.

[@JulianIlett]

Well, there's a LTO version which I'm guessing has a lower cutoff voltage.

[@fredflintstone1]

Hmm being Green but the cct board is black?? puts me off as I do not like XR or road blockers, but this looks fun:-) how much carbon is used to produce this against standard resistor one LOL

[@JulianIlett]

Depends if they're carbon film resistors ;)

[@fredflintstone1]

@@JulianIlett surely if they are Carbon resistant that is green:-)

[@Rich-on6fe]

What a balancing circuit is practically useful for is avoiding over-volting (during charge) the one dying cell in your pack that has a reduced capacity. A cut-off lets you avoid under-volting it when discharging. When charging up, it doesn't save any significant energy doing this balancing by pumping charge out of the smaller (higher voltage) cell into the others. It's really just as effective to just load the cell a bit to avoid overcharging it: your pack is still effectively no greater in capacity than the smallest of your cells, which ever way you go about it. BUT! When discharging the pack, it comes into its own: it can keep the smaller cell up in a healthy voltage range, for longer, by stealing a little bit of charge from each of the other cells. This can result in a significantly greater output from the pack. Let's say in a 10 cell pack, if one cell is 15% lower capacity: during charging, 1.5% of the total charge energy would be wasted by a resistive balancer on the bad cell. Bugger all. But: upon discharging, rather than having to stop when you get down to 15%, (to avoid reversing the bad cell), a pumping balancer would allow you to take the whole pack down to empty. It props up the small cell as you go. So really it's about maximising pack usage, not about charging efficiency.

[@Sparky400]

@@Rich-on6fe I have a thery that something like this would also help even out cell wear over time by supporting the weeker cells like a splint. So even added to a pack. The worst cell would have less stress alowing the others to catch up. I was thinking of installing this into my ninebot One electric unicycle. It's got a 15S 1P battery. As the voltage goes down it becomes more and more unstable. I suspect it really only takes one or two weaker cells to bring down the entire pack. I'd hope something like this would help even it out especially if I let it rest for a bit to catch up.

[@Rich-on6fe]

@@Sparky400 It could well help. What I'd do if I were you is instrument the pack so you can see exactly what's happening to the cells as it is run up and down: sometimes I find I've made assumptions about this kind of thing, but once I've got hard data it tells a different story. When I was diagnosing something similar I set up an Arduino to measure all the voltages once a second and send it all out through Bluetooth to my phone. In this case it showed that the cells weren't actually the problem.

[@trevortrevortsr2]

Each cell is connected to an individual capacitor then they are decoupled and the capacitors are connected all in parallel to even out and then recoupled to the individual cell once again - the process repeats over and over on a kind of PWM slowly shuffling differences about till they all even out- inrush resistors limit excessive high currents - as cells approach parity their capacitor potential difference fall and the corresponding current transfer reduces - difference across cells is not limited to adjacent cells through adjacent cells tend to receive the lions share.

[@keithking1985]

Must watch this when I get back home. Looks bloody awesome 👌 👏

[@JulianIlett]

Hope you enjoy it!

[@amrishhirani6096]

Thanks 👍 Happy Diwali

[@JulianIlett]

Thank you too

[@hi-tech-guy-1823]

I did see one Smart Active Bidirectional Equalizer (Serial + Bluetooth) but you would just label it as overkill

[@JulianIlett]

But was it green? ;)

[@solventtrapdotcom6676]

I noticed some of these have "extra" capacitors and some don't...

[@JulianIlett]

Me too. The later revision PCBs seem to have more capacitors.

[@Sparky400]

I thought I was super on board with these but now I'm less sure. Interesting in how they work, not what I expected. Would love to see a diagram brake down. A bit worried if they have no current limiting. I guess you still want a resistive balancer then something to engage this periodically when they are not too far out of balance. I have seen a number of people have these short out and cause major issues. I thought maybe they got a bad one bit it happened hours or days after installation. After seeing this I'm wondering if they went bad with a design flaw. I'm also surprised how hot they get given they are meant to be a better option to a resistive balancer.

[@GoatZilla]

They apparently have current limiting through the polyfuse. Even when the polyfuse isn't "blown", it has a significant resistance. If you run the calcs it looks like the polyfuses must have something like 0.180 ohms normal resistance.

[@bobflyer4346]

Link to purchase Flying Capacitors balancing Module..Keep of the good video's

[@JulianIlett]

Links added to the description :)

[@bobflyer4346]

@@JulianIlett Thanks

[@ats89117]

Check out this video for capacitor based balancing: kzbin.info/www/bejne/eIPIq6ithqiUgas&ab_channel=SamBen-Yaakov

[@JulianIlett]

Seen it - good stuff )

[@keithking1985]

this is bloody interesting stuff

[@murraymadness4674]

They draw way too much current to leave them on, so 'eco friendly' I don't think so! Also, those caps will wear out pretty quickly, especially if they get hot. The ideal most efficent is to charge each cell independently using a transformer to isolate the cells, after they are topped

[@RobinRastle]

they have a switch to turn off balancing with a draw of

[@immrnoidall]

My cells stay balanced until close to full or empty. then it goes way off and shuts down because of one cell being full/empty before the rest. giving me a great lithium battery that shuts down way before min /max voltage of everything used by it. charging shut down at 27v and discharge shutdown at 23v - 24v. To make things more aggravating, I bought this "5 amp" balance to find out it don't do crap amperage at the low and high place when it goes out of balance. it helps more than the built in BMS balancer, but nothing even close to 5 amps. [china marketing mathematics] In my mind, no battery should shut down with 27 available volts when the protective limit of everything being used, is 21-22v. If one cell is low / high only that cell needs attention. when will A real BMS be manufactured. A big part of cell price is exact capacity matching to stop cell shutdown. price of 8 capacity unmatched 105 Ah , about $350 , matched double it. that is crazy. the variation of the unmatched cells is really small . but screws up everything because stupid BMS shutdown design. Now I ordered 6 more 105 Ah cells to hopefully find one slightly higher capacity, than my one slightly lower capacity cell. hopefully the remaining 4 will be close enough for a working 4s 12v pack . lol. whats the odds?

[@XFolf]

But what if...flying SUPER capacitors? When cell balancing must be instantaneous. Joking aside, it could be an interesting concept for balancing much bigger and scarier super cap banks or charging at much higher currents. ESR becomes very important very quickly, however.

[@zaprodk]

Well they misspelled NMC :D

[@Richardincancale]

4:22 Have you visited Notre Dame recently? :-)

[@JulianIlett]

My hunch is in my head, not on my back :)

[@BajanAlan]

You can tell Battgo to keep screen on I beleave!

[@JulianIlett]

If they've added that feature, I'd happily buy another one. Not possible on my unit.

[@nightshadelenar]

60Ah?? more like 6Ah, since that's more a reasonable guess. 60Ah balancers would be waaaay larger.

[@JulianIlett]

"Static current of about 12 mAh, the recommended battery capacity to 60-300AH battery applicable!" Quoted from this eBay listing: www.ebay.co.uk/itm/384385385409

[@nightshadelenar]

@@JulianIlett quite odd, i must say. I mean, maybe they do work for 60Ah batteries, though the current would need larger terminals if its more than 0.5v.

[@JulianIlett]

I think what they're trying to say is: 12mA consumption will drain a small battery quite quickly, so make sure you only put it on a big battery. It's nonsense really - if your battery is connected to a reasonable sized solar panel, that 12mA consumption is pretty irrelevant. And the current consumption in standby is only 0.1mA - so a small battery might be drained by this device, but it won't be damaged.

[@nightshadelenar]

@@JulianIlett exactly. Chinglish is wonderful to understand.

[@LhDiy]

Look at this video if you plan to measure the balancing current : kzbin.info/www/bejne/hnnKkKJpa5qSmrM

[@JulianIlett]

I watched it - it was fun :)

[@piconano]

Send it to Clive so he can reverse engineer it, and get a schematic. That would be interesting to see how this is designed.

[@user-qf6yt3id3w]

I believe he hold the title of "Sir Clive" now Clive Sinclair has died.

[@Sparky400]

@@user-qf6yt3id3w talking about a different one

[@user-qf6yt3id3w]

@@Sparky400 You're talking about Big Clive, right?

[@benbaselet2026]

Someone caring (and bored) might want to calculate how many millions of times this green thing needs to be used for it to save the amount of energy making it consumed :)

[@JulianIlett]

Not me then ;)

[@Sparky400]

Well the more accurate calculation would be comparing it to a resistive balancer. IE does this use less power than the resistive would use to achieve the same goal. Judging by the standby current I'd say no.

[@me-zz2340]

Absolute waste of money. Using such a big flying polymer caps it should balance your supercap battery in seconds, which we obviously didn't see. The working active balancer must well... keep the battery balanced.

[@arebear4797]

Wrong test equipment for the wrong battery. Useless video.

[@JulianIlett]

Glad you enjoyed it - more battery videos coming soon :)