tag:blogger.com,1999:blog-6095764826828192506.post775329467063875535..comments2023-12-29T01:44:52.114+08:00Comments on Afdhal Atiff Tan: Low Cost Active Battery BalancerUnknownnoreply@blogger.comBlogger66125tag:blogger.com,1999:blog-6095764826828192506.post-65740212056068080652021-08-10T04:55:14.810+08:002021-08-10T04:55:14.810+08:00@9thtale
Yes, connecting it 24/7 while charging i...@9thtale<br /><br />Yes, connecting it 24/7 while charging is ok. Don't use it without a charger as it will drain both cells.<br /><br />The current is not adjustable. The maximum that the chip can handle reliably is around 1A. Any resistance in the wiring will drop the balancing current very significantly.<br /><br />This is because it relies on the voltage difference between the top and the bottom cell.<br /><br />For example, if the difference between cells is 100mV and the resistance of the wire is 1Ω, the balancing current will be 100mA.<br /><br />I do realise that this is a huge drawback. I would suggest you look at other solution.<br /><br />No problem, you asked good questions.Afdhal Atiff Tanhttps://www.blogger.com/profile/18048104720991410262noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-30389487338682838642021-08-10T03:24:11.982+08:002021-08-10T03:24:11.982+08:00And is it okay to keep this circuit attached to th...And is it okay to keep this circuit attached to the battery 24/7 and while its charging?<br />also how much can the transfer current between the cells be? is it adjustable?\<br /><br />Sorry if I'm asking stupid questions, I'm new to battery balancing circuits, till now I've always worked with single cells.9thtalehttps://www.blogger.com/profile/05206863491517513710noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-20035579358950983962021-08-10T03:21:16.603+08:002021-08-10T03:21:16.603+08:00This comment has been removed by the author.9thtalehttps://www.blogger.com/profile/05206863491517513710noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-36096120621605077422021-08-10T03:04:38.528+08:002021-08-10T03:04:38.528+08:00@9thtale
You'll need a synchronous buck so th...@9thtale<br /><br />You'll need a synchronous buck so that the energy can be transferred between 2 cells. Using an asynchronous one will only allow transfer from top to bottom cell.<br /><br />In that config, the bottom should transfer its energy to the top.<br /><br />Afdhal Atiff Tanhttps://www.blogger.com/profile/18048104720991410262noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-30841270763661746372021-08-10T02:40:02.597+08:002021-08-10T02:40:02.597+08:00I was wondering if is it possible to use nonsynchr...I was wondering if is it possible to use nonsynchronous buck converters?<br />Also in 2s config what happens if the bottom cell is fully charged and the top one isn't?9thtalehttps://www.blogger.com/profile/05206863491517513710noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-87453221702461880182020-11-28T06:44:45.536+08:002020-11-28T06:44:45.536+08:00@BobEmerson12
Hi Bob,
May I have a more detailed ...@BobEmerson12<br /><br />Hi Bob,<br />May I have a more detailed spec for the kind of modular balancer that you have in mind?Afdhal Atiff Tanhttps://www.blogger.com/profile/18048104720991410262noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-68846410107096810232020-11-27T06:44:23.125+08:002020-11-27T06:44:23.125+08:00No matter which approach is best, I'm still in...No matter which approach is best, I'm still interested in building or buying a modular cell balancer. Also a modular cell-based low and high voltage disconnect. <br /><br />Whether it's 1s or 4s, modules that could be daisy chained.<br /><br />Good luck with your studies--you're learning a lot with this project.BobEmerson12https://www.blogger.com/profile/13387329291281270410noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-66065019817261462332020-11-15T03:15:31.811+08:002020-11-15T03:15:31.811+08:00@Vanarian
Hi Chris,
Upon further inspection, I d...@Vanarian<br /><br />Hi Chris,<br /><br />Upon further inspection, I decided that the balancer is not suitable for ebike use, mainly because of the high quiescent-current consumed by the voltage reference.<br /><br />In other words, I can only make it either precise or energy-efficient.<br /><br />In precise mode, the balancer will use a voltage reference and maintain the balance in terms of mV (e.g. all cells are balanced within 10mV).<br /><br />In energy-efficient mode, the balancer will use a resistor divider and maintain the balance in terms of % (e.g. all cells are balanced within 1%).<br /><br />I think the former is more suitable for ebike applications.<br /><br />Any suggestions?<br /><br />Kind regards,<br />AfdhalAfdhal Atiff Tanhttps://www.blogger.com/profile/18048104720991410262noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-43423427026992888732020-11-15T02:35:50.668+08:002020-11-15T02:35:50.668+08:00Hi Afdhal,
Sorry for late reply (I don't know ...Hi Afdhal,<br />Sorry for late reply (I don't know why I never get notifications from your posts) ! Nice progress on your project, do you plan on making any blog post about it or for now it remains in schematic state ? Do you think it can be tested as is ? <br /><br />Best<br />Chris/Vanarian :)<br />Vanariannoreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-24871402605230918762020-09-26T22:37:09.660+08:002020-09-26T22:37:09.660+08:00@Unknown
Hi Maarten,
Sorry for the late reply, I a...@Unknown<br />Hi Maarten,<br />Sorry for the late reply, I am not sure how somehow missed your last message.<br /><br />The maximum current practically I believe is about 500mA, if you need something higher, I would suggest you to use BTS7960 or equivalent. But, I don't think this topology is suitable for that purpose though.<br /><br />I am not familiar with the DXF conversion. The circuit is fairly simple, it should be easy for you to rewire it in EasyEDA.<br /><br />Kind regards,<br />Afdhal<br /><br />Afdhal Atiff Tanhttps://www.blogger.com/profile/18048104720991410262noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-39104217564817353462020-09-25T21:15:24.914+08:002020-09-25T21:15:24.914+08:00Hello Afdhal,
Any idea how i can import this proj...Hello Afdhal,<br /><br />Any idea how i can import this project in easyEDA <br />to i can generate the gerber files and BOM list and order pre-assembled prints?<br />I downloaded the project from GitHub and opened with KiCad but cannot export to DXF.<br /><br />Thanks in advance!<br /><br />Maarten<br />Anonymoushttps://www.blogger.com/profile/10772009653969368329noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-24761282592535617282020-09-11T21:02:31.917+08:002020-09-11T21:02:31.917+08:00Hi,
Indeed there are many low-cost bms but these ...Hi,<br /><br />Indeed there are many low-cost bms but these are not modular <br />and if you want an interface they are not cheap anymore.<br /><br />I want to use this for a diy home battery of 80p 7s li-on batteries.<br />Any idea what the maximum ballance current is for this system?<br /><br />Anonymoushttps://www.blogger.com/profile/10772009653969368329noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-62730548834123578982020-09-11T19:21:57.637+08:002020-09-11T19:21:57.637+08:00@Unknown
Hi Maarten,
While it is possible, I don...@Unknown<br /><br />Hi Maarten,<br /><br />While it is possible, I don't plan to include it because of the added cost (MCU, isolation, etc.).<br /><br />There are many low-cost "bms" in the market that handle over/under voltage/current protection.<br /><br />Kind regards,<br />AfdhalAfdhal Atiff Tanhttps://www.blogger.com/profile/18048104720991410262noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-34416292584022835952020-09-11T19:02:16.332+08:002020-09-11T19:02:16.332+08:00Hello Afdhal,
Nice project you are working on, i ...Hello Afdhal,<br /><br />Nice project you are working on, i was looking exactly for something like this.<br />Like the modular design.<br /><br />I was wondering if it would be possible to add an i2c interface to monitor the singe cell voltages and statuses (getting feeded or feeding) to read in with for example nodeMCU, raspberryPI, arduino...<br /><br />This makes it possible to shut down the charging or discharging of the pack at the lowest/ highhest cell voltage with relays.<br /><br /><br />Greets<br /><br />Maarten (Belgium)Anonymoushttps://www.blogger.com/profile/10772009653969368329noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-66925622893583450472020-08-24T03:45:48.432+08:002020-08-24T03:45:48.432+08:00@Proboscide99
Hi Alessandro,
I see your point abo...@Proboscide99<br />Hi Alessandro,<br /><br />I see your point about the voltage clamp. Maybe it is not critical where it is clamped.<br /><br />The BTS7960 could be used in any way (NXP or not). <br />I haven't thought much about it yet.<br /><br />500mV at about 125kHz looks very odd.<br />I will check mine later once I got my pcb done.<br /><br />The V1 circuit forces the output to be at 1/2 Vbatt.<br />The V2 circuit chooses the output between 1/3 and 2/3 Vbatt.<br />So, V2 should give a higher balancing current.<br /><br />But, if the Vfb is already saturated, then, you're right,<br />the balancing current is indeed unbalanced.<br /><br />While driving the inductor may be trivial,<br />measuring voltages on a multi-cell pack is not.<br /><br />Do you have any (low-cost) suggestions?<br /><br />Kind regards,<br />AfdhalAfdhal Atiff Tanhttps://www.blogger.com/profile/18048104720991410262noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-11325450280784875602020-08-24T01:56:57.232+08:002020-08-24T01:56:57.232+08:00I thought the sallenkey filter didn't need the...I thought the sallenkey filter didn't need the clamped voltage, only the MP2307 does, so I connected the first to the output of comparator as it was in your original schematic, probably it does not care so much :-)<br /><br />I see the BTS7960 is a dual power mos P+N which seems to be ideal for the NXP solution, you mean to use it as a power stage for the converter or are you talking about the NXP way?<br /><br />I connected a scope to the Vfb pin, I can see a very large sinusoidal ripple (0,5V !), here is the photo (500mV / 10uS per division). The 0V is the center of the screen (in fact the ripple reaches the 0,9V of the regulator).<br />https://www.dropbox.com/s/nc2b5x3rgq5zbos/2020-08-23%2019.21.25.jpg?dl=0<br /><br />Indeed I think it's due to routing.<br />May you check your circuit to see if Vfb is stable?<br /><br />I believed the asymmetry between buck and boost was a limit of the converter that can't be compensated by driving the feedback, is that wrong?<br /><br />I think in the NXP solution the voltage is sampled by the AD in the microcontroller, which then drives the MOS gates. The T-on duration should be less than the saturation time of the coil, so I think the goal is turning the MOS off when the maximum possible energy has been stored in the coil but before it saturates. Then, the T-off should be long enough for this energy to flow in the lower battery.<br /><br />Probably the microcontroller should be programmed in such a way that the amount of energy transferred at every cycle varies with the battery voltage difference, anyway never more than 50% duty (which corresponds to 100% of energy transfer, I think).<br /><br />Ciao!<br />Alessandro<br />Proboscide99https://www.blogger.com/profile/13133115093144427655noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-55879626754637515082020-08-23T23:57:20.130+08:002020-08-23T23:57:20.130+08:00@Proboscide99
Hi Alessandro,
I am glad to know th...@Proboscide99<br />Hi Alessandro,<br /><br />I am glad to know that you have replicated my result.<br /><br />May I know why you moved the voltage clamp?<br />I think your idea of clamping it early in the signal chain is a good idea.<br />The sallenkey low pass filter is meant to produce a fixed 0.925V (or whatever the reference used by the IC) during runtime.<br />The LPF can be replaced with a fixed 0.925V reference if so desired.<br /><br />If a higher current is needed, I can recommend BTS7960 or equivalent (an external driver will be needed).<br /><br />Do you have an oscilloscope to capture the instability?<br /><br />Idling at 50mA sounds a bit too high for an MP2307.<br />I am in the process of designing a PCB, but this might take 3-4 weekends as I have very little time for personal projects.<br /><br />I recommend 1% tolerance for all resistors. <br />You will need to adjust the reference resistor if you use TL431 to the desired voltage difference.<br /><br />The V2 opamp is not just for enable control,<br />it also will increase the balancing current (U2.3's output is connected to R8),<br />i.e. the window comparator should solve the unbalanced current problem (source/sink).<br /><br />I've seen the AN4428 before. I shy away from it because of the potential cost.<br />Unless all pwm channel is driven at 50%, Fig.2 in AN4428 doesn't describe how the cell voltage is sampled.<br />It does mention the use of analogue multiplexer though.<br /><br />Any thoughts?<br /><br />Kind regards,<br />AfdhalAfdhal Atiff Tanhttps://www.blogger.com/profile/18048104720991410262noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-47330178400123948692020-08-23T18:40:10.089+08:002020-08-23T18:40:10.089+08:00Hello Afdhal, I received the MP2307 boards. I buil...Hello Afdhal, I received the MP2307 boards. I built a mix of previous schematic (using MP2307, which is more powerful) and new schematic (with the additional op amp for turning off the converter when not needed) and it works.<br /><br />I only moved the Vfb clamp away from the sallenkey filter so that the filter sees the entire voltage swing while the MP2307's feedback pin receives the clamped signal (schematic):<br />https://www.dropbox.com/s/vmhb2qa9fxkpipd/schema_prototipo_assemblato.jpeg?dl=0<br />(the manually drawn op-amp is from the V2 schematic, just for reference of pins because I used two LM358 instead of the single i.c.)<br /><br />The balancing current (tipically in the range of 350-400mA) is a little low for the size of my batteries (180Ah). I believe this may be due to overheat protection. I noticed some instability of the MP2307 caused by the unusual wiring of the feedback input, that probably needs to be closely coupled with the output voltage. The MP2307 draws about 50mA instead of 15mA whith the feedback connected normally (for voltage regulation), but this may be due to poor layout of my prototype, maybe that a well designed pcb solves this small issue.<br /><br />I also noticed that the balance current is a little higher when the circuit works as a step-down (upper battery to bottom battery) than when it works in boost mode (bottom battery to upper battery).<br /><br />The window comparator (added in V2) for enable control is a little critical on resistor's tolerance, it can happen that the circuit is turned off when there is a little difference to balance and vice versa.<br /><br />Anyway, the concept of bidirectional conversion works.<br /><br />I'm now wondering if go ahead with this solution (with larger converter) or finding other ways to obtain higher currents.<br /><br />For example, I came across an NXP application note (AN4428) which describes an inductor-based solution: one of two mosfet is driven with a PWM. Depending on which mosfet is driven, the current flows from B1 to B2 or vice versa. The amount of energy is only limited by the size of the inductor and the mosfets (and the operating frequency, of course, but I'd prefer to keep this low to reduce losses). The drives should avoid saturating the inductor by limiting the T-ON time to the necessary.<br /><br />See figure 2 of this document:<br />https://www.dropbox.com/s/4w9z1k0c1i2txts/idea_half_bridge-NXP_battery_balancer_AN4428.pdf?dl=0<br /><br />What do you think about it?<br /><br />Ciao!<br />AlessandroProboscide99https://www.blogger.com/profile/13133115093144427655noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-55546716134932942222020-07-29T23:55:00.139+08:002020-07-29T23:55:00.139+08:00Thankyou for the schematic and for the proboscide&...Thankyou for the schematic and for the proboscide's led :)Proboscide99https://www.blogger.com/profile/13133115093144427655noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-73226209810873524322020-07-27T08:59:04.414+08:002020-07-27T08:59:04.414+08:00Hi @Proboscide99 Alessandro, @Vanarian Chris, and ...Hi @Proboscide99 Alessandro, @Vanarian Chris, and all,<br /><br />Sorry for the slow progress.<br /><br />I have uploaded the V2 draft here: https://github.com/AfdhalAtiffTan/LowCostActiveBatteryBalancer/blob/master/v2_draft_untested.pdf<br /><br />Any feedback will be appreciated.Afdhal Atiff Tanhttps://www.blogger.com/profile/18048104720991410262noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-9912774366251340192020-07-26T19:33:47.088+08:002020-07-26T19:33:47.088+08:00Hello Afdhal, while I wait for the MP2307 boards t...Hello Afdhal, while I wait for the MP2307 boards to arrive, I may start building che circuits. I read that you made some improvements over the initial schematic, may you share them before I start? Doesn't matter of course if you only have sketches and manual drawings.<br /><br />Thanks in any case<br />Alessandro<br />Proboscide99https://www.blogger.com/profile/13133115093144427655noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-87098010604083456072020-07-17T22:34:10.197+08:002020-07-17T22:34:10.197+08:00Hi Afdhal,
Unfortunately we don't live in the ...Hi Afdhal,<br />Unfortunately we don't live in the same country (I'm from Italy) but we can still do something. UK (and Spain) was the closest place where I found the converters.<br /><br />The batteries are for a stand alone small solar plant (750Wp panels, four 6V 180Ah batteries), mainly used for a backup 24V line at home (lighting) and a low power 230Vac line (from a 2kW true sinewave inverter).<br /><br />The batteries are five years old and unfortunately one (n.3) is about to fail, this is why I'm approaching the world of balancers (a little late, you may say) :-)<br /><br />Next pack I'll install them from the beginning.<br /><br />Since the battery n.3 is not completely gone and the other three are operating (almost) normally, I decided to have some fun by trying to postpone the replacement of the pack.<br /><br />Hence, I bought the 24 --> 5V 20A isolated dc-dc converter (the output can be trimmed between 4,5 and 5,5V). It takes power from the whole pack (24V) and supplies a 5,25V to the n.3 (failing) battery. This way, the battery is capable of sustaining the load even when prematurely discharged. The converter prevent the voltage from falling below 5,25V and this avoids further damage to the battery.<br /><br />Then, the opposite problem arises at charging time: the three "good" batteries have far more capacity to restore than n.3, which reaches the final voltage quickly. For that problem I'm building a large passive shunt which is capable to bypass the whole solar charger current (abount 20 Amperes) so the other batteries can keep charging. This assures that the voltate on the failing battery will not rise above 7,15V (which is the shunt threshold).<br /><br />Of course an active balancer whould be better, but currents are a bit too high.<br /><br />When available, I think I will add the MP2307 balancers to take some of the passive dissipation off, but I can't remove the passive shunt from battery n.3 until I replace the pack.<br /><br />Thanks for the credits! :-)<br /><br />Kind Regards<br />Alessandro<br />Proboscide99https://www.blogger.com/profile/13133115093144427655noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-35928607033289367052020-07-17T03:40:42.573+08:002020-07-17T03:40:42.573+08:00@Proboscide99
Hi Alessandro,
I have been caught ...@Proboscide99<br /><br />Hi Alessandro,<br /><br />I have been caught by Murphy's Law in the past as well. 🤣<br /><br />I am in the UK too (Worcester), so maybe we can collaborate on a project one day.<br /><br />I used to use a shunt on my battery before, but the wasted heat generated is too much for me to handle. <br /><br />Are you using the battery on an ebike? The isolated DC/DC converter idea is nice, but a bit overkill in my opinion.<br /><br />I see what you mean with the Vfb, I have to say that I am embarrassed to missed the fact that it can be saturated when one of the cells is off-balance.<br /><br />Thank you for sharing your thoughts, I will add it in the future revision and credit you for it. :)<br /><br />Kind regards,<br />Afdhal<br /><br /><br /><br />Afdhal Atiff Tanhttps://www.blogger.com/profile/18048104720991410262noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-91964682982828074652020-07-16T20:01:17.507+08:002020-07-16T20:01:17.507+08:00Hi again, about the Vfb zener protection, I agree ...Hi again, about the Vfb zener protection, I agree that if the converter operates on normal loads (no batteries), then the MP2307 reacts immediately to Vfb adjustments (cancelling the output voltage error) so the operational amplifier (LM358) output always stays around the 0,925V reference value for the Vfb signal.<br /><br />But if you have a battery, which is very slow, the MP2307 will slowly move to the balance condition. Before that, the LM358 will keep seeing the error and probably the output (sent to the Vfb pin) will be near 0V (in one case) or near +VBATT (in the opposite).<br /><br />I suppose that the near-vbatt condition on Vfb occurs when the gnd-side battery has higher voltage than the other one.<br /><br />In this case, the Vfb will be higher than 0.925V (because we are asking the MP2307 to lower the voltage on VCOM).<br /><br />Since the VCOM voltage cannot change quickly (because we have to move energy between batteries, which will take some time) the error persists long enough for the LM358 output to reach the upper voltage (near the +vbatt supply) because it keeps seeing the error until the balance is done.<br /><br />If this makes sense to you, you may try to check what happens on the Vfb pin during balancing (I can't because I don't have the hardware yet).<br /><br />Ciao!<br />Alessandro<br />Proboscide99https://www.blogger.com/profile/13133115093144427655noreply@blogger.comtag:blogger.com,1999:blog-6095764826828192506.post-71275248463149572732020-07-16T16:46:33.388+08:002020-07-16T16:46:33.388+08:00Hello, I ordered four MP2307 converters (one spare...Hello, I ordered four MP2307 converters (one spare part for murphy law) and expect to receive them from UK in two weeks worst case.<br /><br />It was the fastest possible path as they would take one month from china.<br /><br />In the meantime, since I have a nearly failed battery in the serie, I'm developing a super-shunt that allows recharging the pack even if the failed battery reaches prematurely the final voltage.<br /><br />During discharge I connect an isolated meanwell SD-100B-5 stepdown power supply with 24Vin, 5Vout 20A for the purpose of transferring energy from the working batteries to the weak one :-)<br /><br />I'll keep you posted.<br /><br />Ciao!<br />Alessandro<br />Proboscide99https://www.blogger.com/profile/13133115093144427655noreply@blogger.com