Even Newer 4 to 24-cell Battery Management System (BMS)

[GGoodrum]

Sorry, Steveo. I missed this. The big difference is that there are replacements for the big electrolytic caps. The new ones are small MLCC versions. You don't HAVE to replace them, though.

In the charger control section, there are three new resistors (two 10ks and one 20k...) that replace the 4.7k R6 in v2.3. Finally, all the 820 ohm resistors have been replaced by 1ks, but again, this doesn't have to be done, if you already have the 820s.

-- Gary

 

[AndyH]

Hey Gary

Just a quick question, Has the BOM on mouser Changed for the 2.5version? .. if so what are the extra parts that i would not need to order specifically .. I just want the first channel spare parts ..as i burn out a channel .. when one of my battey packs disconnected. .. i have version 2.3

-steveo

Here you are. This is the BOM for a 4-channel V2.3 board. Andy

//// 4-ch BOM ////
647-UVR1H100MDD1TD|6
581-SA105E104MAR|1
512-1N914T50A|10
78-1N4744A|1
78-TLUV5300|1
859-LTL1CHKFKNN|4
512-KSP94TA|2
511-STP160N75F3|1
512-KSP44TF|1
512-KSA931YTA|4
299-1M-RC|1
299-100K-RC|1
299-10K-RC|4
299-4.7K-RC|1
299-820-RC|11
299-100-RC|9
270-75K-RC|4
270-180K-RC|4
286-6.8-RC|4
299-20K-RC|4
610-2N5060|1
512-LM78L12ACZXA|1
747-IXDF404PI|1
579-TC54VC2102EZB|4
512-FAN431LZXA|4
782-ILD2|4
//// last item ////

 

[Patriot]

I have more strange BMS issues. :?


As I charged up my battery today from a partial discharge, it reached 59.1v while charging, and half of the cells lit up, then the main LED started flickering red/green/orange... or whatever color you call it. Of course, when this happens, the charger starts to slow down its charge because of its auto shutoff. It stays in this flickering mode for hours, and half the leds lit, with the resistors warm. :?


How the heck do I overcome this? The charger voltage is setup above 60v.

 

[recumbent]

Patriot,

Have you checked your voltage readings on the cells that tripped the orange leds? Because my BMS does similar operating problems as yours. Don't let your charger keep flikering the LED's because it will overheat.

My Soneil charger blew the 2 large power transistors and one resistor as part of my learning curve, just saying so those of you that cannot afford to pay the extra $60 for repair on top of the $200 for the Soneil charger, DON'T BUY THIS BRAND of charger!

If you only use a 2amp charger, go 3amp, huge difference.

 

[Patriot]

I think I may need to pull the battery apart, and check all connections. After flickering for about three hours, I decided to squeeze the battery, and all of a sudden, the main LED went red and started charging again. I think I have a loose connection somewhere. That seems odd, since I really torqued everything good. I'm not so sure I can even overload this BMS, because my 2amp charger isn't all that powerful. Even when the channels go into bypass the big resistors don't even get all that hot. They get very warm, but not screaming hot. I don't mind the long charge times. It's so small, and takes about 10 hours to fully charge my 20ah battery.

 

[steveo]

Hey everyone,

thank for specifying the bom file & parts that have changed specifically! ..

Patriot,

Have you checked your voltage readings on the cells that tripped the orange leds? Because my BMS does similar operating problems as yours. Don't let your charger keep flikering the LED's because it will overheat.

My Soneil charger blew the 2 large power transistors and one resistor as part of my learning curve, just saying so those of you that cannot afford to pay the extra $60 for repair on top of the $200 for the Soneil charger, DON'T BUY THIS BRAND of charger!

If you only use a 2amp charger, go 3amp, huge difference.

To be honest .. this bms works well for me! .. but the issue when i charge is i can never get the green light to trip and terminate the charge!! .. I've seen the light go completly green however the bms still attempts to bleed the cell .. and i'll eventually see the light turn orange then green .. its like it never will end ..

I have seen the bms work perfectly once when i first assembled is using a 24s pack .. it charged light lighted up green and charge terminated ... PERFET !!! .. this only has happend 1-2 time in all the time i've charged my pack since the begining of this year .. ever since i reconfigured my pack to 10p & 24s .. I find the bms has never triped to green and shut down like it did before...

I have a second 2.3v bms i assembled .. which i will try out tonight to see if it will work properly ... i have my charger set at 88.9 v to charge 24s of a123 ..

wish me luck..

-steveo

 

[Patriot]

I have a 16s pack, and I haven't changed the configuration since I installed it.

What I'm wondering is what would cause the LED to flicker like that, while leaving some of channels unbalanced?

It never did that before. I noticed, a couple of the channel LED's also flicker once in a while.

Does anyone know what would cause this?

 

[Icewrench]

Been using two of these bms`s for a few months in a 12 s 4 p set up. Using a 2 amp charger all works good.
I wonder how much of that 2 amps is eaten up by the shunt resistors when say 10 channels are lit up?

Patriot try stopping the charge for a few minutes. let the lit leds fade out, then charge it some more. Repeat...

Steveo how many amps will your charger put out?

 

[Patriot]

^^^ I tried that. It didn't seem to do anything.

What did work was squeezing and shifting the battery around. This tells me I may have a bad connection somewhere, causing it to stop charging a series of the cells.

My first bike had no suspension, and really got the daylights beat out of it, by bouncing around so much on rough roads. I wonder if something could have gotten knocked loose, which may have caused what is going on.

Other than that, I would have no idea as to how to proceed with fixing it.

 

[steveo]

Hey everyone,

thank for specifying the bom file & parts that have changed specifically! ..

Patriot,

Have you checked your voltage readings on the cells that tripped the orange leds? Because my BMS does similar operating problems as yours. Don't let your charger keep flikering the LED's because it will overheat.

My Soneil charger blew the 2 large power transistors and one resistor as part of my learning curve, just saying so those of you that cannot afford to pay the extra $60 for repair on top of the $200 for the Soneil charger, DON'T BUY THIS BRAND of charger!

If you only use a 2amp charger, go 3amp, huge difference.

To be honest .. this bms works well for me! .. but the issue when i charge is i can never get the green light to trip and terminate the charge!! .. I've seen the light go completly green however the bms still attempts to bleed the cell .. and i'll eventually see the light turn orange then green .. its like it never will end ..

I have seen the bms work perfectly once when i first assembled is using a 24s pack .. it charged light lighted up green and charge terminated ... PERFET !!! .. this only has happend 1-2 time in all the time i've charged my pack since the begining of this year .. ever since i reconfigured my pack to 10p & 24s .. I find the bms has never triped to green and shut down like it did before...

I have a second 2.3v bms i assembled .. which i will try out tonight to see if it will work properly ... i have my charger set at 88.9 v to charge 24s of a123 ..

wish me luck..

-steveo

I got my second built bms to give the the green light and shutdown after charging!!!! we have great success!!

-steveo

 

[Patriot]

Weird!!!


I rode my ebike to work today, and promptyl started charging when I arrived. After 6hrs, the balancing started. The last twelve channel LED's came on all at once, very quickly. Then, over the next 20mins the last four came on, and the main LED went green as expected. So, the battery charged and balanced as it should.

Odd how it worked fine today, but was giving me all hell yesturday.


CRAZY!!!

 

[Nate]

As you all may know I was having a lot of trouble a few pages back. With the help of Mr. Flechter my BMS is up and running. Turns out I increased the forward voltage on the LEDs by installing them backwards.
Now on to my next problem, I can only get it to latch after manually balancing the cells at the end of the charge. Well, I can let the charger sit there and "cook" in what seems to be a trickle charge state, but I don't want to fry another charger. I have a Soneil 4806 SRF. This charger may just be too smart for this BMS... It senses the end of the charge and the charger goes green before the BMS finishes. It turns off the internal fan, makes a buzzing sound, and starts to get hot. I can unplug it and let the lights go out, over and over, until the balance is close enough to latch the BMS, but of course I want this to work as it should.
So now after all that rambling, is there a way to modify this charger to work? I have maxed out the voltage and current pots... 62.5v and 3.5a.

Thanks everyone.
Nate

 

[ejonesss]

fechter and/or ggoodrum.

how much power does the charge control draw when the second black wire is disconnected?

because i noticed that the 4 cells on the minus end of the pack loses their surface charge more than the others do.

after fully charging the pack then i unplug the charger and later when i plug in the charger the 4 cells on the minus end of the pack are still charging while the rest are fully charged and lit up their leds.

i know it is probably normal for the cells on the minus end of a pack to go lower than the rest but should stay the same even when settling from their surface charge.

 

[fechter]

fechter and/or ggoodrum.

how much power does the charge control draw when the second black wire is disconnected?

because i noticed that the 4 cells on the minus end of the pack loses their surface charge more than the others do.

after fully charging the pack then i unplug the charger and later when i plug in the charger the 4 cells on the minus end of the pack are still charging while the rest are fully charged and lit up their leds.

When the switch wire is disconnected, there should be near zero drain from the control circuit.

When you start charging, the control circuit takes about 20ma (most of this goes to lighting up the red LED). This is why the first 4 cells take longer to light up. They may not lower to start with, but get 20ma less charging current than the others so take slightly longer to get up to the set point.

At the end of charge, when the LED is green, there may be some drain on the first 4, but if the charger is still supplying current, it should supply the control circuit through the shunts in the other cells and not put any drain on the first 4. If the charger cuts off by itself, then there could be a bit of drain until you disconnect the charger.

 

[fechter]

Well, I can let the charger sit there and "cook" in what seems to be a trickle charge state, but I don't want to fry another charger. I have a Soneil 4806 SRF. This charger may just be too smart for this BMS... It senses the end of the charge and the charger goes green before the BMS finishes. It turns off the internal fan, makes a buzzing sound, and starts to get hot. I can unplug it and let the lights go out, over and over, until the balance is close enough to latch the BMS, but of course I want this to work as it should.
So now after all that rambling, is there a way to modify this charger to work? I have maxed out the voltage and current pots... 62.5v and 3.5a.

Thanks everyone.
Nate

Yes, it seems like if you had about 1-2 more volts it would behave the way you want.

I'm not sure about the Soneil. If the adjusting pot has a fixed resistor connected to the high side, you might be able to change the fixed resistor to get a bit more range. I don't know the exact layout in that one.

If the Soniel is kicking out just before the last cell lights up, it's probably balanced well enough that you don't really have to worry about it.

 

[manfred59]

The problem with my BMS is back again.
On page 76 I reported from this problem. After connecting the half discharged 16s/6p battery with the main + and - powerpoles the first(near to the control unit) orange led and the last one (16th led) are lit. I changed the U102 (LM431 2.5V linear regulator) and the Q101 (KSA931 PNP transistor) of both channels and it worked again. I could do about 20 chargings without a problem, but today again these four transistor were damaged, I changed them again, but after the next try of connecting the battery with the main connector at 58V I think all 4 transistors are probably damaged again because the first and the last led are lit when I do the connections .
Is there a special order of connecting the battery to the charging unit?
I did it like this:
1. Connecting the multiple connectors 20 pin
2. connecting the main charging wires
3. plugging the charger to 230V

 

[fechter]

Yes, we should try to figure out why this is happening.

You say it fails when making the charger connection?

Do you disconnect the BMS from the battery when not charging?

I'm trying to understand exactly what you are doing when it blows.

Most people leave the BMS permanently connected to the cells and only disconnect the charger, but that should not really matter.

What kind of connector are you using for the tap wires?

 

[manfred59]

Hello Fechter!

You say it fails when making the charger connection?

Yes exactly!

Do you disconnect the BMS from the battery when not charging?

Yes, I am disconnecting the BMS from the battery after charging.

What kind of connector are you using for the tap wires?

I usee the Molex Mini-Fit connectors 20pin two rows (9A).
Perhaps another reason: the charger I use is a Switching Power Supply 12V and 48V in series at 8A. As in an earlier post you told me to mount a cap at the output. Perhaps I should use a much bigger cap like I have now (1000uF/63V)!

 

[rslter]

Hello. I am new to the forem and to this site. I recently build the 2.2 BMS module, for someone who is building I bike. I am not building it. I am mearly doing electronics work for him. So I assembled the unit(Via soldering), and now I am in stages of actually testing it, to see if I can charge, and monitor the batteries correctly useing this devise. It my be premature to actually ask for advise at this moment especially since I have not read every inch of the forem posts. And I am sorry if I have interupted the flow of other members also putting up posts. For that I apologise. Anyways, I really just want to test the unit to see if it works. I have enabled the board to add/remove jumper wires without soldering useing pin sockets soldered onto the board as so. So I can set it up for a 4,12,20 A123batteries. Right now I only want to charge 4 of them. The outline instructions that came with the BMS monitor is very vague. For example: on page 6 "If the controllers brake inhibit line and ground signal, are connected to the optocoupters output, it will pull down this control line, causing the controller to cut the throttle, which in turn, removes the load from the pack." I am like "what!????". So this kind of turn me off from actually doing the LVC test of the circuit board, which is what I wanted to do first. But let me get back to the 4 battery test. I was actually able to get 3 to light up for quite some time. And I made sure all the LED's, and everything else were soldered correctly. I am useing 15ohms 2W 5% metal film resistors Digikey--PPC15W-2TRND for shunt resistors. I don't have a very good power supply. It is an old Harrison laboratories 6204A DC power supply 0-18V(600mA),18-40V(300mA). I first set the Voltage to 4X3.6V=14.4V, so I set it to 14.8V. I use a 3A Diode and connect the Anode to the positive output, and Cathode to the +pack on the PC board. The 4th batterys +terminal is also connected to the Cathode. I do this so I dont get any feedback voltage going in and ruining the power supply because I am useing it as a charger. I feel this is logical. For hours nothing happens, But I notice that the anolog current meter on the power supply is down to zero, when before it was all the way to 600mA. So I increased the voltage slightly, and instantly one of the LEDs come on, then later two more pop up, and I am like "cool". The resistors get tremendously hot to the touch, but its not melting plastic, or solder. So I put a fan small fan over the unit. Ithe power supply is now dishing 16V, which I don't know if this is the right thing to do or not, but I want to see if I can get the other LED to come on, and the main one next to the gate driver to turn green. Then at least Id know if the board worked. Then something strange happens. The light next to the driver(that has always been red of course) is now changing colors back and forth periodically from red---orange---green and then back to red again over, and over. The needle on the power supplies current monitor also changes with the color change moving from left to right, and then back again. It goes from max--zero when it turns green. I thought it was supposed to turn green after all the leds on each channel are on. One of the LEDs in the middle is still off, and I checked the polarity of LED, its correct. So I turn power supply off. The batteries seem unevenly charged, so I leave the unit on with just batteries pluged in. Now each battery is close to 3.4. One of them is 3.8V. Anyways I hope that I have given enough information. If anybody can help me with my progress, and like I said before; I know its a little early, but some help would be much appreciated. So I hope I am not being rude. Thank you very much for your time.

PS--I wanted to upload an image, but the site gave me some problemsView attachment Batt monitor1.jpg

 

[fechter]

Hello Fechter!

You say it fails when making the charger connection?

Yes exactly!

Do you disconnect the BMS from the battery when not charging?

Yes, I am disconnecting the BMS from the battery after charging.

What kind of connector are you using for the tap wires?

I usee the Molex Mini-Fit connectors 20pin two rows (9A).
Perhaps another reason: the charger I use is a Switching Power Supply 12V and 48V in series at 8A. As in an earlier post you told me to mount a cap at the output. Perhaps I should use a much bigger cap like I have now (1000uF/63V)!

Because you are blowing the ones on the ends, it makes me think there may be a bad connection between the pack and the main + or - charger wires coming from the BMS.

The wires from the BMS to the pack must be connected first before connecting the charger to the BMS. This may be part of the problem too. If the power supply is connected to the BMS board (even if it has no power) the batteries will try to charge the capacitors in the charger when you connect the board to the batteries. Depending on which pins make contact first, you could easily fry cell circuits.

If the BMS board is connected to the cells by itself, there is no inrush of current. After that, you can connect to a charger and the high currents will be passing through the main connections and main FET not through any cell circuits.

If you have any pins left over on your connector, it would be good to have redundant connections on the main pack + and - wires. A loose or bad connection on the main pack wires might allow destructive currents to pass into the first cell circuit.

I don't think adding capacitance will help much (based on testing by others). Using longer wires from the power supply to the BMS board might help by adding a little resistance.
I'm working on a modification to the charge control circuit that will slow down the oscillation and might make certain switching mode power supplies 'happier' during throttling.

 

[fechter]

Now each battery is close to 3.4. One of them is 3.8V. Anyways I hope that I have given enough information. If anybody can help me with my progress, and like I said before; I know its a little early, but some help would be much appreciated. So I hope I am not being rude. Thank you very much for your time.

PS--I wanted to upload an image, but the site gave me some problemsView attachment 1

Not rude. Ask away. That's why we're here.

Your power supply looks perfect for testing. They don't make them like that anymore. 8)

I suggest initial testing with no cells attached. Just hook up the main + and - wires and slowly crank up the voltage until you see the cell LEDs just barely light up. At this point, you can measure the 'cell' voltages. Just be careful not to crank the voltage way past where you expect the shunts to come on. You should see about 100ma or so at this point. The cell voltages should be around 3.7v with that version. Measure each cell connection.

Increasing the voltage a bit more (and it is very sudden), the LEDs will go fully bright and the status LED may start to go green, indicating throttling. Once the circuit starts throttling, voltage measurments will tend to be inaccurate due to the waveform.

If the cell voltages look OK with no cells attached and all the cell LEDs light up, then connect the actual cells.

The first time I ran my pack, there was one low cell that took like a whole day to come up to the set point. Even then, it has a tendency to drop off faster than the others. This is a partially damaged cell. Eventually it got up to the point where the auto shutoff worked.

 

[rslter]

Something terrible happened. I was turning it as slow as possible untill the orange LEDs flashed, and then abruptly went off. But the light did turn green and stay there. I hope at the worst, I just have to replace the LEDS. Did I hurt any Trasistors?

 

[fechter]

Something terrible happened. I was turning it as slow as possible untill the orange LEDs flashed, and then abruptly went off. But the light did turn green and stay there. I hope at the worst, I just have to replace the LEDS. Did I hurt any Trasistors?

Ahh... that's exactly what's supposed to happen if the automatic cutoff is working

If you disconnect the charger, then reconnect it, the auto shutoff resets (and you can repeat the experiment).

To avoid tripping the auto shutoff, put a jumper across any cell (and use less voltage). If any cell does not light up, it won't tigger the auto shutoff. After measuring the voltages, you can move the jumper to a different cell to test the one that was jumpered first.

 

[rslter]

Thank you for all your help. I see whats going on. I overreacted. I actually successfully balanced a 4-battery pack, and the light turns green. So it was very successful. I was also wondering if there is some way to discharge the batteries without ruining them. Can I do this by leaving the batteries connected to the BMS, disconecting the powersupply, and then connecting a motor in series with the pack? I want to try an build a powersupply/charger cheaply from scratch and see if it will work with the BMS. I know that they are allot of them for sale online, but I want to do this anyways. Thanks.

 

[ejonesss]

in a future version it would be nice if everything related to charging (charging, and lighting up the leds ) was powered from the charger.

fechter and/or ggoodrum.

how much power does the charge control draw when the second black wire is disconnected?

because i noticed that the 4 cells on the minus end of the pack loses their surface charge more than the others do.

after fully charging the pack then i unplug the charger and later when i plug in the charger the 4 cells on the minus end of the pack are still charging while the rest are fully charged and lit up their leds.

When the switch wire is disconnected, there should be near zero drain from the control circuit.

When you start charging, the control circuit takes about 20ma (most of this goes to lighting up the red LED). This is why the first 4 cells take longer to light up. They may not lower to start with, but get 20ma less charging current than the others so take slightly longer to get up to the set point.

At the end of charge, when the LED is green, there may be some drain on the first 4, but if the charger is still supplying current, it should supply the control circuit through the shunts in the other cells and not put any drain on the first 4. If the charger cuts off by itself, then there could be a bit of drain until you disconnect the charger.