Even Newer 4 to 24-cell Battery Management System (BMS)
- Thread starter GGoodrum
- Start date
ejonesss said:
Yes, the shunts will pull them down. If you have enough voltage to light up the cell LEDs, this means the shunts are full on and you should be able to see that the control circuit is throttling. This should be at around 500ma. When throttling, you should have around 3.68v across each active cell, however measurement with a DVM will be off a bit due to the throttling. You just want to make sure the circuits are functioning. Measure the exact voltages after connecting the batteries.
i was wondering is the channel leds just indicators or are they critical to the functioning of the board too?
because with no battery cells connected they come on but since the remaining 15 have fixed yet there may not be enough power passing through to the optos to make them turn on enough to cause the master led to turn green.
however if i short r 105 witch is a 100 ohm resistor then the opto turns on and the master led goes green so i know switching circuit works.
because with no battery cells connected they come on but since the remaining 15 have fixed yet there may not be enough power passing through to the optos to make them turn on enough to cause the master led to turn green.
however if i short r 105 witch is a 100 ohm resistor then the opto turns on and the master led goes green so i know switching circuit works.
Yes, the cell LEDs are an active part of the circuit, acting like 2v zener diodes. If they are disconnected (or installed backwards), the optocoupler will never turn on for that cell.
The resistors are supposed to get hot. They should not exceed 100C during normal operation. A heat sink is not necessary but if you want to enclose the circuit in a box, then the heat has to go somewhere. It might work to use silicone glue to attach a piece of aluminum bar to the resistors in order to pass the heat through a sealed enclosure. A small fan would work wonders for heat dissipation also.
The resistors are supposed to get hot. They should not exceed 100C during normal operation. A heat sink is not necessary but if you want to enclose the circuit in a box, then the heat has to go somewhere. It might work to use silicone glue to attach a piece of aluminum bar to the resistors in order to pass the heat through a sealed enclosure. A small fan would work wonders for heat dissipation also.
are all leds created equal?
i ordered 604-L7113ED unfortunately they do not give electrical specs.
they are the standard 5 mm in size so they will be easier to handle without the need for tweezers.
i also in the removal of the old leds i broke a trace (fortunately i can still jumper a piece of wire from the back of the board to fix it).
i ordered 604-L7113ED unfortunately they do not give electrical specs.
they are the standard 5 mm in size so they will be easier to handle without the need for tweezers.
i also in the removal of the old leds i broke a trace (fortunately i can still jumper a piece of wire from the back of the board to fix it).
fechter said:
ejonesss said:
Maybe you should just get or pay someone to build it for you. It would be a lot easier (for all of us).
Or start a new topic, you could call it "ejoness and his backwards LEDs" thread.
Better yet maybe write a book about it.
i have built kits in the past so i know how to assemble electronics.
i got it nearly done i just have to replace the leds and re test it and wire up the batteries and it is good to go.
combine with the fact that i had to refer to more than just the stuff in the instruction manual that stumped me i got nearly up and going..
i like this board because it is sort of universal meaning that it is not specific to anyone's brand of batteries.
most other maker's bms's are specific to their batteries/cells and do not balance and lvc as good
i got it nearly done i just have to replace the leds and re test it and wire up the batteries and it is good to go.
combine with the fact that i had to refer to more than just the stuff in the instruction manual that stumped me i got nearly up and going..
i like this board because it is sort of universal meaning that it is not specific to anyone's brand of batteries.
most other maker's bms's are specific to their batteries/cells and do not balance and lvc as good
velias said:
boostjuice
10 kW
velias said:
Yes, i must admit im getting sick of ejonesss' use of this thread as an online diary. How about you PM fechter so we all dont have to read through your petty problems.
I keep checking this thread for useful information but all i see is endless posts about your lack of initiative in putting together a simple and well documented circuit. Not to mention statements of the bleeding obvious.
rossasaurus
1 mW
I just wanted to thank you guys for designing and offering this BMS to the EV world.
I ordered one for my 20-cell TS 40AH pack today, and expect to order more in time(20-cell+).
I expect you'll continue to be swamped with orders too as there are quite a few folks importing 3-4Kw bikes from China with 20+ cells and no BMS's inside.
I fried a TS cell in my scoot due to no BMS, so this is a great development.
I also want to say you've got the patience of a friggin saint! enough said.
Thanks for the design, the PCB's, the instructions! you guys rock.
Ross
I ordered one for my 20-cell TS 40AH pack today, and expect to order more in time(20-cell+).
I expect you'll continue to be swamped with orders too as there are quite a few folks importing 3-4Kw bikes from China with 20+ cells and no BMS's inside.
I fried a TS cell in my scoot due to no BMS, so this is a great development.
I also want to say you've got the patience of a friggin saint! enough said.
Thanks for the design, the PCB's, the instructions! you guys rock.
Ross
michaelplogue
10 kW
fechter said:
I guess that answers the question I was going to ask - which was: can you take the LEDs and mount them via leads to a centralized panel so they can be viewed remotely. I'm going to guess that putting them on the ends of long wires could result in a degradation of the signal.
boostjuice said:
I agree, Eric, please start your own thread and detail your built however you wish, but lets try to keep the " sticky " as clean as possible... it's already way too long and cluttered.. i will try to clean it up at some point ..
( This post will self destruct in a few hours, )
Some news about the BMS v.1.5 from Gary!
Yesterday I tried to connect my MeanWell supply 48V/6,7A to my 15s6p battery. I have been set the voltage to 56V it is the max. voltage to use for charging. Because of this high voltage the "charger" pulsed from 0A untill 7,8A. So Itried to put a resistor with 0,8ohm in series with the positive out of the charger. As resistor I took about 50 meter of a copper wire with 16awg and it works! :lol: The max. current is now at 7,7A at 50V at the beginning and 56V at the end of charging. I think this would be an easy way to solve the problem of limiting the current.
Next I would take some big resistors putting them on the case over the fan of the meanwell.
How many resistors I have to put parallel, that is my question?
To destroy 400W is really much heat.
Another was very curious: Charging with Garys BMS there was one LED(Nr.1) which was on after 3/4 time of charging but then some minutes later the LED went off and went on a second time untill the end of charging, when all the other Leds came on.
Why this?
Yesterday I tried to connect my MeanWell supply 48V/6,7A to my 15s6p battery. I have been set the voltage to 56V it is the max. voltage to use for charging. Because of this high voltage the "charger" pulsed from 0A untill 7,8A. So Itried to put a resistor with 0,8ohm in series with the positive out of the charger. As resistor I took about 50 meter of a copper wire with 16awg and it works! :lol: The max. current is now at 7,7A at 50V at the beginning and 56V at the end of charging. I think this would be an easy way to solve the problem of limiting the current.
Next I would take some big resistors putting them on the case over the fan of the meanwell.
How many resistors I have to put parallel, that is my question?
To destroy 400W is really much heat.
Another was very curious: Charging with Garys BMS there was one LED(Nr.1) which was on after 3/4 time of charging but then some minutes later the LED went off and went on a second time untill the end of charging, when all the other Leds came on.
Why this?
Tom Tom
100 W
manfred59 said:
Manfred- Is there a thread and/or website about your stand on golf cart rider?
manfred59 said:
If the resistor is 1ohm and the current is 8 amps, then you would be dissipating 64 watts. (P=I^2R)
That works and it's pretty simple and cheap. It would be more elegant to have a true CC circuit so you don't need to dissipate any heat, but that would require some modding on the power supply. Doctorbass started a thread on that somewhere...
As far as the LED going on and then out, that sounds like the board is doing it's job. The first one lit may be still passing a fairly large current, but that might drop when the second one lights up, causing the first one to go out until the actual cell current tapers off, at which time it could light up again. As long as the time from the first one coming on to the time when all of them are on is not too long, it should be good. If you had a weak cell, it might light up first and stay that way for a very long time as the others catch up.
i am asking this here instead of pm since the answer may help others.
i was wondering if putting an led in parallel with the opto's led for the lvc would affect the trip point of the lvc?
it would be nice to know what cell is the culprit for causing the lvc to kick in and would help determine what one is going bad or low.
i was wondering if putting an led in parallel with the opto's led for the lvc would affect the trip point of the lvc?
it would be nice to know what cell is the culprit for causing the lvc to kick in and would help determine what one is going bad or low.
The current needed to trip the LVC is very low and probably not enough to light up a LED enough to see it. The other problem is the LVC usually trips under heavy load, then resets when it disables the controller output. It would blink so fast you would not be likely to see it. If you wanted to add such a thing, you would want to have a separate resistor for the added LED.
GGoodrum
1 MW
I've been in Bangkok all this week, on a business trip, so I haven't had much time to participate here. Thanks to Richard for answering all the questions, etc. 
Anyway, I will be home Sunday night. I'll try to get the third batch of boards out the first part of the week. They showed up on Friday, I'm told.
-- Gary
Anyway, I will be home Sunday night. I'll try to get the third batch of boards out the first part of the week. They showed up on Friday, I'm told.-- Gary
michaelplogue said:
Missed that one.
Yes, you could extend them to a remote panel no problem. The maximum current in them is around 15ma, so even skinny wires would be OK. The voltage drop in the wire will be negligible. You would have to run a pair of wires for each LED. You can't share wires.
Another idea I had was to attach optical fibers to the LED and run a fiber bundle out to a remote. No danger of sorts that way. Waterproof too. I've seen plastic optical fiber (fat stuff) that could be glued onto the LEDs somehow. Drilling a shallow hole in the end of the LED would make it easier to glue on.
michaelplogue
10 kW
fechter said:
The fiber optic trick sounds cool. After thinking about it, I was thinking that trying to fit 120 LEDs on one panel would be a bit silly. I think that if I just ran the main LED from each board, that would probably suffice. Using wire leads, would having crimp-on connectors effect the signal too much, or would it be OK?
One other question. In the picture below from your instruction manual, you've got two leads going to the battery pack. Could you do a quick sketch as to how this would work? I'm a little confused as to how this would work going between the batteries and the controller.
Another concern I have is that I recall you saying a while back that the traces on the board would only handle around 30 amps max (I think you were talking about the charging current at the time). I plan on using these boards in a high amp application on a motorcycle. Can I run the discharge leads directly from the batteries to the controller, or do I have to go through the BMS in order for the LVC to work?
Thanks!
GGoodrum
1 MW
The 30A limit is strickly for charging. The Pack + and - connections, shown in the illustration, are also just for charging. These would be connected in parallel with your main discharge + and - connections.
During discharge, the LVC circuits each have there own separate connections to the cells, and there is only miniscule amounts of current that flows through them. LVC operation has no bearing on how much discharge current is allowed, as this doesn't use active cutoff. that is controlled strickly by the controller, and I guess, by the size of your discharge wires. If an LVC circuit trips, it tells the controller to cut power.
-- Gary
During discharge, the LVC circuits each have there own separate connections to the cells, and there is only miniscule amounts of current that flows through them. LVC operation has no bearing on how much discharge current is allowed, as this doesn't use active cutoff. that is controlled strickly by the controller, and I guess, by the size of your discharge wires. If an LVC circuit trips, it tells the controller to cut power.
-- Gary
michaelplogue
10 kW
So, those 'pack' leads are connected in parallel then to the main discharge leads? And should these wires be of a gauge sufficient for the amps put out by the charger, or does it matter? Just for my own edification, how do these connections function in the overall charging scheme. Don't the leads going to the individual cells handle the charging function?
GGoodrum
1 MW
michaelplogue said:
Yes, to both questions
michaelplogue said:
The current from the charger goes through a circuit that is simply the all the cells in series, and then through the FET that is inserted in series with the negative connection from the charger. During the first part of the charge cycle, the current is maxed out, at whatever the charger can put out. During this "constand current" mode, it becomes harder for the cells to accept this much current, as it "fills", so the cell voltages start to rise at a steady rate. If you had a pack with perfectly matched cells, the they would all rise at the same rate, to the point that the collective sum total (i.e. -- the total pack voltage...) reaches the charger's CC/CV crossover point. Ideally, this will be somewhere around 3.7V x the number of cells in the pack (assuming LiFePO4 cells...), or 44.4V for a 12-cell/36V pack and 59.2V for a 16-cell/48V setup. Once the total voltage reaches that point, the charger will hold it there (the CV mode...) which causes the current to start dropping off. At this point, the cells are about 80-85% full. the current will gradually taper off, down to zero. When it gets under about 100mA, the cells are about as full as they are going to get.
That, or course, is with cells that are perfectly balanced, which is almost never the case. Cell capacities, internal resistances, temperature characteristics, etc., will vary over time, and each cell can change at a different rate. The bottomline is that you almost always end up with cells that hit the cutoff point before the others, and when that happens, the cell voltage starts to rise at a much higher rate. When this happens, it can cause the charger to "see" a pack voltage at the limit, but it might only be one or two cells that are causing the limit to be hit artificially. For these cells, the voltage can and will be a lot higher than the optimum 3.7V point, over 4V even. The cells seem to be able to take this, but it is believed that the long term effect of doing this repeatedly is that cell life will be shortened. In any case, what it also does is to cut the current down quicker than it should, and since all the current has to go through all the cells, the net effect is that many cells will end up not getting a full charge, and this difference between the thower and higher cells can drift apart more and more, over time.
Now, what the BMS does is simply to monitor the voltage of each cell, and if it detects that the cell is at the "magic" cutoff point, it turns on the shunt circuit so that some of the current will start to go through the shunt circuit for that channel instead of causing the cell voltage to start rising. As that first cell gets fuller, the shunt will have to bypass more and more current in order to keep the cell's voltage from quickly rising above the cutoff point (i.e. -- around 3.7V...). The max current the shunt circuit will pass is 1/2A, before it gets overloaded and the cell voltage starts rising again. Just before that happens, the shunt circuit sends a signal to the charger control logic to shut off the FET that is in series with the negative charger power lead. This cuts off all the charge current, which in turn causes the cell voltage to drop a bit, causing the signal from the shunt circuit to turn back off. That causes the FET to be turned back on, which restores the charge current. What happens is that this cycle repeats. This oscillation, or continuous interruption of the charge current keeps the shunt circuit right at the threshold, just before the circuit overloads. This in turn, keeps the cell voltage right at the optimum cutoff point. When these first cells get full, the shunt circuits still allow up to 1/2A to be available for the next cell in series, so that if it is talking longer to get full, that's okay. Once all the cells are in this full bypass mode, which means the cells are now all finally full, another signal is sent to the charger control logic to shut of the charge current, and keep it off, until the system is reset.
Anyway, to answer your question, the wires going from the BMS board to each of the cell junctions only need to be big enough to handle the shunt bypass current, which is 1/2A.
-- Gary
michaelplogue
10 kW
Thanks for the explaination! I think I understand now. I'm not an electrical engineer, but I like to get the basic idea as to how things work before I use them. Less chance of me screwing something up.
Thanks again!
Thanks again!
PJD
100 W
I never understood the purpose of the connections on the board for the charger positive and pack positive leads. Only the negative lead needs to go through the board and then to the negative pack terminal. The charging positive lead can simply be run directly to the positive terminal of the pack.
Am I missing something?
Am I missing something?
Similar threads
