steveo
100 kW
GGoodrum said:
hey gary
will the version 4 bms be avaliable for christmas :lol:
--waits anxiously--
-steveo
Even Newer 4 to 24-cell Battery Management System (BMS)
- Thread starter GGoodrum
- Start date
hey gary
will the version 4 bms be avaliable for christmas :lol:
--waits anxiously--
-steveo
StJaux
1 mW
fechter said:
I like this design of yours - I will expand on it and feed back
StJaux
1 mW
StJaux said:
I have a question please : I assume you will be using a TC54VX21 to detect the 2.1V crossing - why the 5v1 crowbar at the input of the voltage detector?
StJ
If there is a connector to disconnect the pack from the BMS, when reconnecting, there is a chance that some cell circuits could see more than the TC54's rating (12v) for long enough to make it blow up. The zeners protect the TC54s under this condtition. Normally, they don't do anything.
StJaux
1 mW
fechter said:
I see you have been around the block a couple of times. I have taken a slightly different approach - I will post the design in the next day or 2.
StJ
StJaux
1 mW
Here is the Fechter Cascade design populated and simulated with real=world components.
I like this design - it is cheap, consumes almost no power and is LVC Active LOW, independent of the height of the battery stack, provided the overlap daisy-chain is maintained.
I will construct a prototype over the festive season, and if successful, get PCB's made - I will make the Gerbers available.
St.
Cell 01 under test
View attachment Cascade001_01_Schem.pdf
View attachment 5
Cell 02 under test
View attachment Cascade001_02_Schem.pdf
View attachment Cascade001_02_Graph.pdf
Cell 03 under test
View attachment Cascade001_03_Schem.pdf
View attachment Cascade001_03_Graph.pdf
I like this design - it is cheap, consumes almost no power and is LVC Active LOW, independent of the height of the battery stack, provided the overlap daisy-chain is maintained.
I will construct a prototype over the festive season, and if successful, get PCB's made - I will make the Gerbers available.
St.
Cell 01 under test
View attachment Cascade001_01_Schem.pdf
View attachment 5
Cell 02 under test
View attachment Cascade001_02_Schem.pdf
View attachment Cascade001_02_Graph.pdf
Cell 03 under test
View attachment Cascade001_03_Schem.pdf
View attachment Cascade001_03_Graph.pdf
Nice. I constructed a test on a breadboard using actual parts. I used 1M resistors everywhere and it seems to work fine. It still functions properly down to about 1v/cell. Below that, there won't be enough voltage to turn on the transistor and it will signal LVC.
GGoodrum
1 MW
Since I'm seeming to get quite a few connection questions of late, I thought I'd post this connection diagram I'm doing for the new LVC/HVC/Active Cutoff instructions:
It shows a typical 12s2p LiPo setup, with two 6s LVC/HVC/Parallel Adapter boards, a 48V MeanWell supply with the MW Charge Controller and the new "Smart" Active Cutoff module. The latter will still use the technique of grounding the throttle signal, for LVC trips under load, but will now cut power if it detects an LVC trip due to a slow current drain, like leaving a controller on, for instance. This cutoff is latched, until reset.
The Active Cutoff module also includes both inputs and outputs for the throttle signals, and includes the current limiting resistor.
-- Gary
It shows a typical 12s2p LiPo setup, with two 6s LVC/HVC/Parallel Adapter boards, a 48V MeanWell supply with the MW Charge Controller and the new "Smart" Active Cutoff module. The latter will still use the technique of grounding the throttle signal, for LVC trips under load, but will now cut power if it detects an LVC trip due to a slow current drain, like leaving a controller on, for instance. This cutoff is latched, until reset.
The Active Cutoff module also includes both inputs and outputs for the throttle signals, and includes the current limiting resistor.
-- Gary
Kingfish
100 MW
Gary,
I am in the midst of constructing a new battery bag set which has a completely different layout than my present ones of which are located at the rear of the bike in panniers. The new bag will be draped over the triangle area like a saddle bag. There is insufficient room for the BMS boards, however I have the center void between the bags available. My thinking is that if I put the BMS boards in a common plastic enclosure then they would be sufficiently protected.
Question: Does the balancing wire length matter greatly to the circuitry? In other words, I’d like to make as few holes in the new bag as possible, and gang together the bungles. There will be Qty-9 5S-Zippys neatly on a side, possibly one BMS board per 3 Zips. I haven't estimated the length, though possibly up to 8-inch extentions.
Thanks kindly in advance, KF
I am in the midst of constructing a new battery bag set which has a completely different layout than my present ones of which are located at the rear of the bike in panniers. The new bag will be draped over the triangle area like a saddle bag. There is insufficient room for the BMS boards, however I have the center void between the bags available. My thinking is that if I put the BMS boards in a common plastic enclosure then they would be sufficiently protected.
Question: Does the balancing wire length matter greatly to the circuitry? In other words, I’d like to make as few holes in the new bag as possible, and gang together the bungles. There will be Qty-9 5S-Zippys neatly on a side, possibly one BMS board per 3 Zips. I haven't estimated the length, though possibly up to 8-inch extentions.
Thanks kindly in advance, KF
Kingfish said:
The length itself doesn't matter much but the resistance will. 8" should be no problem though.
The balancing current will cause a voltage drop in the wire, which can throw off the voltage by a few millivolts.
The main issue we've seen in the past is the resistance of the wires on the end cells (pack + and -). If the shunts are all active (worst case), the shunt current on the 'middle cells' passes to the next cell instead of going down the tap wires. On the end cells, there is no next cell, so all the shunt current will be present on those wires. The solution for this is to make the end cell tap wires heavy enough to minimize voltage drop. For 1 foot wires, 18ga should be enough. The amount of voltage drop in the wires is typically not enough to cause any problems, but you may notice the end cells have slightly higher voltages at end of charge.
Kingfish
100 MW
Understood, thank you Richardfechter said:
For clarity and comparison, would you know what the typical wire gauge is for balancing leads? Forgive my noobnessness; it's one of those details that I seldom consider
And I wouldn't normally ask, except that I do not see this given in the spec: ZIPPY Flightmax 5000mAh 5S1P 15C
…although we can easily deduce that the battery cables are 12 AWG (not that I was terribly concerned; it can be read in the picture).
Happy Thursday! KF
GGoodrum
1 MW
The Zippy and Turnigy packs typically come with 10-gauge discharge wires, with 4mm bullets, one male and one female. The balance plug pigtails are I think 22-gauge silicon wires with JST-XH connectors. What I do on the combo LVC/HVC/parallel adapter boards is have spots for up to four matching JST-XH connectors and then output pads for larger connectors that can handle 18-gauge wires. There are pads for either the 3.5mm mini-terminal blocks, or for my new favored type, the MTA-156 series. The wire end of these type connectors are much easier to fabricate, as the wires don't need to be stripped and can be simply pushed into the connector with a $10 T-handle tool. No crimping required.
Anyway, by using 18-gauge wires from the pack you can make them as long as a couple feet, and not have issues with balancing errors.
-- Gary
Anyway, by using 18-gauge wires from the pack you can make them as long as a couple feet, and not have issues with balancing errors.
-- Gary
Gregb
100 W
Does anybody know where these are available in numbers fewer than 100???GGoodrum said:
GGoodrum
1 MW
Gregb
100 W
Thanks Gary, but that is the male. I need a couple of in line sockets and the standard packet of crimps they sell is 8000.
Greg
Greg
Hi
Gregb I am getting 100 units of 5,6 and 8 cell balance wires made up with silicone wire they might be what you are looking for the colors are limited to red and white but they are silicone 22 gauge. as for the housings that gary showed you, gary if that is your only sorce and you still want a lot of them drop me a PM and we can do a deal that will benifit both of us.
Geoff.
Gregb I am getting 100 units of 5,6 and 8 cell balance wires made up with silicone wire they might be what you are looking for the colors are limited to red and white but they are silicone 22 gauge. as for the housings that gary showed you, gary if that is your only sorce and you still want a lot of them drop me a PM and we can do a deal that will benifit both of us.
Geoff.
Gregb
100 W
A couple of 8 cells would be good if they are 2.5 mm and fit the JST plugs.. Maybe 4 for some spares.
Greg
Greg
GGoodrum
1 MW
I do have some 9-pin JST-XH pigtails, with about 9" leads. I had these made to use with the CellLogs. If you want some, send me a PM with your particulars.
-- Gary
-- Gary
trappermike
100 W
So do you Gary,or someone else,have a finished and tested BMS for a 24s LiFePO4 battery pack for sale yet?
I will be needing one in a couple of months,and frankly from what I'm hearing the units available from Asia are questionable in quality and effectiveness.I would like one of your new designs when they are ready...
Can someone PLEASE explain to me the difference between the "Capacity Transfer",and the "Bleeding Resistive" types offered. I've looked but not found the difference explained,good points and faults?
I will be needing one in a couple of months,and frankly from what I'm hearing the units available from Asia are questionable in quality and effectiveness.I would like one of your new designs when they are ready...
Can someone PLEASE explain to me the difference between the "Capacity Transfer",and the "Bleeding Resistive" types offered. I've looked but not found the difference explained,good points and faults?
Indubitably
100 W
- Joined
- Jan 9, 2010
- Messages
- 126
So, after looking things over and reading some of the posts since I last logged in, its seems like what I am going to need is the "full" bms, and I'm thinking it might be a good idea to go with the design that will have it embedded in some sort of potting substance becasue I intend to mount it on the bike itself (if I can't roll up to a power outlet and plug this thing in when I get where I'm going, its going to loose a lot of its charm). The configuration I want is going to use either 16 or 24 thundersky lifepo4 cells in series, and I'm guessing that I will want to buy one of the meanwell widgets seperately and string together however many I need to get the desired voltage.
At any rate, I suppose at this point I'm more just wondering if anyone has an update on the availability of the full bms, and or any suggestions about the flavor that would be most effective for my application.
At any rate, I suppose at this point I'm more just wondering if anyone has an update on the availability of the full bms, and or any suggestions about the flavor that would be most effective for my application.
I'm in the same boat... Currently I unplug my 2 12s packs and put them in parallel to charge, and in series to run on the bike. I'm planning to build a proper battery box though, so it's no longer going to be convenient to unhook and parallel the batteries.
I anxiously await a nice simple BMS that will do this.
trappermike: The difference between those 2 types of BMS is how the balancing occurs.
Resistive bleed ones (which this BMS is) work by simply bypassing current once the battery is at the correct voltage. When the first cells get to their max voltage it just burns off the extra energy as heat, until all of the other cells catch up. It balances only during charging.
Advantages are that it's relatively simple and doesn't rely as much on electrolytic caps (which don't last forever)
Disadvantages are that in a heavily imbalanced pack, a lot of energy could be wasted, and a lot of extra heat pumped out. Balancing doesn't continue as the pack discharges, so a weaker cell will still drain until it hits the low voltage cutoff, and shut down the pack, even if there's still power in the other cells.
Capacitive ones work by switching capacitors from cell to cell, charging the capacitors from the highest voltage cells, and supplying that energy to the lowest voltage cells. They usually run continuously.
Advantages are usually better efficiency, less heat, and full time balancing, which could give better range on a pack with a weak cell, since that cell will be continuously topped up by the stronger ones during discharge and idle.
Disadvantages are lots of relatively unreliable caps (assuming electrolytic) and just a more complicated operation (could be more prone to errors in a poor quality design),
It's a bit of a tradeoff really. A cap BMS is nice, because it should stay balanced through the entire discharge period, and should be more efficient during charging. The down side is that under some circumstances it can actually make things worse. If you have a cell that has a higher internal resistance than the others (possibly a bit weak, but still usable), the BMS will balance it properly during idle, but can cause it to imbalance under load/charge.
For example, a pack with one slightly high impedance cell:
During charge, the weak cell will receive less current than the other cells, due to its higher resistance. the voltage across the cell will climb a lot faster, not because the cell is charging faster, but because it's accepting charge slower.
Unfortunately, a capacitive BMS will see the higher voltage on that cell, and start shuttling it away to catch up the lower cells. The end result is that the cell which is already accepting the least current is having current shuttled away from it, causing it to charge even slower.
In theory, all of the cells will reach their full charge voltage around the same time, and after a short time in CV mode, the charger will shut off. (while the weak cell is still less charged than the others)
A resistive BMS, OTOH, would charge them all normally. The high impedance cell will actually end up with a higher voltage across it, (and therefore a lower voltage on the other cells). It will charge, but due to its impedance, will reach it's final voltage at a lower than ideal SOC(state of charge). At this point, the BMS just burns off any overvoltage applied to the cell, keeping it at the final per-cell charge voltage (kind of forcing it into the CV portion of the charge early) it will be held there until all of the other cells are at full voltage, then the charger will go into CV mode and terminate.
It's still not perfect, but at least in this situation it's just kept the weak cell topping up the whole time, which will bring it a lot closer to charged.
It's all pretty variable though. Slower charging would probably allow a more even charge. More time floating at the full pack voltage would allow the weak cells to catch up. (at the expense of holding your healthy cells at a high voltage longer than necessary)
I think if you're running relatively low currents for both charge and discharge, it probably doesn't matter which design you use. Capacitive will be better with uneven capacity cells.
If you're running higher currents, it probably still doesn't matter in most situations, but if there is a problem, it won't make things worse.
I anxiously await a nice simple BMS that will do this.
trappermike: The difference between those 2 types of BMS is how the balancing occurs.
Resistive bleed ones (which this BMS is) work by simply bypassing current once the battery is at the correct voltage. When the first cells get to their max voltage it just burns off the extra energy as heat, until all of the other cells catch up. It balances only during charging.
Advantages are that it's relatively simple and doesn't rely as much on electrolytic caps (which don't last forever)
Disadvantages are that in a heavily imbalanced pack, a lot of energy could be wasted, and a lot of extra heat pumped out. Balancing doesn't continue as the pack discharges, so a weaker cell will still drain until it hits the low voltage cutoff, and shut down the pack, even if there's still power in the other cells.
Capacitive ones work by switching capacitors from cell to cell, charging the capacitors from the highest voltage cells, and supplying that energy to the lowest voltage cells. They usually run continuously.
Advantages are usually better efficiency, less heat, and full time balancing, which could give better range on a pack with a weak cell, since that cell will be continuously topped up by the stronger ones during discharge and idle.
Disadvantages are lots of relatively unreliable caps (assuming electrolytic) and just a more complicated operation (could be more prone to errors in a poor quality design),
It's a bit of a tradeoff really. A cap BMS is nice, because it should stay balanced through the entire discharge period, and should be more efficient during charging. The down side is that under some circumstances it can actually make things worse. If you have a cell that has a higher internal resistance than the others (possibly a bit weak, but still usable), the BMS will balance it properly during idle, but can cause it to imbalance under load/charge.
For example, a pack with one slightly high impedance cell:
During charge, the weak cell will receive less current than the other cells, due to its higher resistance. the voltage across the cell will climb a lot faster, not because the cell is charging faster, but because it's accepting charge slower.
Unfortunately, a capacitive BMS will see the higher voltage on that cell, and start shuttling it away to catch up the lower cells. The end result is that the cell which is already accepting the least current is having current shuttled away from it, causing it to charge even slower.
In theory, all of the cells will reach their full charge voltage around the same time, and after a short time in CV mode, the charger will shut off. (while the weak cell is still less charged than the others)
A resistive BMS, OTOH, would charge them all normally. The high impedance cell will actually end up with a higher voltage across it, (and therefore a lower voltage on the other cells). It will charge, but due to its impedance, will reach it's final voltage at a lower than ideal SOC(state of charge). At this point, the BMS just burns off any overvoltage applied to the cell, keeping it at the final per-cell charge voltage (kind of forcing it into the CV portion of the charge early) it will be held there until all of the other cells are at full voltage, then the charger will go into CV mode and terminate.
It's still not perfect, but at least in this situation it's just kept the weak cell topping up the whole time, which will bring it a lot closer to charged.
It's all pretty variable though. Slower charging would probably allow a more even charge. More time floating at the full pack voltage would allow the weak cells to catch up. (at the expense of holding your healthy cells at a high voltage longer than necessary)
I think if you're running relatively low currents for both charge and discharge, it probably doesn't matter which design you use. Capacitive will be better with uneven capacity cells.
If you're running higher currents, it probably still doesn't matter in most situations, but if there is a problem, it won't make things worse.
trappermike
100 W
Thanks for the tutorial on BMS types,it will help me decide on what type I will need. One supplier I was talking to discontinued selling Capacitive type BMS's,I'm assuming because of problems. I like the simplicity of the resistor type more,to me simplicity usually means reliability in machines I buy or build. I've found a few sources of good quality chargers,I think this may also be key to good battery performance,I really don't care how much more it costs to get the best one,because I think it's good insurance for the money invested in the battery pack.
I'm interested in the Meanwell device if it will benefit me,and some Cell-Logs(3 or for for 24s battery?)if they will benefit me too,these are tools that should last me many years and several battery packs. I'm just unsure yet what exactly the Meanwell and Cell-Logs do,and how they will benefit me.
I love buying and collecting tools,I've amassed more than $30,000 in hand tools and special tools in the last 35 years...it's an addiction too.
I'm interested in the Meanwell device if it will benefit me,and some Cell-Logs(3 or for for 24s battery?)if they will benefit me too,these are tools that should last me many years and several battery packs. I'm just unsure yet what exactly the Meanwell and Cell-Logs do,and how they will benefit me.
I love buying and collecting tools,I've amassed more than $30,000 in hand tools and special tools in the last 35 years...it's an addiction too.
Meanwell is a power supply manufacturer. Gary has a little widget for sale that plugs into the supply, and messes with its feedback signal to turn it from a plain CV supply into a full CC/CV charger.
AFAIK the cellogs are just a cell voltage monitor that has outputs for HVC/LVC and can be used to handle the measurement portions of a BMS design. I'm not too sure about that though. I've never actually played with one.
AFAIK the cellogs are just a cell voltage monitor that has outputs for HVC/LVC and can be used to handle the measurement portions of a BMS design. I'm not too sure about that though. I've never actually played with one.
trappermike
100 W
I thought I read somewhere here that the Cell-logs help you to balance the cells better when charging,I'd use them if they can top up the cells better than with the charger and BMS. Otherwise I won't bother with them if there's not a real advantage over a good charger and BMS,I really aren't into baby-sitting the battery every time it's charging if I don't have to.
Similar threads
