steveo said:
You're crazy man. That's a lot of volts
Yes just the 6 wires for the charge and LVC control bus lines. If you're using the latest version, the voltage regulator should be fine.
Even Newer 4 to 24-cell Battery Management System (BMS)
- Thread starter GGoodrum
- Start date
You're crazy man. That's a lot of volts
Yes just the 6 wires for the charge and LVC control bus lines. If you're using the latest version, the voltage regulator should be fine.
steveo
100 kW
dnmun said:
To complicated .. I will build it so i could link the additional boards.. I will be using 3 psu linked together to get 150v charge.
-steveo
methods
1 GW
I agree.
Having done high voltage for a while I learned that constantly switching back and forth between parallel and series invites all kinds of problems.
Not only do the connectors wear out but every time you take it apart disaster is eminent.
These are very high energy density packs and I like mine to go together and stay together.
Even having different voltage chargers around is trouble if the connectors are compatible.
I hooked up my 100V charger to my 50V lipo a couple weeks back
You want to talk about spectacular fire!!!!!
Luckly I had my lipo monitors running and they started beeping like hell when it crossed 4.22V per cell.
Charging in series allows one to pull up, plug in, and forget it.
This Fall I was in heaven. Charging 2 or 3 times a day with no hassle.
I had a 30' cord running outside and I would come home at lunch and just plug in.
No fooling around.
It was awsome.
-methods
Having done high voltage for a while I learned that constantly switching back and forth between parallel and series invites all kinds of problems.
Not only do the connectors wear out but every time you take it apart disaster is eminent.
These are very high energy density packs and I like mine to go together and stay together.
Even having different voltage chargers around is trouble if the connectors are compatible.
I hooked up my 100V charger to my 50V lipo a couple weeks back
You want to talk about spectacular fire!!!!!
Luckly I had my lipo monitors running and they started beeping like hell when it crossed 4.22V per cell.
Charging in series allows one to pull up, plug in, and forget it.
This Fall I was in heaven. Charging 2 or 3 times a day with no hassle.
I had a 30' cord running outside and I would come home at lunch and just plug in.
No fooling around.
It was awsome.
-methods
yes you do for 2 reasons.
1. with the charger control circuit connected it does put a few milli amp load on the first 4 cells and over a few days (for a 1p pack of a123 cells)
2. to reset the charging circuit status (if you discharge with led lit i think it will not charge.
after the battery charges the main led should go green after a while if not there may be something wrong like a solder bridge or a reversed opto.
1. with the charger control circuit connected it does put a few milli amp load on the first 4 cells and over a few days (for a 1p pack of a123 cells)
2. to reset the charging circuit status (if you discharge with led lit i think it will not charge.
after the battery charges the main led should go green after a while if not there may be something wrong like a solder bridge or a reversed opto.
Patriot said:
Patriot
10 kW
- Joined
- Aug 7, 2008
- Messages
- 511
Oh heavens,
I left mine attached all night, and the main LED was on all night.
Will that damage anything?
I am finishing my charge today, and the final cells are now reaching there full charge/balance. I don't think the cells discharged that much over night with the circuit energized. They're the big 38120S Headway cells, and they did charge for a couple hours before I went to bed.
I left mine attached all night, and the main LED was on all night.
Will that damage anything?
I am finishing my charge today, and the final cells are now reaching there full charge/balance. I don't think the cells discharged that much over night with the circuit energized. They're the big 38120S Headway cells, and they did charge for a couple hours before I went to bed.
Patriot said:
No, that won't damage anything. It will just take a bit longer for the pack to balance on the next charge cycle. In 24 hours, it will drain about .5Ahr from the first 4 cells. If you left it on for several days it could completely discharge the cells to the damage point.
If the charger is connected, it will supply the current and the batteries won't be draining.
methods
1 GW
recommended to set it up so that it is impossible for the negative lead to be connected without the charger being connected.
Better yet make it so that the charger needs to be connected *and* on.
On mine I bypassed the 12V circuitry that leaches the first 4 cells and drove that 12V signal from the charger.
-methods.
Better yet make it so that the charger needs to be connected *and* on.
On mine I bypassed the 12V circuitry that leaches the first 4 cells and drove that 12V signal from the charger.
-methods.
GGoodrum
1 MW
This is why there is a 2nd negative wire that needs to be jumpered. If you don't have the charger plugged in, and on, it can't power the charger control section of the circuit.
-- Gary
-- Gary
Patriot
10 kW
- Joined
- Aug 7, 2008
- Messages
- 511
I think I just fried my BMS. Something went POP when I was testing it.
$120.00 down the drain, and no ebike.
I should never have done this project. I am good at soldering and running wires, etc, but not building circuits. It looks great, but I have neither the knowledge nor skill to repair it, when something goes wrong.
I have no one to blame, but myself.
I'll keep trying to fix it, but I have no idea what I'm doing.
$120.00 down the drain, and no ebike.
I should never have done this project. I am good at soldering and running wires, etc, but not building circuits. It looks great, but I have neither the knowledge nor skill to repair it, when something goes wrong.
I have no one to blame, but myself.
I'll keep trying to fix it, but I have no idea what I'm doing.
dnmun
1 PW
pictures? so sorry, but this is how you are gonna learn.
my worst was a term project for my junior level EE design class. had to build an autoranging DVM from discrete parts, not even MSI level parts. starting with circuit, then SPICE, then build using wirewrap terminals. my entire christmas holiday. everyday, until like 2-3am, for almost 2 weeks, unwrapping, then rewrapping something like 8 times to make it work. the guy still failed me since it was late. but they still took me into grad school in EE, and of course my first job as a grad student was grading papers from circuits class, which i had skipped myself. lucked out and got another job playing with excimer lasers, karma.
did it pop? is there a burn mark somewhere? what happened immediately before it failed and was there another problem it had before that that you were fixing. somebody will help, jonesy is becoming pretty good at helping too.
my worst was a term project for my junior level EE design class. had to build an autoranging DVM from discrete parts, not even MSI level parts. starting with circuit, then SPICE, then build using wirewrap terminals. my entire christmas holiday. everyday, until like 2-3am, for almost 2 weeks, unwrapping, then rewrapping something like 8 times to make it work. the guy still failed me since it was late. but they still took me into grad school in EE, and of course my first job as a grad student was grading papers from circuits class, which i had skipped myself. lucked out and got another job playing with excimer lasers, karma.
did it pop? is there a burn mark somewhere? what happened immediately before it failed and was there another problem it had before that that you were fixing. somebody will help, jonesy is becoming pretty good at helping too.
I'd like to request that we keep the discussion on Patriot's BMS to his build thread. It's kind of duplicated here and gets confusing to figure out where to answer.
flashedarling
100 mW
- Joined
- Jul 21, 2008
- Messages
- 49
So I'm still really torn on what would be the ideal charger to match with this BMS. If I order a Zivan I want to make sure I'm using the best possible combination.
Would an constant current constant voltage curve or IUa curve of say 15Amps with a voltage of 87v (24 Cells) be the best? Or would it be better to have a profile with 15Amps to 85v then 1Amp to 87v for balancing.
I see it this way, if some cells are out of balance some shunts will start to activate around 84-85v so as soon as that happens the current will be forced to reduce to 500mA. I don't know what happens if the charger is still in constant current mode, will it try to feed the cells 15A still because it hasn't reached cutoff voltage or will it backoff and allow it to charge at the lower rate? The exact time the shunts turn on is a little random since every pack may go out of balance differently.
I'm just excited to start testing my 24 Cell BMS (Stupid lack of 72v chargers) and want to make sure I don't buy some $500 charger that will break either itself or the BMS.
Would an constant current constant voltage curve or IUa curve of say 15Amps with a voltage of 87v (24 Cells) be the best? Or would it be better to have a profile with 15Amps to 85v then 1Amp to 87v for balancing.
I see it this way, if some cells are out of balance some shunts will start to activate around 84-85v so as soon as that happens the current will be forced to reduce to 500mA. I don't know what happens if the charger is still in constant current mode, will it try to feed the cells 15A still because it hasn't reached cutoff voltage or will it backoff and allow it to charge at the lower rate? The exact time the shunts turn on is a little random since every pack may go out of balance differently.
I'm just excited to start testing my 24 Cell BMS (Stupid lack of 72v chargers) and want to make sure I don't buy some $500 charger that will break either itself or the BMS.
I think a standard CC,CV supply will be fine as long as it behaves under heavy load. I don't think you need the taper toward the high end, but it wouldn't hurt either. When the BMS throttles, it doesn't really matter how much current the supply makes. It will force it into sort of a CV mode by throttling.
A Zivan should be good as long as you get it programmed with the LiFePO4 profile. The older ones don't have the right software.
There are probably cheaper things that would work, but I can't recommend anything specific.
I know there have been some issues with switching power supplies that go into a 'hiccup mode' under high load. Most bench power supplies with adjustable current limit work very nicely.
A Zivan should be good as long as you get it programmed with the LiFePO4 profile. The older ones don't have the right software.
There are probably cheaper things that would work, but I can't recommend anything specific.
I know there have been some issues with switching power supplies that go into a 'hiccup mode' under high load. Most bench power supplies with adjustable current limit work very nicely.
CGameProgrammer
10 kW
This was probably answered somewhere in the previous one thousand posts in this thread but damn if I'm gonna read every single one! If two PCBs are purchased, can a 4-cell slave section be joined with the first PCB for a 28s setup, or would you need the two masters, thus basically having two BMSs?
If two PCBs are purchased, can a 4-cell slave section be joined with the first PCB for a 28s setup, or would you need the two masters, thus basically having two BMSs?
GGoodrum
1 MW
Yes, that would work fine. You just need to connect the six bus wires with small gauge wires. With the latest version, the 12V regulator section is good for more than 100V, so 28-30s LiFePO4 should be fine.
-- Gary
-- Gary
CGameProgrammer
10 kW
Oh, does the cell count not have to be a multiple of 4? Do the instructions explain how to wire a slave for only two cells?
GGoodrum
1 MW
CGameProgrammer said:
No, it doesn't have to be in multiples of four at all. I just needed to pick some number to group into "tearoff" sections, and four seemed reasonable. You can use this with any number between 4 and 24 channels. If you wanted to do 22 channels, you just don't populate the last two with parts. If you wanted to do 30 channels, for instance, you could simply break off an 8-channel section from another board, and just populate six of those 8 channels, and then connect the six bus lines from the second card to the main board.
-- Gary
Just a notice to those building boards, it seems Mouser is fresh out of 180k resistors.
I found some at Newark:
http://www.newark.com/multicomp/mf12-180k/resistors-0-125w-1-180k-resistors/dp/38K5434
Over 7,000 in stock right now.
I found some at Newark:
http://www.newark.com/multicomp/mf12-180k/resistors-0-125w-1-180k-resistors/dp/38K5434
Over 7,000 in stock right now.
If this has been mentioned previously and I missed it, I apologize but are there any plans in the near future for a turnkey version of this BMS? If so, a LiFePO4 and a LiPo would be great. If no LiPo, maybe at least a LiFePO4 version that can be easily modified for LiPo.
Something I don't quit understand regarding the BMS and packs built of series and parallel cells. If I have a 7s2p pack, does this BMS need a channel connected to every single cell?
Thanks for everyone's great work including those that have shared their experience with their construction. Its been a wealth of information but I'm trying to digest it all so I understand it correctly.
Something I don't quit understand regarding the BMS and packs built of series and parallel cells. If I have a 7s2p pack, does this BMS need a channel connected to every single cell?
Thanks for everyone's great work including those that have shared their experience with their construction. Its been a wealth of information but I'm trying to digest it all so I understand it correctly.
CGameProgrammer
10 kW
No, you connect it to each of the seven 2p pairs. The only thing that matters is the number in series.
Similar threads
