Even Newer 4 to 24-cell Battery Management System (BMS)
- Thread starter GGoodrum
- Start date
myzter said:
I don't see why not. Normally you would never get enough solar power to make the shunts go on if you were riding. If you were going to park in the sun all day, then you could. In this case it would be best to run the PV into the charger connections on the BMS.
michaelplogue
10 kW
Since we are on the subject of charger compatibility:
I plan on making several 60V packs using BMI/PSI cells and running them in parallel (total of 60AH). Would I be able to charge them in parallel using a single charger using these BMS's? This is the charger I'm looking at....
http://elitepowersolutions.com/products/product_info.php?cPath=9&products_id=32
- Input: 110VAC single phase
- Output: 72.6V/20A DC
- Suitable for 20 cell pack (connected in series)
- Recommended Application: 20X 60Ah LFP battery pack with 3 hours of charging time
.
I plan on making several 60V packs using BMI/PSI cells and running them in parallel (total of 60AH). Would I be able to charge them in parallel using a single charger using these BMS's? This is the charger I'm looking at....
http://elitepowersolutions.com/products/product_info.php?cPath=9&products_id=32
- Input: 110VAC single phase
- Output: 72.6V/20A DC
- Suitable for 20 cell pack (connected in series)
- Recommended Application: 20X 60Ah LFP battery pack with 3 hours of charging time
.
GGoodrum
1 MW
michaelplogue said:
If you connect all the cell junctions together, so that the cells are paralleled first, and then connected in series, you can then use one BMS board, and charge them all together with the one charger. You could also keep the serial strings separate, using Schottky's to protect each string, but then you'd need one BMS board for each serial string.
-- Gary
michaelplogue
10 kW
GGoodrum said:
I was planning on using a 2p20s configuration for each pack, giving me three 60V 20AH packs, each with their own BMS. I would then connect them in parallel using schottkys as you surmised. I want to have multiple packs in case I get one bad cell that causes the brake cutoff, I can isolate that pack and continue on with the remaining two packs.
I guess my question is: Do I need a charger for each BMS, or can I connect the charge leads in parallel and use a single charger?
GGoodrum
1 MW
michaelplogue said:
That's a good plan. Although we haven't tried it, in theory you could paralel the packs/BMS units, and they should charge fine.
Here a picture of my 15s6p battery with Garys BMS on it.
Untill now I am charging with a Schulze charger http://www.schulze-elektronik-gmbh.com/isl8-d.htm
at max. 5A. After charging with this one I balance the pack with Garys BMS using the Meanwell PSU. This balancing lasts 3 - 4 min.
I have been reading the thread about "PSUs have a feedback loop to set the output voltage" from http://endless-sphere.com/forums/viewtopic.php?f=14&t=6749&start=0&st=0&sk=t&sd=a.
My Meanwell does not have this TL 494 but a ML 4800CP on board.
Which pin would be the feedback node ? Any ideas?
Untill now I am charging with a Schulze charger http://www.schulze-elektronik-gmbh.com/isl8-d.htm
at max. 5A. After charging with this one I balance the pack with Garys BMS using the Meanwell PSU. This balancing lasts 3 - 4 min.
I have been reading the thread about "PSUs have a feedback loop to set the output voltage" from http://endless-sphere.com/forums/viewtopic.php?f=14&t=6749&start=0&st=0&sk=t&sd=a.
My Meanwell does not have this TL 494 but a ML 4800CP on board.
Which pin would be the feedback node ? Any ideas?
Attachments
Nice work!
What is the box made of? How did you attach the edges?
Here's the datasheet for the ML4800: http://www.fairchildsemi.com/ds/ML/ML4800.pdf
It looks like that one lives on the hot side, so you don't want to make any direct connection to it.
Toward the end of the datasheet, there's a typical power supply schematic, which should be similar to the one you have.
You need to activate the emitter side of the opto coupler when the current reaches the desired limit. The emitter, which is just a LED, might be tied to ground or to Vcc, so you may need to measure the voltages on both sides when the supply is on to figure out where to tie in.
What is the box made of? How did you attach the edges?
Here's the datasheet for the ML4800: http://www.fairchildsemi.com/ds/ML/ML4800.pdf
It looks like that one lives on the hot side, so you don't want to make any direct connection to it.
Toward the end of the datasheet, there's a typical power supply schematic, which should be similar to the one you have.
You need to activate the emitter side of the opto coupler when the current reaches the desired limit. The emitter, which is just a LED, might be tied to ground or to Vcc, so you may need to measure the voltages on both sides when the supply is on to figure out where to tie in.
Doctorbass
100 GW
Nice work manfred59 !!! it is like the Killacycle battery pack style! 
Plexi, aluminium.. transparent box...wow !
If that could help, i have a 48V 8.3A (400W) H S E china made power supply fro ebay... very similar to the meanwell case but the internal board is different. they seems to all have the PWM control board seperated from the mainboard. On this little board there is a IC with the same number of pins that is a KA7500B that is exactly the same IC than the popular TL494. And... the Pin 1 is the same...
So i would say that many PSU use this common SMPS Controller IC that could have different name in it , but generally the same number of pins that is 16 pins.
some little wallmound psu have a basic 8 pins version and they also have this feedback pin input that affect the V out of the PSU.
Plexi, aluminium.. transparent box...wow !If that could help, i have a 48V 8.3A (400W) H S E china made power supply fro ebay... very similar to the meanwell case but the internal board is different. they seems to all have the PWM control board seperated from the mainboard. On this little board there is a IC with the same number of pins that is a KA7500B that is exactly the same IC than the popular TL494. And... the Pin 1 is the same...
So i would say that many PSU use this common SMPS Controller IC that could have different name in it , but generally the same number of pins that is 16 pins.
some little wallmound psu have a basic 8 pins version and they also have this feedback pin input that affect the V out of the PSU.
Attachments
Doctorbass
100 GW
I noted something on your serie wiring on the cells. maybe i would suggest you to connect the serie tab on both end of each row. That would help current sharing between parallel cells to be equal. Cause with the link wiring on only one end from a parrallel row to the next, i believe that the cells will not drain the same current during heavy load. The cell closer to the link between each parallel row will draw more current than the cell far from this link. and they could age differently.
It is very important to preserve a great low internal resistance for every parallel row. When paralleled cells have different current draw from one to another, their internal resistance will evoluate differently as the cell become aged. and this current imbalance will make your cell to heat differently. The cell closer to the serie links will heat up more than the cell farter.
Here is a suggestion you could do to correct that.
Also, the balancer could work faster with that.
:wink:
It is very important to preserve a great low internal resistance for every parallel row. When paralleled cells have different current draw from one to another, their internal resistance will evoluate differently as the cell become aged. and this current imbalance will make your cell to heat differently. The cell closer to the serie links will heat up more than the cell farter.
Here is a suggestion you could do to correct that.
Also, the balancer could work faster with that.
:wink:
jeffkay
100 W
- Joined
- Sep 14, 2007
- Messages
- 115
I tested my Elite charger and Gary's original 16 channel BMS. The charger cuts out--meaning thinks the pack is full before the BMS can finish working. Here's why: The charger in this case for 16 cells, is putting out a final voltage of 58.2v. The reason for this is that Elite, in their wisdom, is "under-charging " the Thunder Sky Lifepo4's.
Why? Because in a motorcycle or something not pulling huge amps, the cells will stay relatively well balanced and have long life.
So what can be done? First, the charger can be opened and examined to see if there is a master pot for the voltage--turning it up to 60ish volts would be ideal, again for this size pack... If that is not doable, you could adjust the BMS shunt turn-ons to go on earlier. This would require some different resistor values per channel I believe.
The Elite charger seems to be a very good value in general. Using two 110v lines, the Elite and a large variac, I charged at over 30 amps! I then drove the electric Beetle for about 28 EV miles--or is that smiles?
Jeff K.
Why? Because in a motorcycle or something not pulling huge amps, the cells will stay relatively well balanced and have long life.
So what can be done? First, the charger can be opened and examined to see if there is a master pot for the voltage--turning it up to 60ish volts would be ideal, again for this size pack... If that is not doable, you could adjust the BMS shunt turn-ons to go on earlier. This would require some different resistor values per channel I believe.
The Elite charger seems to be a very good value in general. Using two 110v lines, the Elite and a large variac, I charged at over 30 amps! I then drove the electric Beetle for about 28 EV miles--or is that smiles?
Jeff K.
jeffkay said:
Yes, you'd have to change a resistor on each cell circuit to change the voltage set point. I don't think you'd really want to go this route.
59.5v or more should be enough to fully balance 16 cells. It should be possible to increase the charger voltage, but I don't know the insides of that particular model. It would probably work great for 15 cells as is.
Tiberius
10 kW
Hi Manfred, Docbass
That 400W PSU looks very similar inside to my 350W one. Except for the Tochter-board with the IC on. My guess is that the 400W version is a modification of the 350W design.
As fechter says that IC is probably on the input side, so you should not try to connect to it. You will have to reverse engineer the circuit. The best place to start is the trimpot for adjusting the output voltage. If you can trace out some of the circuit around that it then we may be able to see how the voltage control feedback circuit works.
Or there may be someone on here who already knows about that PSU.
Nick
That 400W PSU looks very similar inside to my 350W one. Except for the Tochter-board with the IC on. My guess is that the 400W version is a modification of the 350W design.
As fechter says that IC is probably on the input side, so you should not try to connect to it. You will have to reverse engineer the circuit. The best place to start is the trimpot for adjusting the output voltage. If you can trace out some of the circuit around that it then we may be able to see how the voltage control feedback circuit works.
Or there may be someone on here who already knows about that PSU.
Nick
GGoodrum
1 MW
fechter said:
Instead of changing resistors, one thing you could do is tack on a "piggy-back" resistor to the 75.0k ones on the board now, of say 500k or maybe 1M. This will lower the shunt turn-on point a bit. Anyway, tacking on a second resistor is a lot easier than replacing them.
Here's a picture of the Ver 2.1 board with 4 channels installed for validation testing.
It seems like everything worked this time. :wink:
The thing with the red wires is a transistor switch to allow the charging connector to work with 3 pins instead of 4. The 3rd wire goes through a jumper to ground, which turns on the control circuit. Not only does this allow a smaller connector, but an accidental connection to the control line won't damage anything. We plan to incorporate this feature in the next version board. It is easy enough to "dead bug" one onto earlier versions of the board like I did.
The switch unit was later insulated with heat shrink tubing and the wires were zip tied together to make it more rugged.
The 2.1 board works just like the previous version unitil all the shunts are on, at which time the SCR gets triggered and shuts off the charging current. The batteries will be fully charged at this point. The SCR resets when the charger is disconnected.
Here's a closeup of the switch. It's a 2N3906 and a 20k resistor.
It seems like everything worked this time. :wink:
The thing with the red wires is a transistor switch to allow the charging connector to work with 3 pins instead of 4. The 3rd wire goes through a jumper to ground, which turns on the control circuit. Not only does this allow a smaller connector, but an accidental connection to the control line won't damage anything. We plan to incorporate this feature in the next version board. It is easy enough to "dead bug" one onto earlier versions of the board like I did.
The switch unit was later insulated with heat shrink tubing and the wires were zip tied together to make it more rugged.
The 2.1 board works just like the previous version unitil all the shunts are on, at which time the SCR gets triggered and shuts off the charging current. The batteries will be fully charged at this point. The SCR resets when the charger is disconnected.
Here's a closeup of the switch. It's a 2N3906 and a 20k resistor.
GGoodrum
1 MW
I'm placing the order for boards this morning. It is going to take about a week. That means early next week for availability.
In this version, I've added the "3-pin" jumper change, which Richard describes above, but the board is otherwise identical to the 2.1 version Richard just tested. Here's the layout and the schematic:
In this version, I've added the "3-pin" jumper change, which Richard describes above, but the board is otherwise identical to the 2.1 version Richard just tested. Here's the layout and the schematic:
michaelplogue
10 kW
Yay! Put me down for five!
methods
1 GW
He can have 5 so long as I get 2 first. . .
Can I get mine on a pink PCB?
-methods
Can I get mine on a pink PCB?
-methods
rkosiorek
100 kW
so just a question when you are using it for a 4 cell pack do you still use the LM78L12 regulator?
Would you set the Charger voltage slightly higher to ensure that the regulator does not drop out? or do you substitute a 78L10? would there be a problem with either of these approaches?
rick
Would you set the Charger voltage slightly higher to ensure that the regulator does not drop out? or do you substitute a 78L10? would there be a problem with either of these approaches?
rick
Ypedal
100 TW
Oh man.. i'm so screwed.... that's alot of parts in a small space... i sure hope it's easier than it looks to assemble... :lol:
GGoodrum
1 MW
Ypedal said:
Oh man.. i'm so screwed.... that's alot of parts in a small space... i sure hope it's easier than it looks to assemble... :lol:
That picture is not full-size, so it isn't that bad.
Ypedal
100 TW
*Puts on courage boots ... :lol:
Step 1.. i need a new iron :
http://cgi.ebay.ca/936-12-Hakko-Solder-Station-ESD-Safe-Factory-Fresh-New_W0QQitemZ380067473238QQcmdZViewItem?hash=item380067473238&_trkparms=72%3A1215%7C39%3A1%7C66%3A2%7C65%3A12%7C240%3A1318&_trksid=p3286.c0.m14
Anyone recommend something better ? Or does this baby do the job just fine ?
Step 1.. i need a new iron :
http://cgi.ebay.ca/936-12-Hakko-Solder-Station-ESD-Safe-Factory-Fresh-New_W0QQitemZ380067473238QQcmdZViewItem?hash=item380067473238&_trkparms=72%3A1215%7C39%3A1%7C66%3A2%7C65%3A12%7C240%3A1318&_trksid=p3286.c0.m14
Anyone recommend something better ? Or does this baby do the job just fine ?
GGoodrum
1 MW
Ypedal said:
That should be fine. I use a 25W model from Radio Shack that looks like about the same size.
rkosiorek
100 kW
Hakko makes some nice irons. that station comes with one of 3 possible soldering pencils. they don't specify which one is included. they come in "small" which is really only good for the SMT stuff, "Medium" for all round and "Large" which really tranfers a little too much heat for the SMT stuff.
i'd recommend that you get the "medium" size one if you have a choice and a couple of different size tips.
rick
i'd recommend that you get the "medium" size one if you have a choice and a couple of different size tips.
rick
rkosiorek said:
Yes, you should still use the regulator as the control circuit is referenced to the charger, not the pack negative. During throttling or when the charge is complete, you would have the full charger voltage going in. If the regulator starts to drop out, it shouldn't be a problem unless it gets below 10v. The regulator is nice to protect against accidental shorts etc. I suppose you could get away without it as long as the charger never goes above 20v peak.
Similar threads
