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

[Jay64]

That is GREAT news! I will definately be buying some of those boards then. Thank you to Gary and Richard and everyone else that gave input for this project. It is a great service to the EV community. And yeah, pics for us noobs is really good. :lol:

 

[steveo]

Hey everyone

Could anyone please advise what the max amps I can charge with this bms? Is there a detailed spec on all the feature of this bms somewhere?

thanks
-steveo

 

[methods]

I charge at 103V and 15A

I think I heard Garry say 30A at one point. . . Better check on that.
The charge power just passes through a fet and other than that it is a straight shot.

I would not hesitate to charge at 20A on my board. . . But then I also charge 100V 10Ah lipo packs next to my bed :wink:

-methods

 

[ZapPat]

There are a few ways to do this.
One way would be to isolate the 12v bus line and the any shunts line with diodes so you can supply the line separately from the charging circuit. You would need to supply a voltage (like 5v or 12v) from the controller that would get pulled down by the optocouplers when a shunt goes full. The signal on the any shunts line would be normally high and go low when you need to kill the regen. You would need to cut the traces on the board and jump the cuts with diodes.

This Plan A looks good - How much current would the 5V/12V supply need in this situation?

Plan "B" would involve powering the control circuit whenever the controller is on. This would drain about 10ma or so. Once the control circuit is powered, the any shunts line will be pulled up by the BMS. To backfeed the voltage regulator, you can tie in a single diode and find some voltage supply from the controller that's in the allowable range. I think 5v would be too low. 12v would work, but the normal state for the any shunts line will be a few volts lower due to drop in the regulator.[...]

I could use my controller's Vdrv (~12V) to do this, but if I undertstand this solution would use more current than Plan A?


So this solves the "first cell charged" Over-Voltage cutoff, but what of the first cell Low Voltage Cutoff signal? It seems that plan B would work I guess since it powers up the circuit anyways, but what of plan A?

BTW, is there any complete schematic available for the BMS? Not to pry or anything, I just figure it might avoid some stupid questions on my part...

Thanks!
Pat

 

[fechter]

This Plan A looks good - How much current would the 5V/12V supply need in this situation?


I could use my controller's Vdrv (~12V) to do this, but if I undertstand this solution would use more current than Plan A?


So this solves the "first cell charged" Over-Voltage cutoff, but what of the first cell Low Voltage Cutoff signal? It seems that plan B would work I guess since it powers up the circuit anyways, but what of plan A?

BTW, is there any complete schematic available for the BMS? Not to pry or anything, I just figure it might avoid some stupid questions on my part...

Thanks!
Pat

Plan A does not draw any current at all until a cell goes low, then it's less than 1ma.

Plan B draws around 10ma all the time.

The LVC bus is completely isolated from the charging circuit so you can use tie it in either way.

I think this is the current schematic. It is constantly changing.
(click to view)

 

[fechter]

Hey everyone

Could anyone please advise what the max amps I can charge with this bms? Is there a detailed spec on all the feature of this bms somewhere?

thanks
-steveo

The maximum charge current is limited by FET heating.
If you're using a IRFB4110 with no heat sink, I think around 20 amps will be the max.

With a large enough heat sink, you could take it up as high as 40-50 amps where the thickness of the copper on the board may become the limiting factor. The FET is rated for up to 70 amps continuous. Additional FETs in parallel would reduce heating at higher currents.

Be sure you do not exceed the safe limits for your batteries.

 

[GGoodrum]

We think the max current, based on the traces for the FET and wire size holes, etc., is around 30A. If you wanted to do higher, you'd need to beef the traces up, at least. I don't know anyone who has even tried 30A yet, so I don't know for sure.

The new version board (v2.3) is available now here: http://www.tppacks.com/products.asp?cat=26. I've updated the BOMs, and the instructions, which can be found here: 4-24 Cell LiFePO4 BMS Assembly and Test Instructions. The latest schematic is v2.3, which is here. It is functionally the same as the one Richard posted, but that one has an error, as there are two Q2s.

-- Gary

 

[Kinks]

Awesome! I'm getting 4. I though you were going with active cutoffs for low voltage and not an ebrake line for this version for some reason...

 

[methods]

I think this is the current schematic.

Fechter. . . Is that a 3.2ohm resistor on that schematic?
Have you tried that circuit out?

I am using 3.9ohm resistors and they make the circuit so temperature dependent that it goes into thermal runaway if I don't manage it correctly.
I dont think 3.2ohms will work without a mandatory fan.

-methods

 

[methods]

How many volts over HVC are you setting your supply to get it to "kick off" when using the 3.2ohm resistors?
I have found that if I don't set the supply at least a few volts over HVC it tends to run off into the weeds.
Garry said 0.5V should do it but trying that with the 1A shunts leads to endless charge time.
It gets really hot then gets stuck in an under-damped response pattern.
Cooling the resistors is a must to get it to kick off.
Either by a fan or by turning it off for 2 or 3 minutes then back on.

Now I dont have any experience with normal 1/2A boards so maybe the things I have noted are normal.
Is it normal for the circuit to kind of drift around, one channel going up, then down, then up. . .
Or is the circuit usually fairly linear?
My expectations are for the circuit to kick off a few minutes after the LED's are all light bright
I have heard you guys speak of hours. . .

I had assumed that mine was a little "loosey-goosey" because of my over-amping.
Perhaps not.

-methods

 

[velias]

I haven't been following the thread or schematic too closely.
What are the main differences between this new (V2.3) version of the BMS and the previous version from a couple of months ago?

 

[methods]

For anyone using the 1A version:

disclaimer: 1A is Gangster, stick to 1/2A if you don't want to go the extra mile for thermal management.

I have been making up some acrylic cases to show off the board and keep it cool.
I started with the testing rig on the right. Under the painters tape is a big brushless fan.
Up on the left you see a later version where I integrated a 5V fan.
My Project Mayhem bike blew up and so did my Chopper so this got side-lined for a while

Thermal management is easy. Do it and you will be rewarded. Looks cool, is cool.

Sorry about the cell phone pictures, I lost the charger for my digital camera

 

[methods]

Sorry to be a spammer. . .

DOH! I blew up my BMS board again. . . ARG!
http://endless-sphere.com/forums/viewtopic.php?f=6&t=7169&p=125027#p125027

How many things are wrong with this circuit?

View attachment Another_Project_Mayhem_Setback_002.jpg

Red = +100V
Green = +12V
Black = Common Ground




-methods

 

[fechter]

Doh!, I hate it when that happens!

If you put the positive terminal in the middle of the connector, reversing the connection will be less likely to cause damage.

 

[Roy Von Rogers]

I'm ready to purchase one of those boards. I'm using A123's, anything I need to do different in the parts count. I know the lvc is set to 2.1, is that a bit close for a A123 low voltage cut off ??
And whats the high voltage set at ??

Roy

 

[GGoodrum]

I'm ready to purchase one of those boards. I'm using A123's, anything I need to do different in the parts count. I know the lvc is set to 2.1, is that a bit close for a A123 low voltage cut off ??
And whats the high voltage set at ??

Roy

The a123 cells work fine with s 2.1V cutoff. They have such high C ratings, the voltage stays up high, all through the capacity, and then dumps quick at the end. All the LVC needs to do is detect this quick drop before the cell dumps completely. I've tried the 2.7V chips and the 2.1V versions, and I can't really notice a difference.

On the charging end, the shunts keep the charge voltage at about 3.68V.

-- Gary

 

[Roy Von Rogers]

Thanks for the info, seems I remember you mentioning that somewhere, but I have done so much reading here, after a while it all seems to melt together..lol And I forget what specs belonged to where..grin

I stuck the bom in to Mouser and as others have said, they have some stuff on backorder, I just stuck it in to projects.
Not going to be doing anything soon, I'm slated to get butchered wednsday and have no idea how long it will take before I can do anything. Just hope I come out better then when I went in..lol

Not going to make any major moves on building an ebike til I see how this all works out. I know one thing, I paid good money for this dang A123 developers pack, and I'm going to electrify something, even if I have to order a hub kit for a wheel chair..chuckle.


Roy

 

[ZapPat]

One way would be to isolate the 12v bus line and the any shunts line with diodes so you can supply the line separately from the charging circuit. You would need to supply a voltage (like 5v or 12v) from the controller that would get pulled down by the optocouplers when a shunt goes full. The signal on the any shunts line would be normally high and go low when you need to kill the regen. You would need to cut the traces on the board and jump the cuts with diodes.
View attachment 1

Well since this method (plan A) is quite simple to do and uses less current, it is the winner. I'll have to order at least one board before you are out of stock once again!

Have you considered making a surface mount version for faster assembly times and reduced board area? And only once the design is stable that is. The most difficult thing I see as to part size reduction is the power resistors of course, where maybe some form of heatsinking could help.

Thanks for the help, Richard!

 

[silicium]

Hi Fechter,

Congratulation for your design 8)

Let me ask a few questions:

How limited the intensity during the charging ?
And during the balance?
There is a current generator?
In my BMS circuit, the intensity is limited to 0.5 A during the balancing but is not limited during the charge (You must use an external power supply limited in intensity).
This is managed in the same way?

 

[fechter]

Hi Fechter,

Congratulation for your design 8)

Let me ask a few questions:

How limited the intensity during the charging ?
And during the balance?
There is a current generator?
In my BMS circuit, the intensity is limited to 0.5 A during the balancing but is not limited during the charge (You must use an external power supply limited in intensity).
This is managed in the same way?

Hi Silicium,
I owe you a lot of credit on the design, which I borrowed heavily from yours.

The charge current is not limited until at least one cell gets up to the shunt voltage. The charging supply must have some current limiting for bulk charge. Some people have run it up to 30 amps.

Once any cell gets up to shunt voltage, the current will start to taper and eventually be limited to whatever the shunt is designed to handle (500ma). The charge current is limited by PWM on the FET, which does not need to dissipate much heat since it is in switching mode. Current will stay at 500ma until all cells reach the shunt voltage (balance mode).

 

[fechter]

Have you considered making a surface mount version for faster assembly times and reduced board area? And only once the design is stable that is. The most difficult thing I see as to part size reduction is the power resistors of course, where maybe some form of heatsinking could help.

Thanks for the help, Richard!

Yes, we are moving toward a fully built surface mount version. I found the cost of the surface mount power resistors to be prohibitive, so I think the resistors will stay through-hole parts, but the rest of it can be smd.

 

[flashedarling]

I've gone and ordered one. I figure I have time before it is warm enough to ride so before I order my batteries I'll see if I can manage to build one of these.

Just a few questions. Is there any real advantage to breaking it up and building 2 36V boards instead of a single 72V board?

Also for the 3.9 Ohm resistors would getting a higher wattage, like the " 588-TUW15J3R9E" 3.9Ohm 15W resistor, make any difference for heat dissipation. Or should I just get the 5W resistors because they are cheaper. Other than the resistors should I make any change to the BOM for the higher amp shunt version.

 

[ejonesss]

a way that you can get the board smaller and keep the costs down is to vertically mount the resistors as shown below.

the only problem here is you will have to re design the board layout (probably too late now) and re design the enclosure if there is any.

the advantage a smaller horizontal footprint (taking up only the area of the end of the resistor.

Have you considered making a surface mount version for faster assembly times and reduced board area? And only once the design is stable that is. The most difficult thing I see as to part size reduction is the power resistors of course, where maybe some form of heatsinking could help.

Thanks for the help, Richard!

Yes, we are moving toward a fully built surface mount version. I found the cost of the surface mount power resistors to be prohibitive, so I think the resistors will stay through-hole parts, but the rest of it can be smd.

 

[silicium]

Hi Fechter,

Congratulation for your design 8)

Let me ask a few questions:

How limited the intensity during the charging ?
And during the balance?
There is a current generator?
In my BMS circuit, the intensity is limited to 0.5 A during the balancing but is not limited during the charge (You must use an external power supply limited in intensity).
This is managed in the same way?

Hi Silicium,
I owe you a lot of credit on the design, which I borrowed heavily from yours.

The charge current is not limited until at least one cell gets up to the shunt voltage. The charging supply must have some current limiting for bulk charge. Some people have run it up to 30 amps.

Once any cell gets up to shunt voltage, the current will start to taper and eventually be limited to whatever the shunt is designed to handle (500ma). The charge current is limited by PWM on the FET, which does not need to dissipate much heat since it is in switching mode. Current will stay at 500ma until all cells reach the shunt voltage (balance mode).

I am pleased that my BMS scheme could help, your schedule is easy! therefore easier to achieve.
I did not understand the system hovered. On my BMS, there is also an alternating load and balance but the time is 5 seconds ...
Bravo for the tip, it is simpler than my system!

 

[fechter]

Just a few questions. Is there any real advantage to breaking it up and building 2 36V boards instead of a single 72V board?

Also for the 3.9 Ohm resistors would getting a higher wattage, like the " 588-TUW15J3R9E" 3.9Ohm 15W resistor, make any difference for heat dissipation. Or should I just get the 5W resistors because they are cheaper. Other than the resistors should I make any change to the BOM for the higher amp shunt version.

A single 72v board will perform the same as two 36v boards.

I'm not sure I would recommend using the 3.9 ohm resistors. That will get very hot. 15W resistors will be fine from an electrical standpoint, but I don't think they will physically fit on the board (too big). At higher currents the Q101 transistor needs to be the TO-126 type. This is helpful even at 500ma. Something like a MJE170/171/172 would be good.