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

[GGoodrum]

also congrats and assume that is degrees F and not C ???? most of the world has dropped F

We don't give a rat's ass what the rest of the world is doing, we are bound and determined keep our inches, feet, yards, miles, grams, ounces, pounds, tons, teaspoons, tablespoons, cups, quarts, gallons and our Fahrenheit thermometers.

I remember that back in the '70s, there was a big push to go metric, but it didn't stick. There really hasn't been much push to change since then.

As Andy has pointed out, we are done with the cell circuit portion, and have the HVC working with both the relay-based charge controller, and with the new FET-based variant. I've been testing the latter, and it works quite well. I think the balance times will be better because we've switched the on-off logic around, so that when the HVC hits, there's a short delay before the FET cuts power and then when the HVC resets, it is instantly turned back on. The relay version cuts power as soon as the HVC is hit, and then forces a small delay before it turns the relay back on. The net result is that the power stays on a bit longer during each cycle, with the FET version, which lets the low cell(s) catch up a bit quicker. As you can see in Andy's last graph, eventually the low cell will get close enough that the shunts for the other cells can keep the high cells in check on their own, so the HVC stops tripping. The low cell then catches up with the rest fairly quick.

As I've said, I'm testing the new FET-controlled HVC throttling portion of the charge controller. All that is left is to add in the low current detection/auto-shutoff function, and the LED logic. Richard is working on this portion. As soon as he does some more testing in this area, I'll incorporate whatever changes he makes into a new run of boards.

-- Gary

 

[Gregb]

also congrats and assume that is degrees F and not C ???? most of the world has dropped F

We don't give a rat's ass what the rest of the world is doing, we are bound and determined keep our inches, feet, yards, miles, grams, ounces, pounds, tons, teaspoons, tablespoons, cups, quarts, gallons and our Fahrenheit thermometers.

I

Well at least you seem to have dropped rods, poles perches and chains which I had to learn and you at least had decimal currency before us. Guineas, Pounds, crowns, half crowns, shillings and pence. And farthings just before my time. (1/4 of a penny. 12 pennies to a shilling, 240 to a pound, 21 shillings to a guinea.....was a nightmare altho we seemed to cope.

 

[gblast123]

Soooo, pardon this very nooob question but

Is it done yet?

 

[GGoodrum]

Soooo, pardon this very nooob question but

Is it done yet?

I think we are very close. We've fixed the last remaining "weirdness" issues, and tested the HVC operation with the new FET-based controller logic, and it works well. Richard has made some improvements in the rest of the control logic that is used for the auto EOC shutdown and the auto-start function that is used in place of a start/reset button. I've finished laying out the new board, which is shown below. I will turn these in tomorrow.



When we get the boards, we'll do one last set of verification tests, and then we should be good to go.

-- Gary

 

[chxs]

Hi !! I'm very happy to hear this !! so this is possible for some one like me to get 28S version ?? Tanks
Chris

 

[gblast123]

Another noob question. About how many amps, continuous, will this BMS be able to discharge under "normal" circumstances and is there a way to beef up the current delivery in the event that we want a high power application? (some of us are a little calorically challenged so a regular motor/bms simply will not last as long unless we build in a safety factor)

I define High Power to be above 40 amp continuous.

David

 

[Gregb]

The BMS has nothing to do with the discharge RATE that is a factor of the battery chemistry and your use of the throttle and the type of motor controller. The BMS function is to control the charge rate and the voltage that the cells attain. When they get to a certain point it "throttles them back" to prevent damage from over charging and overvoltage. HVC. It's other major function is to prevent any parallel group of cells from being over discharged which can also destroy cells. LVC . It is up to you to purchase cells with the discharge profile you need. This is discussed in other threads. You must also buy a charger with a voltage/current profile suitable for your pack

 

[fechter]

Unlike most chinese units, this BMS does not limit the discharge current. That is up to your motor controller.

 

[gblast123]

Unlike most chinese units, this BMS does not limit the discharge current. That is up to your motor controller.

I guess i don't understand very much, but if the purpose of the LVC is to prevent the cells from overdischarging, how does if cut off the current if the current is not flowing through the circuit?

Or to put it another way, if the current is flowing through the circuit, isn't there some limit as to how much current the circuitry can handle?

again, please excuse my lack of knowledge.. I have only some basic fundamentals of circuits (I was a chem major).

I tried to read through this thread and I kept on getting lost when things got very technical.

David

 

[Arlo1]

Unlike most chinese units, this BMS does not limit the discharge current. That is up to your motor controller.

I guess i don't understand very much, but if the purpose of the LVC is to prevent the cells from overdischarging, how does if cut off the current if the current is not flowing through the circuit?

Or to put it another way, if the current is flowing through the circuit, isn't there some limit as to how much current the circuitry can handle?

again, please excuse my lack of knowledge.. I have only some basic fundamentals of circuits (I was a chem major).

I tried to read through this thread and I kept on getting lost when things got very technical.

David

In short and simple it monitors voltage and not current for LVC so when the battery gets down to a set voltage it does something to stop you from discharging any further but it does not care how much current flows through the curcit. It just trys to keep the battery cells with in a safe range of voltage.

 

[Alan B]

The LVC sends a signal to the controller to shut down, either via the ebrake or throttle inputs to the controller.

I believe there are two setups with this V4 system. One has FETs to interrupt power, the other sends signals to the controller. If you read back into this thread you can find the setup, it has several TO220 FETs in parallel, so it has pretty high current capability. If it is not high enough just bypass that and use the throttle/ebrake shutdown instead.

 

[Mesuge]

Unlike most chinese units, this BMS does not limit the discharge current. That is up to your motor controller.

That's understood, however what about voltage,
is it going to work also with those "mega pack" setups e.g. like long string ~700VDC ?
Driving: +/- 336VDC => [+ 6s(16s LiFePO4@3.5VDC) - 6s(16s LiFePO4@3.5VDC)]
Charging: block of 16s LiFePO4@3.5VDC per one charger

Or where is the treshold is it "safe" at least for ~200-250VDC pack?

 

[fechter]

The optocouplers are rated for over 5000v, so the cell circuits should be OK.
The problem would be for the control circuit. It would be possible to do a custom control circuit that uses an external 12v supply, but you need some way to interrupt the charging current on HVC. The 'standard' control circuit is rated for 100v (or 150v with alternate FETs), but you could possibly use solid state relays to switch off the AC input to the chargers. Many high end chargers also have an interface that can be used to limit the charge current.

 

[bigmoose]

Gents with respect to the high voltage usage of any circuit board, mind the creepage distance on PCB traces, they often are the limiter and not the breakdown of the opto isolator itself. Here is a nice summary of the various specs in worldwide use for reference: http://www.smps.us/pcbtracespacing.html
Nice little on line creepage design tool: http://www.creepage.com/
IPC2221 Creepage table: http://www.smpspowersupply.com/ipc2221pcbclearance.html
A good conformal coating after wash and bake-out helps also.

Hope this helps if you are applying anything up near the kilovolt level, extra care is warranted.

 

[fechter]

Thanks for pointing that out. I didn't mean to imply that the circuit could be used at 5kV. With the spacing present on the opto pins, it looks like 300v should be within the safety guidelines.

 

[GGoodrum]

Andy is building up one of the new boards right now, and will be testing it this week. If things go like they are supposed to, I'd then declare these finished, and we'll make them available.

-- Gary

 

[Arlo1]

Andy is building up one of the new boards right now, and will be testing it this week. If things go like they are supposed to, I'd then declare these finished, and we'll make them available.

-- Gary

AWESOME

 

[AndyH]

It's looking good so far!

 

[jk1]

Does this BMS do Cell level balancing during charging ?

I.e with this BMS could we simply use an SLA or NICD charger and do a bulk charge with say a 10 series Lifeop4 cells pack ?

Or do we still need a balancing charger ?

 

[Gregb]

I really think you should read some of the earlier posts in this thread, but to answer your question, yes it does. However a normal sla or similar charger is seldom CC/CV and you will need one for correct charging. It does its balancing at the end of the charge regime.

 

[chxs]

Can i buy 28S Preassembled board ??? Tanks

 

[manfred59]

Hello!
I hope somebody could help me with the problems I have, perhaps Fechter
As you see on the photo the second channel does not work as it should. Only the second led is lit and the main led is red. But shorting the pin 7/8 of the optos it gets green.
I have the meanwell plugged (about 61 V) and no cells attached . The LED and the LM431 and the Q101 (KSA 931) I have allready changed!
What could be fried too?

 

[fechter]

It's a little hard to read the numbers on the picture.
Are those voltages with no cells connected?
7v across the first channel is abnormal. At 5v across the second channel, the LED should be lit and it is apparently operating the HVC properly.

 

[manfred59]

Are those voltages with no cells connected?

no cells connected!
Please check the voltages on this photo, I took it with the camera in a better resolution!

 

[fechter]

Something is definitely not right there.
It seems like the control circuit part is OK, because you can get it to go green when you short opto pins 7-8.
Do the other channels above have 'normal' voltages?
What would help is to measure the voltage across the shunt resistors on the affected channels. Measure those and let me know what you get. Also see which resistors (if any) are getting hot.

There is just not a lot of circuit there, so not much else that can go wrong.
Here's what should be close to your schematic:

Here's a 2.3 diagram of the cell circuit test points. Your board should be about the same.


On the output side of the optocouplers:

One thing I did figure out about this and all similar boards (Ver.4, Signalab and others) is this:
You can easily blow up the board if you make connections between the board and the cells and the charger is connected (even if it is turned off). When making connections between the BMS and the cells make sure the charger is disconnected from the board!

The reason for this is the charger typically has large capacitors across the output. When making cell connections, voltage from the cells can pass through the BMS and charge the capacitors in the charger. You will see a spark when this happens. Depending on the order of cell connection, you could be placing near full pack voltage across a cell circuit and pump it with several amps of current (or more) as the capacitor is charging.

If all the cell connections to the BMS are made first, then the charger is connected, there way less chance of a cell circuit getting blasted.

If you absolutely must make cell connections with the charger attached, either turn on the charger first to charge the capacitors (won't always happen with all chargers), or place a diode between the charger and BMS to prevent backfeeding the capacitors. This will prevent some chargers from starting, as they need to 'see' the pack voltage to turn on.