Low cost BMS

How do you store and manage your electricity?
User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Thu, 26 Dec 2013, 03:32

4Springs wrote:Can anyone see anything that I'm doing wrong?

Nope. Whatever's wrong must not be shown on either of the two diagrams you gave. Is your battery floating relative to the BMS master's power supply? Does the input to the master use an opto in the same way the cell modules do? Did you have a programmer or anything else connected to any of the cell modules at the time? Did the smoke happen when you connected 6 to 5 or when you connected 5 to master? The fact that 3 and 6 are damaged, as well as the one you saw smoke from (5) suggests that the only reason you didn't see smoke from the others might be that the master wasn't connected to them for long enough.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

arber333
Groupie
Posts: 56
Joined: Mon, 11 Mar 2013, 02:29
Real Name: Arber
Location: Slovenia
Contact:

Low cost BMS

Post by arber333 » Thu, 26 Dec 2013, 03:33

Hi 4springs

First of all you should start numbering your cells in some way. I start from first + terminal. This is how the modules have to be connected. So you should connect module no. 4 with module no. 5 and then from 5 to 6 and out.

I guess the output from the module is not isolated, hence the negative line can add potential across comm lines when you connect additional cells. Neville warned this in his instructions. I also found it the hard way when i inverted the comms of my last module in a box. It fried the last PIC in line, but not one before which is separated by opto.

What kind of power supply did you use for the master? Is it separate 12V battery? Also check the isolation in your car. Is the battery minus realy separate from chassis?

Also if you tried to supply master from the pack your optos are gone...

If you connect broken module to a battery, does it shunt right away?
Check the PIC12F between 2nd and 3rd pin. Mine got melted there.

Try to connet modules like you connect + contacts of the cells. I now use small molex connectors that cannot be reversed.

http://www.ebay.co.uk/itm/110891034520? ... 1438.l2649

A
Last edited by arber333 on Wed, 25 Dec 2013, 16:40, edited 1 time in total.

User avatar
4Springs
Senior Member
Posts: 568
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Low cost BMS

Post by 4Springs » Thu, 26 Dec 2013, 12:12

weber wrote:Is your battery floating relative to the BMS master's power supply?
Yes, in theory, although I haven't actually tested it recently.
weber wrote:Does the input to the master use an opto in the same way the cell modules do?
Yes, in fact the master wasn't connected directly to these cells. There are cells before and after these cells in another battery of six. This second battery is not connected to the first battery, except via the comms wires.
weber wrote:   Did you have a programmer or anything else connected to any of the cell modules at the time? Did the smoke happen when you connected 6 to 5 or when you connected 5 to master? The fact that 3 and 6 are damaged, as well as the one you saw smoke from (5) suggests that the only reason you didn't see smoke from the others might be that the master wasn't connected to them for long enough.
Hmmm, it gets a bit hazy here. The programmer was not connected, and I'm not sure exactly when the smoke happened. Smoke is not an instant telltale, and it could have come from 3, 5, 6 or somewhere else...

Ok, good. I was a bit worried there that I was not understanding something, but what you have said fits with what I "know". So I will go back out now and check very carefully for shorts between boards and interconnects. I'll also check for connections between the battery(ies) and chassis.

User avatar
4Springs
Senior Member
Posts: 568
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Low cost BMS

Post by 4Springs » Thu, 26 Dec 2013, 12:25

arber333 wrote: First of all you should start numbering your cells in some way. I start from first + terminal. This is how the modules have to be connected. So you should connect module no. 4 with module no. 5 and then from 5 to 6 and out.
Yes, but my cells will be numbered in a different order than my modules. This is because I have three battery boxes distributed around the car, and I am running comms cables in a different order to the power cables.
That is difficult to explain, but the diagram is a small example - the modules sometimes do not follow the drive current...
I could wire the modules in the same order as the cells, but this would mean more comms cables running back and forth. Each cell uses an optoisolator, so wiring them in a different order shouldn't matter.
arber333 wrote:What kind of power supply did you use for the master? Is it separate 12V battery? Also check the isolation in your car. Is the battery minus realy separate from chassis?
The master is on a separate 12V battery. I will check that isolation.
arber333 wrote:If you connect broken module to a battery, does it shunt right away?
No, I checked each module separately beforehand, tested voltage, temp, shunt and comms.
arber333 wrote:Try to connet modules like you connect + contacts of the cells. I now use small molex connectors that cannot be reversed.
Yes, I use those same connectors. And I keep on double & triple checking that the red wire is on the left, black on the right!

Thankyou for your comments, I'm about to go out and re-start troubleshooting. Have to get some work in before The Hobbit screens later on today...

User avatar
4Springs
Senior Member
Posts: 568
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Low cost BMS

Post by 4Springs » Thu, 26 Dec 2013, 14:15

I found the problem.
The interconnect on cell 5 -ve terminal is a different style, and is a bit longer than the others. This meant that it sat underneath the Din connector for the BMS circuit board. This is one of the few through-hole components on the board, and I hadn't snipped off the legs underneath. So the two legs were probably both connected to the -ve terminal.
I have now put heatshrink on that interconnect, snipped off the offending legs (from a new circuit board) and put some insulation tape over the bottom of the board. It is all assembled now, and the heatshrink does not touch the tape.
I replaced the BMS modules on cells 3, 5 & 6, and it all works now. I'm not sure that the fault explains why all those modules now don't work, but oh well... I have a small but growing pile of BMS modules that need to be fixed.


User avatar
coulomb
Site Admin
Posts: 3779
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Low cost BMS

Post by coulomb » Thu, 26 Dec 2013, 17:04

Edit: Oops! Had some internet problems this morning, and didn't see the flurry of replies ending with the "found it" post.
4Springs wrote: The first time I decided that maybe I had connected two outputs to each other. But this time I am sure I have the connectors the right way around.
Well, connecting output to output would fry the processor. But you say that didn't happen at least this time.

Could this be something to do with the fact that I am connecting modules in a different order than the cells? I'm no expert on optocouplers, but I thought that I could connect them in any order, since they should be isolated from each other?
Yes, that should be no problem, from what I can see.

I can't see why module 3 would be affected, unless something happened *through* module 4.

Mysterious.

Next step is possibly to figure out exactly what has died. Why do you say that module 3 isn't transmitting? Did you connect the master directly to it, while not connected to 4 at all?
Last edited by coulomb on Thu, 26 Dec 2013, 16:45, edited 1 time in total.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Thu, 26 Dec 2013, 20:16

4Springs wrote: I found the problem.
The interconnect on cell 5 -ve terminal is a different style, and is a bit longer than the others. This meant that it sat underneath the Din connector for the BMS circuit board. This is one of the few through-hole components on the board, and I hadn't snipped off the legs underneath. So the two legs were probably both connected to the -ve terminal.
I have now put heatshrink on that interconnect, snipped off the offending legs (from a new circuit board) and put some insulation tape over the bottom of the board. It is all assembled now, and the heatshrink does not touch the tape.

I was a little confused there for a second. Thinking "DIN connector?" (German standard). But I soon realised you're referring to the "Data in" header.

Having found your layout, I do wonder why all those parts are jammed so close to a terminal. The only part that benefits is the thermistor.

I'm also horrified to see a beautiful optocoupler with 7.6 mm creepage and a 600 V continuous rating, with its input unnecessarily surrounded by parts at cell potential in such a way that the creepage is reduced to about 1.3 mm and therefore possibly only good for a 72 volt battery, unless the PCB is conformally coated.

Image

http://www.ptr.eu/en/technical-know-how ... distances/
http://www.creepage.com/
http://www.ni.com/white-paper/2871/en/
http://www.ce-mag.com/ce-mag.com/archiv ... afety.html

I understand that standard FR-4 PCB material is Material Group IIIa. The Pollution Degree is at least 2 in this application. Although some of the above references seem to give special dispensation for PCBs as opposed to the rest of the equipment, which I don't understand.

In normal operation, the voltage stress occurs at two places: The opto on the first cell that receives comms from the master, and the opto in the master that receives comms from the last cell.

I note that, under NCOP14 sections 2.7 and 2.8, unless your traction battery is less than 60 volts, the comms wires coming from that last cell are at battery potential and are therefore hazardous and must have insulation rated for the full battery voltage and be coloured orange, and if they enter the passenger or load areas they must be protected by orange conduit. The BMS master unit will also require an IP4X case with a Hazardous Voltage warning label. The same requirements apply for any comms cables between battery boxes.

That's why we at first had special end-of-box opto-couplers, and now use optic fibre.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
BigMouse
Senior Member
Posts: 600
Joined: Thu, 28 Oct 2010, 02:39
Real Name: Vincent Tannahill
Location: Silicon Valley
Contact:

Low cost BMS

Post by BigMouse » Fri, 27 Dec 2013, 00:12

weber wrote:In normal operation, the voltage stress occurs at two places: The opto on the first cell that receives comms from the master, and the opto in the master that receives comms from the last cell.
I've been following this and other BMS designs and am curious what the reasoning behind the optocouplers on the individual cells is. As you say, the stress occurs only at either end of the pack. It seems like a more direct data link between cells would be beneficial. Of course, fibre optics are the "gold standard" here.

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Fri, 27 Dec 2013, 01:55

BigMouse wrote:
weber wrote:In normal operation, the voltage stress occurs at two places: The opto on the first cell that receives comms from the master, and the opto in the master that receives comms from the last cell.
I've been following this and other BMS designs and am curious what the reasoning behind the optocouplers on the individual cells is. As you say, the stress occurs only at either end of the pack. It seems like a more direct data link between cells would be beneficial. Of course, fibre optics are the "gold standard" here.

Good question.

One reason is that the massive and rapidly changing traction currents, and the PWM noise, can cause spiky voltage differences between adjacent cell monitoring units (CMUs). One of Tritium's early experiments used capacitive coupling with Manchester encoding. Worked fine on the bench. Utterly useless in the car. Optos are a pretty cheap way to do the required level-shifting, and although not as cheap as capacitors they don't suffer from the same dV/dt problems.

Another reason relates to my proviso "in normal operation". Abnormal operation, e.g. a high resistance joint between cells under load, can cause nearly the full battery voltage to appear between any two cells. (A high-R joint can also make the full battery voltage appear across any cell monitoring unit (in reverse), but optos don't help you with that.)

The plastic fibre system we use (Industrial Fibre Optic) isn't quite as costly as gold. :-) Sure the LEDs and phototransistors with their fibre-clamping collet and knurled nut are about $5 each in small quantity but you only need one of each per battery box. The fibre itself is only $2.20 per metre [Edit:] from Digikey. And you just cut it at right angles with a sharp knife. No polishing. No gluing connectors on. You just poke it into the collet and gently tighten the knurled nut with your fingers.
Last edited by weber on Thu, 26 Dec 2013, 15:06, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
4Springs
Senior Member
Posts: 568
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Low cost BMS

Post by 4Springs » Fri, 27 Dec 2013, 02:47

weber wrote:Having found your layout, I do wonder why all those parts are jammed so close to a terminal. The only part that benefits is the thermistor.

I'm also horrified to see a beautiful optocoupler with 7.6 mm creepage and a 600 V continuous rating, with its input unnecessarily surrounded by parts at cell potential in such a way that the creepage is reduced to about 1.3 mm and therefore possibly only good for a 72 volt battery, unless the PCB is conformally coated.

The board was a copy of Neville's, but was extended for a much wider cell. With the benefit of hindsight I would have liked to re-arrange things a bit. The next one I make will be much better!
weber wrote:In normal operation, the voltage stress occurs at two places: The opto on the first cell that receives comms from the master, and the opto in the master that receives comms from the last cell.
The optocoupler on the master board is laid out better. In theory, the traction battery is isolated from the 12V battery, so the voltage should not be high in these two locations. I understand that theory is not always practice though, so point taken. I will also think about the order of the BMS module boards. If they followed the the same connection order as the battery, then there would only ever be 4V between them. But I am intending to connect them in a different order, so I need to carefully consider how much voltage there is between each one.
weber wrote:I note that, under NCOP14 sections 2.7 and 2.8, unless your traction battery is less than 60 volts, the comms wires coming from that last cell are at battery potential and are therefore hazardous and must have insulation rated for the full battery voltage and be coloured orange, and if they enter the passenger or load areas they must be protected by orange conduit. The BMS master unit will also require an IP4X case with a Hazardous Voltage warning label. The same requirements apply for any comms cables between battery boxes.

Yes, I realised after soldering up 100 comms wires that I should have used orange wire for every last one of them. I actually used red and black utp wire. I really should do them all again, and try to source two different colours of orange wire...
I decided to put the BMS master box under the bonnet. I did this for several reasons:
1. The entry of HAZV into the box.
2. It is really quite large, and I couldn't find anywhere to put it.
3. The distraction of the thing. I know that I'd want to monitor the voltages all the time while I'm driving!
I also made sure that the comms go into the master box via a different connector than the 12V stuff, since I wasn't sure on the voltage rating between different pins on the same connector.

User avatar
coulomb
Site Admin
Posts: 3779
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Low cost BMS

Post by coulomb » Fri, 27 Dec 2013, 03:36

4Springs wrote: I will also think about the order of the BMS module boards. If they followed the the same connection order as the battery, then there would only ever be 4V between them. But I am intending to connect them in a different order, so I need to carefully consider how much voltage there is between each one.

But again, this is only under normal conditions. If a cell goes open circuit, then nearly full pack voltage appears across that BMU.

This is also a problem with the monitors that look after a group of cells. They often use ordinary thin wire, because hey it's only 4 volts between any pin or wire and the next. But if a cell goes open circuit, it's ~ 140 V. Edit: or much more, depending on the pack voltage.
Last edited by coulomb on Thu, 26 Dec 2013, 16:43, edited 1 time in total.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Fri, 27 Dec 2013, 03:53

4Springs wrote:In theory, the traction battery is isolated from the 12V battery, so the voltage should not be high in these two locations. I understand that theory is not always practice though, so point taken.
Right. Even in theory, the best you can hope for is that half the battery voltage appears across each of these two optos. But since they're floating, there's no reason for it to stay that way. So all of it could easily be across a different one of them at different times.
Yes, I realised after soldering up 100 comms wires that I should have used orange wire for every last one of them. I actually used red and black utp wire. I really should do them all again, and try to source two different colours of orange wire...
There is no need for those comms wires that are enclosed inside battery boxes to be orange, so long as the battery box lid is IP4X and has a Hazardous Voltage label.

For the others, you could just wrap them in orange PVC insulation tape.
http://www.rexel.com.au/products/PVC-TA ... 80380.html
Don't bother wrapping it 'round and 'round. Just run the tape along it and fold the tape in half onto itself.
I decided to put the BMS master box under the bonnet. I did this for several reasons:
1. The entry of HAZV into the box.
2. It is really quite large, and I couldn't find anywhere to put it.
3. The distraction of the thing. I know that I'd want to monitor the voltages all the time while I'm driving!
I also made sure that the comms go into the master box via a different connector than the 12V stuff, since I wasn't sure on the voltage rating between different pins on the same connector.

Very wise.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
4Springs
Senior Member
Posts: 568
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Low cost BMS

Post by 4Springs » Sat, 18 Jan 2014, 01:38

weber wrote: There is no need for those comms wires that are enclosed inside battery boxes to be orange, so long as the battery box lid is IP4X and has a Hazardous Voltage label.

Are you sure?
Section 2.7 of NCOP 14 just says that all wiring connected to HAZV must be coloured orange.
I'd really really like it to be true that I don't have to replace them!

User avatar
4Springs
Senior Member
Posts: 568
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Low cost BMS

Post by 4Springs » Sat, 18 Jan 2014, 02:24

I've been a-programming!
Here is a picture of the 5 display modes:
Image

The first screen shows that cell 10 is the highest temperature, at 27. Cell 5 is the lowest at 24.

The second screen shows the last 5 faults. Current fault is 0 (no fault). Before that there was a fault 2 (comms fault) that originated at cell 21. Before that was a comms fault on cell 17, etc. (this was me unplugging various comms connectors). New faults appear at the front, old codes drop out the back.
Fault codes are:
0 = no fault
1 = master comms fault (master did not receive anything)
2 = cell comms fault (a cell did not receive anything)
3 = a cell Voltage has exceeded the low limit
4 = the total Voltage has exceeded the high limit
5 = a cell Voltage has exceeded the high limit
6 = a cell has exceeded the high temperature limit

The third screen shows what happens when one unplugs a comms connector between cells. In this case it is saying "module 6 has reported that it is not receiving data from module 7". Modules 5 through 1 have relayed that message on all the way back to the master. All cells stop shunting in this mode, and the charger is disabled.

The fourth screen is the Voltage display mode. Highest cell Voltage is 3.29, lowest is 3.23. Total Voltage is 67.2. The cell with the highest Voltage is cell 16, and the lowest is 4.

The final screen shows the same high and low cell Voltages on the first line, and the second line shows how many cells are currently in balancing (shunt) mode.

I have implemented a PWM output to control the charger output. The PWM will be converted to 0-5V via an optocoupler, capacitor and resistor as described in the TC Charger manual. So the charging current will be scaled back as the first cell modules start to shunt.

Testing is ongoing. I haven't hooked it up to the charger yet.
The fault code stack is proving useful. One of my modules throws a low voltage out every few minutes. I'll have to remove that board and re-do the soldering.

arber333
Groupie
Posts: 56
Joined: Mon, 11 Mar 2013, 02:29
Real Name: Arber
Location: Slovenia
Contact:

Low cost BMS

Post by arber333 » Sat, 18 Jan 2014, 03:49

Excellent work!


Error codes are just the thing i wished my modules to report. Specially the info which cell the fault originated at.
Can i PM you for the source code of the fault reporting?

I havent made much programming other than just trying to sync my charger to the BMS. And it works. Charger now does 40A at 3,45V/cell and cells stay there for 5h. But then one cell starts to go trough 3,55V i set as the limit... I then programmed the charger to lower amps to 5A and all cells finish in cca 15min.

I also found out 1A is about max what is practical in enclosed box. At 2A shunting in cca 20min cells show 50deg!!! Sometimes this causes comms to fall out - probably PIC12 is reset. So i had to find the optimum amps to fill all cells while still being able to hold them below MaxCellVoltage i set in code.

I am hard at work to modify my charger for 3phase full bridge, to use whole power 3phase can deliver. I hope for 10kW charge! Also i hope to connect to those new Mennekes 22kW stations they are seting up in Europe. When i am done i can put my mods on this forum...


A
      
Last edited by arber333 on Fri, 17 Jan 2014, 19:37, edited 1 time in total.

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Sat, 18 Jan 2014, 04:34

4Springs wrote:
weber wrote: There is no need for those comms wires that are enclosed inside battery boxes to be orange, so long as the battery box lid is IP4X and has a Hazardous Voltage label.

Are you sure?
Section 2.7 of NCOP 14 just says that all wiring connected to HAZV must be coloured orange.
I'd really really like it to be true that I don't have to replace them!
No, I'm not sure. The fact that HazV wiring must be coloured orange even when it's in orange conduit gives me pause, and seems bizarre. Can anyone suggest a reason for that?

My understanding is that emergency services know not to cut into any cable or conduit that's orange or any enclosure that has a HazV sticker. So why should it matter what colours are inside the conduit or enclosure?
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
4Springs
Senior Member
Posts: 568
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Low cost BMS

Post by 4Springs » Sat, 18 Jan 2014, 12:41

arber333 wrote: Can i PM you for the source code of the fault reporting?

Yes, sure, I've sent you an email. I'll give my code away to anyone who wants it. When it is completer I'll post it up somewhere for all to access.
arber333 wrote:I also found out 1A is about max what is practical in enclosed box. At 2A shunting in cca 20min cells show 50deg!!! Sometimes this causes comms to fall out - probably PIC12 is reset. So i had to find the optimum amps to fill all cells while still being able to hold them below MaxCellVoltage i set in code.

Good, we are doing different things!
I have reduced my bypass current to about 0.2A. I have heard on this forum and elsewhere that once balanced, LFP cells don't need much bypass to stay balanced. At 16A my charger is also delivering a lot less current than yours. We'll be able to see the difference between our different approaches.

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Sat, 18 Jan 2014, 13:28

4Springs wrote:Section 2.7 of NCOP 14 just says that all wiring connected to HAZV must be coloured orange.
I'd really really like it to be true that I don't have to replace them!
You still don't have to replace them. Just buy some orange electrical tape and fold it over them.

Coulomb and I have orangified lots of stuff that way. In particular, orange corrugated conduit is heavy duty and therefore very hard to bend in a tight radius. So we have used the grey medium duty conduit in some places, and wrapped it helically in orange PVC electrical tape.

But I figure, if someone is using a tool to remove a lid marked "DANGER ELECTRIC SHOCK HAZARD INSIDE", or with the lightning bolt in the yellow triangle, then they should either know what they are doing, or should assume that everything inside there is hazardous.

Will you be making your high current links between cells orange too?
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
4Springs
Senior Member
Posts: 568
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Low cost BMS

Post by 4Springs » Sat, 18 Jan 2014, 13:50

weber wrote:Will you be making your high current links between cells orange too?

Ha!
The NCOP gives special dispensation for a bit of red or black heatshrink at the ends of the cables, so the links fall into that category! Image

Yes, thanks for the tip about the orange tape, I bought some and have been using it on my comms cables where they have to exit the battery boxes.

I do like the theory that once inside a HAZV box you don't need orange wires. I'm not about to pull apart my new charger to make sure all the internal wiring is orange, I'll just whack a HAZV sticker on it. I'm sure most engineers would agree - after all, the NCOP 14 is just a guideline, and common sense should prevail?

arber333
Groupie
Posts: 56
Joined: Mon, 11 Mar 2013, 02:29
Real Name: Arber
Location: Slovenia
Contact:

Low cost BMS

Post by arber333 » Sun, 19 Jan 2014, 03:07

4Springs wrote:
Good, we are doing different things!
I have reduced my bypass current to about 0.2A. I have heard on this forum and elsewhere that once balanced, LFP cells don't need much bypass to stay balanced. At 16A my charger is also delivering a lot less current than yours. We'll be able to see the difference between our different approaches.


Well actually i believe i hurt one of my cells.
When i charge at 40A it is the first one to reach 3,55V, but when i lower to 5A it falls down among other cells. Usually at this time i charge at 5A since other cells start to balance. In about 15min i have 42 cells leveled save one. This one - same one that was 3,55V at 40A now climbs slowly from 3,40V to 3,55V. If i disconnect charger it will drift to 3,43V in a matter of hours.
I figure internal resistance must be higher than others. Yet it holds charge as well as others. I guess a year of bottom balancing caused some damage...


User avatar
4Springs
Senior Member
Posts: 568
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Low cost BMS

Post by 4Springs » Mon, 20 Jan 2014, 01:02

I've connected all the cells of my BMS in the car, and I'm having some comms problems. I haven't driven anywhere yet, so it shouldn't be interference from controllers etc.

I have one battery pack in the rear and two in the front. It appears that the run from front to back (or more precisely from back to front) is too long for reliable communications between modules.
The Master is in the front (under the bonnet). It transmits to one module in the front, which transmits to the rear battery pack. The data then comes back to the front, and is passed around the front cells before coming back to the Master.

The symptom was that I was getting a lot of high and low cell voltages. I found that I could get the faulty readings to stop by bypassing the rear modules. No faults at all with just the front modules, but if I added just one rear module then there were lots of faults.
I thought that perhaps I could ignore the bad packets of data, if I could tell which ones they were. So today I implemented CRC checking on the packets of data. Each cell does a very simple CRC (just adds up all the bytes), and adds the CRC to the end of the packet. If a received CRC does not match the calculated one, the cell does not pass on that data.
The result is that I no longer get corrupt data, but it also means that very little data gets through!

I had a look back through previous pages of this topic, and eventually found where Neville has mentioned how long his cable runs were. I had thought that they would be longer than mine, but he only gets up to 2m! Mine is about 4m.
I also found that Arber put a 0.01uF capacitor over the resistor on the output of the optocoupler. Neville suggested this particular value after viewing the waveform on an oscilloscope. I tried this, but it doesn't seem to make much difference.

Any suggestions on getting a long cable run to work?
I could borrow an oscilloscope to look at the waveform, and fine tune a capacitor.
I could reduce the baud rate.
I could investigate using those fandangled fibre optic thingos that Coulomb & Weber use.
Would a higher voltage be more likely to succeed? I could make up a module running from the 12V system for this long run.


arber333
Groupie
Posts: 56
Joined: Mon, 11 Mar 2013, 02:29
Real Name: Arber
Location: Slovenia
Contact:

Low cost BMS

Post by arber333 » Mon, 20 Jan 2014, 01:35

4Springs wrote: I've connected all the cells of my BMS in the car, and I'm having some comms problems. I haven't driven anywhere yet, so it shouldn't be interference from controllers etc.


You are probably picking up your 12V power cables. Check your cables:
1. are you using twisted pair for communication? normal UTP cable is the best since it is already twisted in pairs, is sleeve protected and has enough strands to use for return comm. It is cheap also.
2. DO NOT use cables in the same bunch to transfer comm and 12V!!! I had a lot of headaches when i used one line to power master along comms.
4Springs wrote: I have one battery pack in the rear and two in the front. It appears that the run from front to back (or more precisely from back to front) is too long for reliable communications between modules.
The Master is in the front (under the bonnet). It transmits to one module in the front, which transmits to the rear battery pack. The data then comes back to the front, and is passed around the front cells before coming back to the Master.


I dont think cable length is important. Quality of cables is more important though, as well as connectors you use. I also use 3m cable in the middle and it is not a problem anymore. As long as i dont rout 12V trough it.
4Springs wrote: The symptom was that I was getting a lot of high and low cell voltages. I found that I could get the faulty readings to stop by bypassing the rear modules. No faults at all with just the front modules, but if I added just one rear module then there were lots of faults.
I thought that perhaps I could ignore the bad packets of data, if I could tell which ones they were. So today I implemented CRC checking on the packets of data. Each cell does a very simple CRC (just adds up all the bytes), and adds the CRC to the end of the packet. If a received CRC does not match the calculated one, the cell does not pass on that data.
The result is that I no longer get corrupt data, but it also means that very little data gets through!


Yes i tried something like that also. In the end i just set up the exact number of cells in code and master stopped adding cells on its own. Try not to use CRC, since Neville set comms up so they would cycle quickly. Lets try to find problem at the root.
4Springs wrote: I also found that Arber put a 0.01uF capacitor over the resistor on the output of the optocoupler. Neville suggested this particular value after viewing the waveform on an oscilloscope. I tried this, but it doesn't seem to make much difference.


Make RC filter for master input opto like you saw me. BUT you must use exactly the same optos as per specs. Otherwise you should select the valuse of cap with your scope, since the top of the opto curve will not be smooth enough.

Also there are some things you can try.
You make power filter for your master. 12V can cause trouble. This can be fought in different ways.

1. Use 12V - 12V galvanicaly separated power supply. 3W is enough i think. BUT this means you would have to supply power (also ground plane) to relays from main board!
Here i am thinking for some time to use a 240VAC 12VDC cell phone charger. It is galvanicaly isolated and works from 100V up to 240VAC as well as DC. Usually they supply 750mA and are very cheap.

2. Make or use an LC filter for your 12V power supply. You can buy car audio LC filter....

I also soldered one 0,1uF ceramic SMD cap across 12V input on master. It helped initialy however the best i got was one 0,1uF cap on the R1 of the master opto.

There is one other thing. When do you receive those bad comms? Do you have charger connected? Is your charger grounded to chassis? Check there.
Is your battery - isolated from 12V gnd chassis?

If all else fails, you could use optic omms for the last box. However then i recommend using both in and out lines.
Check here: http://www.rcgroups.com/forums/showthread.php?t=500234

A


Last edited by arber333 on Sun, 19 Jan 2014, 14:45, edited 1 time in total.

User avatar
weber
Site Admin
Posts: 2624
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber » Mon, 20 Jan 2014, 01:58

4Springs wrote:I found that I could get the faulty readings to stop by bypassing the rear modules. No faults at all with just the front modules, but if I added just one rear module then there were lots of faults.
When you say "bypassing the rear modules" do you mean that the signal was still travelling from the front to the rear and back again, but just not going thru any modules at the rear?

Can you point us to the manufacturer's datasheet for your comms cable?
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
4Springs
Senior Member
Posts: 568
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Low cost BMS

Post by 4Springs » Mon, 20 Jan 2014, 02:21

weber wrote: When you say "bypassing the rear modules" do you mean that the signal was still travelling from the front to the rear and back again, but just not going thru any modules at the rear?
No, for it to work reliably I had to disconnect the long cables. So:
1. 12 front modules working just fine.
2. Add 1 rear module (and two long cables) and I get lots of errors.
weber wrote:Can you point us to the manufacturer's datasheet for your comms cable?
To start with I used two of the pairs in some cat 5 cable bought from Jaycar. Neville swears by manually twisted UTP though, so I also tried that. No discernible difference with either or both as UTP.

User avatar
4Springs
Senior Member
Posts: 568
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Low cost BMS

Post by 4Springs » Mon, 20 Jan 2014, 02:41

arber333 wrote: You are probably picking up your 12V power cables. Check your cables:
1. are you using twisted pair for communication? normal UTP cable is the best since it is already twisted in pairs, is sleeve protected and has enough strands to use for return comm. It is cheap also.
2. DO NOT use cables in the same bunch to transfer comm and 12V!!! I had a lot of headaches when i used one line to power master along comms.
Can you post a picture of your UTP? My understanding of UTP was "Un-Twisted Pair" - just a single wire.
I don't have any 12V anywhere near the comms at the moment. There is almost nothing powered in the car, probably just the clock!
arber333 wrote:I dont think cable length is important. Quality of cables is more important though, as well as connectors you use. I also use 3m cable in the middle and it is not a problem anymore. As long as i dont rout 12V trough it.
This is good to know. Thanks. So I should be able to get it working!
arber333 wrote:Try not to use CRC, since Neville set comms up so they would cycle quickly. Lets try to find problem at the root.
My code is different from Neville's because I used a different compiler. My comms work half duplex - modules can only send or receive, never both at the same time. So they buffer the entire data packet anyway. A CRC was easy to implement, but I agree it would be better if it wasn't needed.
arber333 wrote:Make RC filter for master input opto like you saw me. BUT you must use exactly the same optos as per specs. Otherwise you should select the valuse of cap with your scope, since the top of the opto curve will not be smooth enough.
I am using the same opto that Neville suggested. I saw in your notes though that you were filtering on the Master. Did you ever do it for the modules? I am having the trouble from one module to the next.
arber333 wrote:There is one other thing. When do you receive those bad comms? Do you have charger connected? Is your charger grounded to chassis? Check there.
Is your battery - isolated from 12V gnd chassis?

Charger is not connected. Battery is isolated from 12V & chassis. Bad comms are only when the long cable run is connected.
Thanks for your help!

Post Reply