Low cost BMS

How do you store and manage your electricity?
Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Low cost BMS

Post by Nevilleh »

Well, it is what it is. There is no compunction for you to use it if you decide you don't like it.

I have had personal experience of an electric motorcycle cutting power just because the bms detected a cell going too low and it very nearly killed me, so I won't countenance such a silly feature just to save the possible cost of a single cell. But if you want to put people's lives at risk, its on your own head.

cts_casemod
Groupie
Posts: 76
Joined: Fri, 28 Sep 2012, 17:59
Real Name: Carlos
Location: UK - London

Low cost BMS

Post by cts_casemod »

Nevilleh wrote: Well, it is what it is. There is no compunction for you to use it if you decide you don't like it.

I have had personal experience of an electric motorcycle cutting power just because the bms detected a cell going too low and it very nearly killed me, so I won't countenance such a silly feature just to save the possible cost of a single cell. But if you want to put people's lives at risk, its on your own head.


Hi Neville

Please don't get me wrong, as I said your BMS is the best design I have found to date.
I am probably just going to make some changes to the main controller and re-design the smaller modules to suit my needs.
I am just making you aware that it may be a problem if other people are using the car.
I know a guy near me that told me some horror stories about friends crushing gears trying to change them without clutch, just as an example, so I want to have things as normal/safe as possible

A low battery warning when the charge is only 20%, plus a warning 30 seconds before shutting down its more than enough to stop the car on safety. If you wont, on my case, the controller will cut the power anyway.

I had similar problems with my e_bike, the old BMS and batteries were a piece of crap and during acceleration it would cut often. In fact this is one of the reasons I don’t like the commercial BMS's because I understand it may be a very dangerous situation on a car

I had enough when I almost had a fire on my bike, due to a short circuit on a chaffing cable.
The batteries themselves were of poor construction, without separators between them. After this I have found the metallic connection to the positive terminal got so hot it melted a bit of the top insulation I was lucky it didn’t short the top of the cell with the body or the cell next to it. They were lipo batteries. There was a fuse on the battery box, but no cables were welded into it.

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 »

cts_casemod wrote:I am also thinking in a weay to use a multiplexer to read the actual cells to make the circuit a little bit more simple for a large array of cells.


You've mentioned the multiplexer a couple times, and I'm curious what you're planning. How will you use a multiplexer to read the voltages of cells which all have a different (and increasing with each incremental cell) "ground" potential?

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Low cost BMS

Post by Nevilleh »

The first bms I designed was a "central" board with 8 inputs - for 8 cells - and used two quad op-amps. The ones I used were rated at 40-odd volts, so could be powered by the 8 cells in series OK. It used the differential inputs and the high CMR allowed it to work very well. The 8 op-amp sections did level shifting and scaling and the outputs fed an AVR micro with 8 A-D inputs, that was the multiplexor. The micro had a UART and did serial comms back to a control unit.

I actually built several of these boards and installed a couple in my car, but decided the long wires running from the cells to the modules were a real pain in the backside and then the new regulations came out which required all of those wires to e fuse protected. At that point I abandoned that configuration and went with the single module per cell scheme that I am now using.

Its difficult to make a central module that could handle more than 8 cells because the voltage goes up alarmingly! The level shifting and scaling needed becomes more and more difficult once you get over 36 volts or so. So if you limit it to 8 cells, then for a 192 cell battery you still need 24 of these modules. I think its more complex and difficult that using single, small and relatively inexpensive cell modules. I think, from memory, that my single cell module design worked out quite a bit cheaper than the 8-cell version. Its certainly far simpler to install and get going.

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 »

Nevilleh wrote:It used the differential inputs and the high CMR allowed it to work very well. The 8 op-amp sections did level shifting and scaling and the outputs fed an AVR micro with 8 A-D inputs


Ahh, that makes sense. There's no reason something like that couldn't be done in a similar way that Weber and Coulomb's BMS is layed out (one board spanning several cells), even though theirs is actually distributed. Still the benefits probably aren't there, as you say. Not every pack is going to have an even multiple of 8 cells either.

My BMS is due for a redesign. The cost is right, and the feature set is there, but the implementation needs reworking. I'll get around to it eventually Image

User avatar
coulomb
Site Admin
Posts: 3833
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Low cost BMS

Post by coulomb »

BigMouse wrote: Not every pack is going to have an even multiple of 8 cells either.

No, though of course that's no problem, you just cut the boards and use a three-wire jumper from one set of cells to the next, or optic fibre if it has to go a long way.

The squiggle joins are just a time saver, that's all. Though they do mandate a certain cell pitch, of course. You could always redo the boards to take whatever pitch you needed.

I'm a little surprised that throughout all this, no-one has mentioned our BMUs (till the post above). Perhaps it's perceived as being too fancy, or too specific to 40 Ah cells, or something.
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.

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Low cost BMS

Post by Nevilleh »

coulomb wrote:
I'm a little surprised that throughout all this, no-one has mentioned our BMUs (till the post above). Perhaps it's perceived as being too fancy, or too specific to 40 Ah cells, or something.


That may well be because its embedded in you MX-5 thread and people looking for info on BMS(s!) don't find it so easily. But its definitely worth a mention. Maybe you could stick a link in here so anyone reading this can go straight to it.

cts_casemod
Groupie
Posts: 76
Joined: Fri, 28 Sep 2012, 17:59
Real Name: Carlos
Location: UK - London

Low cost BMS

Post by cts_casemod »

BigMouse wrote:
cts_casemod wrote:I am also thinking in a way to use a multiplexer to read the actual cells to make the circuit a little bit more simple for a large array of cells.


You've mentioned the multiplexer a couple times, and I'm curious what you're planning. How will you use a multiplexer to read the voltages of cells which all have a different (and increasing with each incremental cell) "ground" potential?


Hi Big Mouse,

That idea has been given by my teacher at uni. I haven't tested yet, but the aim is to use optocouplers attached to an analog multiplexer to read the voltages, so they would be isolated.

This had the advantage that the main CPU would be able to read all the voltages on the array without the extra need for an individual CPU on the cell boards and from then it could decide what to do (alarms, cut off, etc), just like Neville board. If I wanted I could also have a remote unit to display the user the soc and voltage of the batteries. (so using 2 CPUs on my case would reduce the cost from 200£ to 30£).


I am yet to find if the temperature coefficient of the optocoupler would give me an accurate voltage reading from -10 to 50C, maybe I cut add some compensation with an temperature sesnor.

Another feature could be to use an extra pair of optocouplers to take care of the balancing, if required or even as a way to isolate the pack from the circuit to have a "shut down" (would only draw the battery if the transistor was on by an external power source).

You guys seem to be more familiar than me on this kind of stuff, so is it viable?
Last edited by cts_casemod on Fri, 16 Nov 2012, 20:07, edited 1 time in total.

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Low cost BMS

Post by Nevilleh »

Woohoo! 192 linear optocouplers would cost you something like $US800 (HCNR200) plus a dual op-amp (another $1 each)for each one. You'd still have to make a pcb for each cell with at least those parts on it and then feed all the outputs to a vast combination of multiplexor chips. Could be done, cost it all out! I think you'd have to remove the "Low cost" from the forum title though.

User avatar
weber
Site Admin
Posts: 2641
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Low cost BMS

Post by weber »

Thanks for the offer Neville. Here is some info about the Coulomb & Weber open-source BMS:

What drove us to begin designing our own BMS 3.5 years ago were two deficiencies of BMS available at the time. They weren't designed with adequate isolation for our 750 V battery and, except for some very expensive designs, they weren't designed for cells in inaccessible locations where there was no chance of ever seeing LEDs. And as time went on we became aware of other safety issues and convenience issues with existing BMS, and incorporated solutions to them.

Incidentally, we would probably not use a battery voltage as high as 750 V again, unless it was a racing car or drag car. The economic optimum appears to be around half that, as we found out the hard way. Since readily available components are rated to 400 Vdc you want to keep the voltage on full-charge at or below that.


Image

Description

A battery management unit for the protection of Lithium Iron Phosphate battery cells as used in electric vehicles. A small printed circuit board is mounted atop each cell, which performs bypassing and monitors for high or low cell voltage and temperature and voltage across links between cells.

This design has been optimised for high voltage isolation, and to withstand high forward and reverse voltages across the BMU under fault conditions without creating arcs that could ignite flammable vapour, and for autonomous (i.e. masterless) operation, and to allow the use of low-cost fibre optics for signalling, and not to require physical access for reset or firmware update, and to have low cost.

Features

    Reports cell over and under voltage -- 1 millivolt resolution
    Reports cell over and under temperature -- 1 degree Celsius resolution
    Reports excessive volt drop across links between cells -- 1 millivolt resolution
    Bypasses 0.4 amp for cell balancing -- max hotspot 30°C above ambient
    Designed for 40 Ah ThunderSky/Winston/SkyEnergy/CALB cells but easily adapted
    Designed for up to 900 V battery
    Designed to be arc-free even if cell or link goes open circuit under load or charge
    Daisy-chained comms is 9600 baud serial ASCII for ease of debugging
    Comms between adjacent cells is by opto-isolated flexible join between circuit boards
    Comms outside battery box is by low-cost fibre-optic for safety and noise immmunity
    Firmware can be updated, or processors reset, on all cells at once over comms
    Autonomus operation -- can be used masterless -- optional audible alarm on every cell
    Blue, red and yellow LEDs -- activity, distress and bypass
    Based on Texas Instruments MSP430 microcontroller -- 16 bit ultra low power

This post describes how to obtain the sources.
viewtopic.php?t=980&p=36932#p36932

When we have tested them to our satisfaction in the MX-5, we may offer them for sale, for those who don't want to build their own. At the moment we're guessing we could sell them for around $15 to $20 per cell if we got orders for sufficient quantities.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

cts_casemod
Groupie
Posts: 76
Joined: Fri, 28 Sep 2012, 17:59
Real Name: Carlos
Location: UK - London

Low cost BMS

Post by cts_casemod »

Congratulations guys, that circuit is outstanding! Its about the same price as a normal BMS with only HVC and LVC.

Its also good to see that provisions have been taken to deal with 750VDC. There aren't many out there rated for that.
Nevilleh wrote: Woohoo! 192 linear optocouplers would cost you something like $US800 (HCNR200) plus a dual op-amp (another $1 each)for each one. You'd still have to make a pcb for each cell with at least those parts on it and then feed all the outputs to a vast combination of multiplexor chips. Could be done, cost it all out! I think you'd have to remove the "Low cost" from the forum title though.


Nevilleh
I did part of that circuit today.
It works with normal optocouplers, as long as the transistor is not Saturated the output current is a percentage of the input current driving the IR LED, so this can easily be converter into voltages again with the use of some voltage dividers.

Anyway, no need to reinvent the wheel on this one, I don't see any obvious advantages compared to your circuit.


I have been doing some minor changes on your board for my circuit with headways.

The board is going to be assembled on a panel, so I have redesigned the circuit in a way that both communication ports and power ports for the batteries can be interconnected.
This allows the circuit to be modular, makes things simpler to wire and less prone to noise. I can also use a resistor fuse instead of a wire giving me added safety.

Modified the board to save space.
Removed the Electrolytic capacitor (I am reading a battery so I believe I only need the ceramic one)
Removed two of the SMD resistors
Added holes for normal through hole resistor assembly (For better heat dissipation of the 200Cells)
Added holes for assembly of wires for remote temperature sensor.

This is the first rev, if someone thinks I would be better of changing something please let me know.

I am also looking to implement a battery gauge on it. I have ordered some samples from Texas and maxim. That will be a separate circuit. I will either wire it into the master BMS controller or directly into the car fuel gauge.

Updated Version

CellTopBMSCS1.rar

Image

Here is a panel with 24Boards to be used with another one on top (Total 48Cells 153V)


Image
Last edited by cts_casemod on Mon, 19 Nov 2012, 03:59, edited 1 time in total.

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Low cost BMS

Post by Nevilleh »

Been a while and I didn't see your last post until just now. You are right in that the output current of an optocoupler is proportional to the diode current, but when you see CTR figures like 100 - 300 % you realise that each one has to be calibrated. And then when you find that they drift all over the place when the temperature changes, you begin to understand why the manufacturers go to the trouble of fitting another emitter and use some complicated circuitry to stabilise the things when they advertise them as "linear". But never mind.
Glad to see you have done some work to produce a new pcb layout to fit your Headway cells. Mine was only done for prismatic ones and the whole point of making it freely available was so that people like you could adapt if for their own application. Keep up the good work! By now you may have made some and tried it out?

cts_casemod
Groupie
Posts: 76
Joined: Fri, 28 Sep 2012, 17:59
Real Name: Carlos
Location: UK - London

Low cost BMS

Post by cts_casemod »

Nevilleh wrote: Been a while and I didn't see your last post until just now. You are right in that the output current of an optocoupler is proportional to the diode current, but when you see CTR figures like 100 - 300 % you realise that each one has to be calibrated. And then when you find that they drift all over the place when the temperature changes, you begin to understand why the manufacturers go to the trouble of fitting another emitter and use some complicated circuitry to stabilise the things when they advertise them as "linear". But never mind.
Glad to see you have done some work to produce a new pcb layout to fit your Headway cells. Mine was only done for prismatic ones and the whole point of making it freely available was so that people like you could adapt if for their own application. Keep up the good work! By now you may have made some and tried it out?


Hi neville,
I am quite busy with the end of Semester at uni, so not a lot of free time available.
Thanks for the heads up! Yeah I understand that now, the other day I was discussing at another forum about noise on a similar circuit because the resistor driving the optocoupler was too small and there is a large variation on these (some work fine, some dont!).
I havent touch the circuit yet. I have 96 A123 ordered and I am trying to see how I could combine your circuit with another one Ive found here

Image

BMS

I am also planing how to use a LTC6802 to be able to read a string of 12Cells with each slave board. This would reduce the power consumption, component count and board size.

I will probably come up with a mix of both, if time allows ;)

zeva
Senior Member
Posts: 422
Joined: Sat, 15 Dec 2007, 05:09
Real Name: Ian Hooper
Location: Australia
MSN: sigmunky@hotmail.com
Contact:

Low cost BMS

Post by zeva »

cts_casemod wrote: I am also planing how to use a LTC6802 to be able to read a string of 12Cells with each slave board. This would reduce the power consumption, component count and board size.

I'll be interested to hear how you get on with the LTC6802 chips. Did you get the 6802-2 addressable version?

I read up on them a while ago and would be interested to have a play with one myself. Initially I was hesitant about them because they're 10x more expensive than say an 8-bit microcontroller, but in retrospect I think it's worth it, e.g having power supply and shunt MOSFETs built in pretty much justifies the cost, and would result in an elegantly simple circuit.
Ian Hooper
--
"Never doubt that the work of a small group of thoughtful, committed citizens can change the world." - Margaret Mead
http://www.zeva.com.au

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Low cost BMS

Post by Nevilleh »

Yes, the LTC6802 is an excellent choice - it does everything my modular bms does and handles 12 cells in a central system. BUT that's why I didn't use it! If you want to build a centralised bms and run sense wires to each cell then its the way to go. Even at $42 each = $3.60 per cell compared with the less-than-$2 for a PIC micro.

I originally considered that approach and ruled it out after building a prototype (with an Atmel micro and some op-amps, not the LTC6802)because of the huge bunch of wires needed, all of which are now required to be individually fused. My choice then became for a modular system where each module can bolt directly to the terminals of a prismatic cell and no fuses are needed. Much simpler and safer wiring. Obviously it needs to be re-designed to use with cylindrical cells, but my reservations about all those wires and fuses still stand.

zeva
Senior Member
Posts: 422
Joined: Sat, 15 Dec 2007, 05:09
Real Name: Ian Hooper
Location: Australia
MSN: sigmunky@hotmail.com
Contact:

Low cost BMS

Post by zeva »

Ah yes I heard that individually fused rule in NZ.

For a completely "centralised" BMS, i.e having trace wires from every cell all coming to a single location in the vehicle, I agree with the ruling; there's no safe alternative to individual fusing each of those trace wires, using fuses rated to full pack voltage. I'm not a fan of this degree of centralisation - too much extra wiring, too onerous to make safe.

But on the other hand picture a sealed battery enclosure containing 12 cells and one multi-cell BMS module monitoring them. It would seem safe to use say 50VDC rated fuses (i.e the maximum potential of cells within the enclosure), since there would be no exposure to potentials outside. They're cheap and readily available. It's a similar amount of wiring to fully distributed systems (having a short trace wire from each cell, instead of a signal wire between each BMS module), and potentially much cheaper having a single PCB instead of 12. Higher voltage enclosures may just involve a small degree of compartmentalisation, e.g groups of 12 cells with an insulating layer between (not too difficult).

From what I've seen, this sort of "semi-distributed" system is the route most automakers are taking, and I imagine what was intended for chips like the LTC6802.

Anyway just my 2c on the subject. Incidentally the LTC6802s are down around $12 in quantity now. Add the cost of the PCB and a smattering of other components and you'd have a BMS down around $3/cell:

http://www.digikey.com/product-detail/e ... ND/2036846

(Edit: Fixed URL)
Last edited by zeva on Tue, 18 Dec 2012, 07:00, edited 1 time in total.
Ian Hooper
--
"Never doubt that the work of a small group of thoughtful, committed citizens can change the world." - Margaret Mead
http://www.zeva.com.au

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Low cost BMS

Post by Nevilleh »

That sounds like a reasonable scheme - 12 cells per battery box, I mean. Of course, you would need to design the system with that in mind at the start, but that shouldn't be too hard.
The price you quoted was 100-up and they are still $US22 each in less than 25 qty which is what the home builder would be buying. Also, that 4-bit addressing scheme looks a bit complicated to implement with "digital isolators" everywhere and the on-chip balancing fet can't take much current - did I see 20R "on" resistance in the spec? Hence the need for a chip temp sensor.
Its an interesting device all the same, but I like my modular system better - for now anyway.

cts_casemod
Groupie
Posts: 76
Joined: Fri, 28 Sep 2012, 17:59
Real Name: Carlos
Location: UK - London

Low cost BMS

Post by cts_casemod »

Most likely I will use the 6803. Its the same as the 02, but the chip may get the power from the master, so when the car is off there is no power consumption from the pack.

I will be linking a battery gauge to have state of charge on each of the 4 modules I will be using. This way the charger may "know" how much left to top up and if there is any difference between the expected and the actual capacity, in case a cell fails I will know before this will turn into a problem.

Having 200Cells is a nightmare. Good point with the fuse, you also need to use a zenner in case a cell opens. The zenner will make the fuse link blow and the BMS will be disabled. Ive seen cells going open, burning the controller and letting the balance resistor with more than 10V across it.

On my system I will run 4 modules with 48Cells each in sealed boxes. the plan is for each one to have BMS, fuel gauge, low voltage buzzer and cut out relays. The master just displays the information

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Low cost BMS

Post by Nevilleh »

Another fellow (Arber) who has built my bms is dead set on having top balancing and suggests that turning on shunts for cells that get to, and above, say 3.5V while charging might go some way towards doing this.
My original scheme was to turn on shunts for cells that reach 3.6 V (the max charge volts)and then cut the charger back to 1/2 A so the low ones continue with that low charge rate, but that requires a special charger. If you just have a good old 10 or 15 A thing, then his suggestion might be good. That would cut the charge rate from say 15 A to 14.5A for cells with the shunt ON so they approach Vmax more slowly than the lower cells. The charger would still be cut off totally when a cell gets to Vmax, but that can be done with a 230V relay driven by the charger control signal so you still have an unmodified charger.
A bit of fiddling with the actual voltage where the shunts turn on might make it come into balance really well.
Any opinions?

User avatar
Johny
Senior Member
Posts: 3729
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Low cost BMS

Post by Johny »

I think it would work but you would not get good balance straight away. I think it would take many cycles to gradually get the pack in balance.

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Low cost BMS

Post by Nevilleh »

Yes, of course, but if the battery was nearly balanced to start off with, the balancing would be done as the charge is completed rather than as an extra thing to do after. Its a simple s/w mod, I think I'll try it out. Watch this space!

antiscab
Senior Member
Posts: 2499
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Low cost BMS

Post by antiscab »

Nevilleh wrote:If you just have a good old 10 or 15 A thing, then his suggestion might be good. That would cut the charge rate from say 15 A to 14.5A for cells with the shunt ON so they approach Vmax more slowly than the lower cells. The charger would still be cut off totally when a cell gets to Vmax, but that can be done with a 230V relay driven by the charger control signal so you still have an unmodified charger.


This is how the BMS on all my lithium packs works. It's probably the best from a battery service life perspective as you are not holding the cells at a relatively high voltage for nearly as long.

My charger takes the battery to an average of 3.65v (42 cells so 153.3v), but balancing happens when a cell reaches 3.5v

The elcon/TC Charger charger's with the enable option can be throttled if you wanted to slow down the charge rate.

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Low cost BMS

Post by Nevilleh »

I modified my bms to do top balancing and I've been running it a few times to see what happens. I've set the shunts to come on at 3.4 V and the charger to shut off at 3.6 V and I've cycled it four times now. Interesting that the charger reports that it has delivered about 1.6 AHrs each time and I'm now finding that 6 cells are going into shunting - that leaves 39 that aren't! When the charger shuts off, the lowest cell is still at 3.36 V, so there is a way to go yet before they are balanced. Not surprising as they have been bottom balanced all along.
I can't cut the charger back below 10 A so I tried turning on all the dc loads so the inverter sucks a fair bit - the idea being that reduces the charge current while the shunts still bypass 1/2 amp so the balancing should happen more quickly ie the % difference is bigger. Seems to be working.
Last edited by Nevilleh on Wed, 02 Jan 2013, 06:19, edited 1 time in total.

Nevilleh
Senior Member
Posts: 773
Joined: Thu, 15 Jan 2009, 18:09
Real Name: Neville Harlick
Location: Tauranga NZ

Low cost BMS

Post by Nevilleh »

The bms outputs cell voltage data and I have started to collect and analyse it with an interesting result: Cell 9 has a higher internal resistance than the others, giving a greater voltage drop under load. I have not, as yet, checked the terminals or links - could be a slightly high resistance connection - but it does show a potential problem.
Well, I was going to upload a .jpg of the graphs, but the system said it failed the security scan! "May contain malicious code". So you don't get to see it. I didn't know .jpg files could contain executable code.
Anyway, the curves show all (45) of the cell voltages over a short drive and 44 of the cells are all within a band of 25 mV and track each other nicely. Cell 9 is within that band except when under higher load ie accelerating with a 500 A current draw - when it drops around 100 mV below the others.

User avatar
coulomb
Site Admin
Posts: 3833
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Low cost BMS

Post by coulomb »

Nevilleh wrote: I didn't know .jpg files could contain executable code.
They don't normally, but they can attempt to overflow buffers in commonly used browsers and gain control that way.

Perhaps you could convert to a gif, and try again? Gif is probably more suitable for computer generated images like graphs anyway.
Last edited by coulomb on Sat, 19 Jan 2013, 02:28, 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.

Post Reply