Power in Flux
Likes Likes:  7
Page 1 of 7 123 ... LastLast
Results 1 to 10 of 65

Thread: Interest in a cell-level monitoring system?

              
   
   
  1. #1
    Seņor Member podolefsky's Avatar
    Join Date
    Oct 2010
    Location
    Boulder, CO
    Posts
    3,890
    Post Thanks / Like

    Interest in a cell-level monitoring system?

    I've been batting this idea around after a couple recent threads. Seems like people have their different opinions on whether you need balancing, and a few folks are OK with just pack voltage, but most people would like to at least have cell level monitoring. I'm wondering if there would be interest in a monitoring and protection system, a BMS without balancing, if you will. It would be small, simple, and relatively cheap.

    I've done my best Google-jitsu and only found a few products like this, and they all seem to be defunct or not exactly what I have in mind (either complicated or pricey or clunky). If there's something out there already, I'd love to know about it.

    I know you can kind of do this with Celllogs, but they're pretty low on features and you only have one alarm output.

    No idea if this will come of anything, just putting feelers out. I got some feedback from Frodus already, and a basic feature set would be:

    - monitor all cells for high voltage and low voltage
    - digital output to control the charger and controller, either directly or via a relay
    - powered in such a way that it discharges the entire pack equally, preferably 12V isolated (so you don't have it running off of the first 4 cells or something)
    - analog output to control charge and discharge current (for chargers / controllers that can use it)
    - current sensor and SOC calculation, along with some way of outputting SOC (like an analog signal to drive a display, or it's own display) - this could be optional, keep it cheaper by making the current sensor an add on
    - alarm output, to hook up to warning lights or buzzers

    Would be nice:
    - CAN bus or serial comm
    - outputs to drive displays for voltage and current (like analog 0-5V)
    - temperature sensing

    I'm thinking it would be centralized. With no balancing resistors, the unit could be really small. The one thing that irritates me about my distributed BMS is that if a board goes, I have to take the whole pack apart to fix it...and they're kind of fragile little things.

    It would cost more than a few Celllogs (which are about $45 for a 24s pack), but way less than some of the $400-500 BMS out there.

    Let me know what you think.
    Last edited by podolefsky; 08 March 2014 at 0941.
    - Noah Podolefsky -
    The GSX-E

  2. #2
    Member jsantala's Avatar
    Join Date
    Feb 2014
    Location
    Pori, Finland
    Posts
    38
    Post Thanks / Like
    I think it would be really nice to have a properly done BMS (as in Battery MONITORING System) that you could count on not to upset your carefully done hand balancing.

    Some thoughts:

    - Most important feature would be to be absolutely certain of equal cell load. Preferably the system would work off of separate 12V and only measure the cells.

    - For a centralized system you'd need to have a fuse at each wire before it leaves the battery box (or boxes), preferably as close to the cell as possible. So the monitoring wires should have a very small, preferably replaceable fuse at the cell end of each wire. Otherwise you will have pack voltage without fuses flopping about. Alternatively each wire could have a "Tesla-style" built in fuse, that would melt the very end of the wire very quickly and the system would come with a bunch of extra wires. Of course it could also simply use such a thin wire that the whole wire would be a fuse in itself.

    - Instead of monitoring each cell individually you could have the option of monitoring a bunch of cells. Like in my case I have 25 cells, so I could easily live with splitting the pack into five sets and monitor just each of those five. If any of them differ, or combined voltage is too low of too high, you can pretty much draw the same conclusions as if you were monitoring each cell individually. It would make things much cleaner with only six wires needed. Even less for even packs.

    - In any case the system should monitor each cell or set of cells continuously to detect anomalies while discharging. In a balanced pack all cells should be at the same voltage when you hit that throttle, so if a cell or set of cells starts to have different readings you can be pretty sure there's something going on.
    Last edited by jsantala; 08 March 2014 at 0427.
    Blog at kWsaki dot com. Converted a Kawasaki GPX750R -87 and working on a Citroën Xsara -97 5-door liftback.

  3. #3
    Seņor Member podolefsky's Avatar
    Join Date
    Oct 2010
    Location
    Boulder, CO
    Posts
    3,890
    Post Thanks / Like
    Cool, thanks for the input.
    - Noah Podolefsky -
    The GSX-E

  4. #4
    Senior Member
    Join Date
    Aug 2010
    Location
    Pacifica, California
    Posts
    2,886
    Post Thanks / Like
    Speaking of cell monitoring, can someone tell me what is the cell-balancing goal for a BMS? My Zero seems to like to balance its cells at either 3mV or 2mV, which seems pretty good to me. So, for the best battery life and performance, what is a good range for cell balance within a pack of Lithium cells?
    Richard - Current bikes: 2018 16.6 kWh Zero S, 2016 BMW R1200RS, 2011 Royal Enfield 500, 2009 BMW F650GS, 2005 Triumph T-100 Bonneville, 2002 Yamaha FZ1 (FZS1000N) and a 1978 Honda Kick 'N Go Senior.

  5. #5
    Seņor Member podolefsky's Avatar
    Join Date
    Oct 2010
    Location
    Boulder, CO
    Posts
    3,890
    Post Thanks / Like
    10 mV is pretty standard.
    - Noah Podolefsky -
    The GSX-E

  6. Likes Richard230 liked this post
  7. #6
    Member Barron's Avatar
    Join Date
    Oct 2010
    Location
    Mesa, AZ, USA
    Posts
    32
    Post Thanks / Like
    I had a basic cell level monitoring requirement too, so put MAX11081 circuit together for 36V but simply daisy chained for higher voltages.
    LVC and HVC circuits are opto-isolated and HVC circuit disconnects charger from battery.
    Not much in the way of parts so it was even practical just to perfboard the circuitry.

    http://www.endless-sphere.com/forums...barron#p849127

    It doesn't do any of the nice to have on your list but I think it more or less covers the must have. It will use power from the first cell, but only during HVC or LVC events and, from what I can tell on the MAX11081 datasheet, during initial power up until it establishes its own internal 3V regulator.
    Last edited by Barron; 08 March 2014 at 0905.

  8. #7
    Seņor Member podolefsky's Avatar
    Join Date
    Oct 2010
    Location
    Boulder, CO
    Posts
    3,890
    Post Thanks / Like
    I added temperature sensing to the "would be nice" list.
    - Noah Podolefsky -
    The GSX-E

  9. #8
    Seņor Member podolefsky's Avatar
    Join Date
    Oct 2010
    Location
    Boulder, CO
    Posts
    3,890
    Post Thanks / Like
    Barron, that's a really cool project with the MAX chip. You could add a microcontroller and get the rest of the wish list. Nice that the interface is so simple.

    The chip I've had my eye on is the Linear LTC6804. It's a pretty full featured BMS on a chip, with HVC, LVC, analog and digital I/O, serial comm, etc. - so you could measure a current sensor with it, control other components. Interface with a microcontroller and you've got the possibility of user adjustable settings, output to LCD, SOC calculation.

    It also has balancing, but you could just leave that off.

    It's 16-bit, so very high resolution. Robo pointed out to me that many microcontrollers, like what's on the Arduino, are 10-bit. That's right on the edge of what you would want to have 10mV precision.

    Lots of options out there. Main point is, it's not incredibly hard for someone that knows their way around chips. Mostly a question of whether it's something enough people would want to put in the effort producing it.
    - Noah Podolefsky -
    The GSX-E

  10. #9
    I should be working! furyphoto's Avatar
    Join Date
    Jan 2012
    Location
    Harrop, BC, Canada
    Posts
    336
    Post Thanks / Like

    My Social Networking

    Add furyphoto on Facebook
    I would be really interested. My thought on a pack has kind of always been to bottom balance to protect the cells if I ever run out of juice, and monitor at the (paralleled) cell level. That way you get redundancy at the top and bottom end of the pack. On the bottom, BmS shut off for low voltage, and bottom balanced cells so they can't eat each other. At the top, pack voltage cut off on the charger, and BmS high voltage cut off to terminate charging (whichever comes first).

    I was always interested in something simple like "Methods" BMS from Endless sphere it was a simple HV/LV signal that could be used to shut down the pack or stop charging. It was designed for RC-Li-po packs and pre-programmed for RC-Lipo cut voltages. It was quite slick, he made it plug and play using the rc lipo balance leads. I don't think he's making them anymore.

    I have always wanted a current counter to drive an analog fuel gauge for SOC, I even rounded up a breadboard, and arduino to try and figure it out, although admittedly I haven got to it yet. I think that would be a great add on.

    I would like to see the ability to set the HV/LV cut voltages, and potentially some other alarm voltages for "almost empty" I think having the ability to adjust voltages makes it a more universal system, and provides the ability to personalize, so I can use it on a bottom balanced system, where I want to undercharge a bit, but someone else who might want a different LV cutoff could use it on a top balanced system.

    I'm a big fan of having the BmS isolated form the pack voltage as far as powering the system. No parasitic load please.


    You'll notice I have coined the contraction BmS vs BMS to distinguish between Battery monitoring System, and Battery Management System because I hate not knowing what people are talking about. Feel free to continue to use that
    -Andrew

    http://www.andrewdoran.com
    mail(at)andrewdoran.com

    My ElMoto Project "Electric Hurricane" - 1987 Honda CBR600 F1: Check out my Build ALBUM
    My ICE Cafe Racer Project "My Precious" - 1983 Honda CM400 Classic

  11. #10
    Seņor Member podolefsky's Avatar
    Join Date
    Oct 2010
    Location
    Boulder, CO
    Posts
    3,890
    Post Thanks / Like
    Cool, found Methods BMS (big M, since it has balancing - 2A at that, which is about 10x what most other balancing circuits do - seems extreme).

    http://endless-sphere.com/forums/vie...6602&start=315

    Looks like it's using the LTC chip I was talking about.

    I'll use BmS. Question remains, is the plural BmS or BmSs?
    - Noah Podolefsky -
    The GSX-E

Page 1 of 7 123 ... LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •