Welcome to 3DRPilots.com

Sign up and join the largest 3D Robotics drone community!

  1. pious_greek

    Joined:
    Nov 15, 2016
    Messages:
    180
    Likes Received:
    98
    Location:
    Fresno, CA
    Battery was defective and I got to keep it without having to return it so figured I'd tear it down. This has probably been done before but couldn't find it online. It looks like maybe a capacitor went bad, but I'm not sure.

    With no intention of reassembling, my method was unnecessarily destructive.... With batteries as cheap as they currently are, and given that this one was nonfunctional and free of cost, I had no reservation doing this....

    20170107_142523.jpg
     

    Attached Files:

    #1 pious_greek, Jan 7, 2017
    Last edited: Jan 16, 2017
  2. pious_greek

    Joined:
    Nov 15, 2016
    Messages:
    180
    Likes Received:
    98
    Location:
    Fresno, CA
    additional pictures
     

    Attached Files:

    taos and VCD like this.
  3. RichWest

    Joined:
    Jun 18, 2015
    Messages:
    2,095
    Likes Received:
    1,017
    Location:
    Chouston, Tejas
    Great set of photos, thanks!
     
  4. Don H.

    Joined:
    Jun 14, 2016
    Messages:
    432
    Likes Received:
    72
    Looks like a puffy pack one the first pick of the second set.
     
  5. VCD

    VCD

    Joined:
    Nov 7, 2016
    Messages:
    52
    Likes Received:
    9
    Good to finally see what's inside, could you please post some close-ups on both sides of the main PCB (longest) - and the LED/indicator/power button PCB)
    I'd like to identify the components, estimate the balancing current it's using and more..

    Is your impression that it is glued together, or do you see some click-mechanism that keeps it together ?
     
  6. Frank Foster

    Joined:
    Jul 15, 2015
    Messages:
    216
    Likes Received:
    55
    Thank you! I too had been wondering what was inside. Now I see why they cost so much! Makes me wonder if the reason a lot of the companies have gone to proprietary type batteries, is because of the hazards of lipo's and improper charging/storage. ???
     
  7. pious_greek

    Joined:
    Nov 15, 2016
    Messages:
    180
    Likes Received:
    98
    Location:
    Fresno, CA
    The two halves are bonded together. I believe there are also snap or tabs to help align the halves probably during bonding.

    At first, I attempted to methodically separate the glue joint with a razor blade, but I have limited patience for some types of operations.
     

    Attached Files:

  8. pious_greek

    Joined:
    Nov 15, 2016
    Messages:
    180
    Likes Received:
    98
    Location:
    Fresno, CA
    I think proprietary batteries are just another avenue companies use to make money once you're locked into their ecosystem; but you're right, there is a lot of R&D that goes into it, and a lot of times they're designing batteries for specific needs: not just power uses but volume/shape/weight constraints as well.
     
  9. VCD

    VCD

    Joined:
    Nov 7, 2016
    Messages:
    52
    Likes Received:
    9
    most - like DJI , do it to be able to charge whatever they like.

    3DR has no "verification" of the power source, ArduCopter is not even checking the serial - it's there purely because it the *BEST* , and maybe "only way it should be in 2016"
    A battery with internal current & voltage monitoring, can automatically tell the charger exactly what it's settings/limits are, but most importantly, it "knows" always it's true capacity, and when one of these gets old, the Solo does not only need to rely on voltage - or theoretical capacity - to calculate safe RTH, an old battery will report it's actual(reduced) capacity. - know it's internal resistance, temperature if going bad.
     
  10. VCD

    VCD

    Joined:
    Nov 7, 2016
    Messages:
    52
    Likes Received:
    9
    Thanks for the additional pictures.
    It seems like there is a low side deep discharge protection there too ! (it's a really neat job, going the little extra length , unusual in flying things, but not a problem when set low enough)
    - I would love to analyze it, and its trigging threshold.
    - I am also very curious whatever the current/voltage/cell monitoring is calibrated (if the SMBus writes to flash to fine-calibrate those)
    -finally, does it have active discharging when left idle - the balancer part could do it - yet nobody confirmed it.

    If anyone has such set of electronics, will throw it away - and don't mind to drop it in an envelope to Norway, please tell :)
     
  11. pious_greek

    Joined:
    Nov 15, 2016
    Messages:
    180
    Likes Received:
    98
    Location:
    Fresno, CA

    VCD, Where do we start with the testing (assuming it's fully functional)?
    20170115_184111.jpg
     
    #11 pious_greek, Jan 15, 2017
    Last edited: Jan 15, 2017
  12. VCD

    VCD

    Joined:
    Nov 7, 2016
    Messages:
    52
    Likes Received:
    9
    Simulate a battery (connect GND, VBatt , and balance to cell voltages.) - depending on how many lab PSU's/outputs you've got , it would be best to use four independent outpult at 4v , like cells.
    -measure reported(I2C) voltage vs actual voltage , & cell
    -apply some load - (100ma is plenty) , reduce vbatt till it shuts down (note when)
    - at normal cell voltage, ramp one up to 4v , let the others be 3.8 ) - check current - we would like to know if it's balancing only during charging, or always.
    - at full voltage, (4x4,2v) - let it be idle for a few days, check for current or look at the PCB's using thermal camera, - see if its discharging to 3v9 or lower.
    - load with ~10A - check current reported vs a proper instrument.

    That's the basic stuff I would check :)

    I would also check the I2C reported values, and see if I find calibration registers. If you are not messing around with that, you can just connect it's SMBUS to Solo and read out voltage/current to verify accuracy.
     
  13. VCD

    VCD

    Joined:
    Nov 7, 2016
    Messages:
    52
    Likes Received:
    9
    Another question is whatever the balancer is active only when the battery is on (long press) , or is it independent
    when on: is it active only during charging, or also when discharging.
     
  14. pious_greek

    Joined:
    Nov 15, 2016
    Messages:
    180
    Likes Received:
    98
    Location:
    Fresno, CA
    Science! Sound like fun. Let me see what I can put together as I have time over the next couple weeks....
     
    #14 pious_greek, Jan 16, 2017
    Last edited: Jan 16, 2017
  15. RichWest

    Joined:
    Jun 18, 2015
    Messages:
    2,095
    Likes Received:
    1,017
    Location:
    Chouston, Tejas
    Curious, did you cut the rear plate? If you did and for the purpose of reassembly, what would be the correct location for the cut?

    I cut my battery case about 6mm up from the edge, around the sides and front. Used a dremel with a saw bit. Other than the depth of cut it was fairly clean, since it was a saw rather than an abrasive cut. It would be easy to rebuild using an adhesive or @PdxSteve 's acetone/plastic paste trick.
     
  16. pious_greek

    Joined:
    Nov 15, 2016
    Messages:
    180
    Likes Received:
    98
    Location:
    Fresno, CA
    Rich, i'm not a hundred percent sure on which side is the rear plate, but....

    If reassembly is a factor, I believe your method would be ideal unless there is a chemical means to debond the adhesive. If i were trying it, I would want to retain in tact as much of the housing as possible especially where it locks/clips into the solo. So my guess is that the ideal situation is to cut the battery as close to the seam and only on the shiny surfacing of the upper housing; none of the matte surfaces.

    I would use a fine bladed rotary tool and cut approximately 6mm (correction per @RichWest ) from the seam; this is probably ideal for the two longer seams as well as the seam nearest the LEDs. For the edge nearest the molex connector, I would cut across the top along the shiny surface a minimum of 9mm from the cantilevered/overhanging edge. The actual seam is through the middle of the face with the Molex connector, above or behind the serial number sticker, but I would be more concerned about retaining the adequacy and functionality of the solo's battery clip and would not cut this seam.

    The next issue that I would face is that the LiPo pack is strongly adhered to a plastic liner (the part that has all the markings depicting solo's capacity, volts and cautionary warnings) that in turn is strongly adhered to the lower plastic housing. The lower plastic housing has a plastic 'rib' dividing Solos capacity/voltage markings and the cautionary warning markings. The plastic liner is strongly adhered to that rib and various other parts of the lower housing. I believe this area allows the LiPos room to expand as they do when they fail or gas off or worse.

    I tried pushing the existing LiPos out by pushing the areas with the markings, but this definitely deforms the LiPos as the adhesive is relatively strong. I had to cut the 'rib' to facilitate removal, but if i did this with the intent of reassembly, I would try to dissolve the adhesive with small dabs of alcohol or perhaps Ronsonol first (being incredibly careful), and if that didn't work i would try cutting the adhesive with a razor or exacto blade.
     
    #16 pious_greek, Jan 17, 2017
    Last edited: Jan 17, 2017
  17. RichWest

    Joined:
    Jun 18, 2015
    Messages:
    2,095
    Likes Received:
    1,017
    Location:
    Chouston, Tejas
    At 6mm from the edge I was able to clear the lower frame's interface to the top shell, just above the glued seam between the lower and upper shell. There are actual catches/hooks on the upper shell that engage to the lower shell. I had about 5 sets remain intacted and there is one under the leds.

    The back plate, at the battery terminals, I just can't tell how far in are the battery cells. I want to cut at the seam, but I'm thinking all the Li-Po juice is going to be spewing on me as I start cutting into it... at least with my luck.

    Previously I'd tried to separate the two shells at their glued seam, unfortunately the plastic deforms when it heats up from the effort. I'd presume the glued area is about 4mm width from the edge, so a fairly good bonding surface.

    Having the basic lower frame intact was following yours and @franknitty69 dissection. As it would need to maintain its intended function, not just for the battery interface, but for the structure of the over all bird's frame. Give the frame a twist without the battery installed...

    Anyway I think I could find a larger saw bit and build a dremel router of sorts. Build a jig to consistently remove the cover and reduce any risk for cutting too deep. But considering the cost for batteries presently, who cares... This is future talk...
     
  18. pious_greek

    Joined:
    Nov 15, 2016
    Messages:
    180
    Likes Received:
    98
    Location:
    Fresno, CA
    You're right, I'll edit my post to correct that (measure twice, post once...) On the Molex face, can you see the rectangular vent hole? It's right under the battery tab indentation. You can see the black battery liner just beyond, and maybe a hint of white battery housing peaking out behind or above the liner. You've probably got about 3mm of clearance from the molex face to the battery, and less from the battery to the upper housing.
     
    RichWest likes this.
  19. dml931

    Joined:
    Jan 1, 2017
    Messages:
    98
    Likes Received:
    33
    Location:
    NE Ohio

    BWAHAHAHAHAHA!!!!!!! I can totally relate here!
     
  20. Maddog

    Joined:
    Aug 24, 2015
    Messages:
    3,433
    Likes Received:
    902
    Location:
    Phoenix, Az
    So, are you guys taking these apart for the electronics or to try to replace the lipo cells? Has anybody even found a suitable replacement cell?