TMC is an independent, primarily volunteer organization that relies on ad revenue to cover its operating costs. Please consider whitelisting TMC on your ad blocker and becoming a Supporting Member. For more info: Support TMC
Start a Discussionhttps://teslamotorsclub.com/tmc/tags/

FYI: Teardown of busted UMC

Discussion in 'Technical' started by dave4444, Sep 30, 2017.

  1. dave4444

    dave4444 Member

    Joined:
    Aug 12, 2017
    Messages:
    25
    Location:
    NH USA
    I had a busted UMC lying around that I salvaged the cord from for another project.

    Left with a useless UMC box, I decided to open it up and take a look around. This particular UMC met its demise by having the cord forcibly pulled out of the UMC box while actively charging.

    20170930114653_DSC03994a.jpg

    The metal and rubber outer shell was easy enough to get off, however the assembled plastic box appears to have been surrounded by an injected plastic that sealed the inner box in a cover of melted plastic sealing all the gaps and entombing the inner box.

    20170930114724_DSC03996a.jpg

    Suffice it to say, the inner top and bottom parts of the plastic box were welded or glued very well, I had to drill out one entire side and then pry around the rest. Luckily the inside was not injected with plastic like others have found in the handle and other parts.

    20170930125419_DSC03997.jpg

    A quick look reveals that as the cord was pulled out it bent the high voltage output connections driving one of them into the logic board where it intersected a ground plane (among other things). The magic smoke was released in the process. Also note (not shown) that in the process, the plastic around one of the pins inside the box-wall cable to NEMA interconnection melted slightly, undoubitly due to the ground short inside the UMC box.

    20170930130250_DSC04006a.jpg

    All of the AC wires (and ground) are crimped onto copper plates before screwing into the power board. Makes a good connection but it was also the point of failure.

    20170930132433_DSC04012a.jpg

    There are 3 circuit boards. The power board, the logic board, and the LED board. The logic board's 3 pin connector has the purple sense pin from the NEMA adaptors at the top, the blue power output to the handle's RF module in the middle, and the purple pilot pin to the handle at the bottom. The latter two of mine were obviously pulled out along with the rest of the cord.

    20170930131040_DSC04010a.jpg

    logic board removed:

    20170930131420_DSC04011a.jpg

    Both boards removed. The AC inputs to the right, relay and GFCI CT at the left. The power and logic boards are interconnected with a 12 pin header. Both boards are covered with a thin layer of clear epoxy.

    20170930132433_DSC04012b.jpg

    back:

    20170930132624_DSC04013a.jpg

    Side, interestingly the AC relay (Tyco T92S7D12-24) is listed for only 30A at 277V. The datasheet however does show this model can be used at 40A at upto 277VAC but only for resistive loads and at a much reduced lifetime (only 6000 cycles).

    20170930132918_DSC04014a.jpg
     
    • Informative x 22
    • Like x 2
  2. mongo

    mongo Member

    Joined:
    May 3, 2017
    Messages:
    367
    Location:
    Michigan
    General background info for anyone interested

    Relay limits are bounded by a few factors, one is the type of load being switched. Capacitive loads have high current when the relay energizes, inductive loads can arc when the relay opens, and resistive loads are the kindest to the contacts. In this system, the relay has an easy job since the vehicle's charger doesn't start out trying to pull power, that occurs after the relay is providing voltage to it and the pilot signal indicated allowable current draw.

    The steady state capability of a relay is partly limited by heat due to contact resistance. If the coil drive is modulated after activation, its self heating is reduced and can increase the relay capacity.

    Edit: Thanks for the write up of the tear down!
     
    • Informative x 12
    • Helpful x 1
  3. dave4444

    dave4444 Member

    Joined:
    Aug 12, 2017
    Messages:
    25
    Location:
    NH USA
    Ya, agree, under normal conditions the vehicle is drawing no load when the relay is turned on (and usually off) unlike say an A/C condensor, fan motor, etc... which would use a relay like this to actually act as a switch drawing full load at the time of contact.
     
    • Informative x 1
  4. arnis

    arnis Member

    Joined:
    Apr 13, 2015
    Messages:
    326
    Location:
    Estonia
    Specs are for what relays are usually used for - switch under load.
    Relays inside EVSEs (incl UMC) are switched at 0A load. First, charger stops current draw, then relay disengages and lastly socket disengages plug locking mechanism.
    With Type1 connector (J1772) pushing the lever stops the current draw, then while plug loses connection with pins EVSE disengages relay (also 0A load). Due to that, EVSE relays will last for very long time.
     
    • Informative x 3
  5. mongo

    mongo Member

    Joined:
    May 3, 2017
    Messages:
    367
    Location:
    Michigan
    Exactly. Although there is one edge case that will cause switching under load. If the internal ground fault circuit detects a current imbalance, it will de-energize the relay during operation. Hopefully, no one has that happening often.
     
  6. scaesare

    scaesare Well-Known Member

    Joined:
    Mar 14, 2013
    Messages:
    5,116
    Location:
    NoVA
    Thanks for posting these...

    Minor point of clarification: I don't believe those cables are crimped to the lugs, but rather ultrasonically bonded. Those types of welded connections are typically superior to crimps for high-current connections.
     
  7. hcsharp

    hcsharp Active Member

    Joined:
    Jun 7, 2011
    Messages:
    2,776
    Location:
    Vermont
    Generally the welded connections are better than crimped but I've been surprised at how many UMCs I've seen where the welds fell apart. I've also been surprised at how many of the relays have worn out despite not switching under load. IMO they need better strain relief and more rubust relays/contactors. But I have to admit the price is right.
     
  8. scaesare

    scaesare Well-Known Member

    Joined:
    Mar 14, 2013
    Messages:
    5,116
    Location:
    NoVA
    I seem to recall there was an issue with substandard welds early on as well...
     
  9. arnis

    arnis Member

    Joined:
    Apr 13, 2015
    Messages:
    326
    Location:
    Estonia
    Maybe this is why Model 3 UMC's are limited to 32A. This would actually help with reliability if they can't get them perfect cheaply.
     
  10. FlatSix911

    FlatSix911 918 Hybrid

    Joined:
    Jun 15, 2015
    Messages:
    3,903
    Location:
    Silicon Valley
    Great write up :cool: Was this UMC energized and attached to a moving vehicle when it failed?
     
  11. Plugsuvohio

    Plugsuvohio Member

    Joined:
    Jul 28, 2016
    Messages:
    91
    Location:
    Cleveland ohio
    This is really interesting info for me. I work with a company manufacturing heavy duty GFI devices. Curious. What approval ratings does this UMC show for Europe? There is a very difficult 4000 volt insulation test required there I believe
     
  12. dave4444

    dave4444 Member

    Joined:
    Aug 12, 2017
    Messages:
    25
    Location:
    NH USA
    This UMC is the 1 or 2 phase US/Canada version. The 3 phase EU version is completely different.
     
  13. Bill25cycle

    Bill25cycle Member

    Joined:
    Mar 31, 2011
    Messages:
    6
    Very interesting - I knew the UMC for my Roadster was highly unreliable, - I never used one but it was common for people to go through 3 or 4 of them.... The main problem with the S UMC was the Nema 14-50P adapter doesn't meet NEMA standards for current density and was designed by an incompetent. After a few fires, Tesla deflected saying there was no problem with the car, but wisely first limited charging current to 32 amperes if there was too much pressure drop, and much more importantly - installed a fusible link to shut off the juice if a fire was imminent. Pushing the rating of a 30 amp relay up to 40 amps seems par for the course, - of course, 6000 operations is around 15 years of use, and it doesn't look like the rest of the cheapie ckt boards would last that long anyway.

    Not picking on them for making such a cheap device, but in the Roadster's case $1500 was a lot of happiness. These are much cheaper but not as cheap as they should be seeing as how crappy the parts are. I'm wondering how often the typical "S" owner has to replace these things if they charge at the full 40 amps?
     
  14. xrayangiodoc

    xrayangiodoc Member

    Joined:
    Apr 1, 2014
    Messages:
    262
    Location:
    Amherst, NY
    I have used the NEMA 14-50 outlet in my son's garage on multiple occasions with no apparent issues. I've had my UMC for 3 years 6 months. Of course, I always make it a point to avoid excessive bending or pulling of the cables.
     
  15. Plugsuvohio

    Plugsuvohio Member

    Joined:
    Jul 28, 2016
    Messages:
    91
    Location:
    Cleveland ohio
    Actually, generally with the 3 Phase in Europe, only 1 phase is used at 230V. Power distribution at a home is usually single phase and the distribution generally alternates between the 3 phases from home to home in an effort to level loads. But 380V can be hooked up also with that they call the RED plug and those wires brought up the cable to an EVSE even though not used. Because of that the regulations push the insulation protection levels to 4000V. But I'm looking to confirm all that with study of some European EVSE specs. I assume the Tesla is shipped with an appropriate UMC for that area.
     

Share This Page