Results 1 to 10 of 10
  1. #1
    Join Date
    Jan 2014
    Posts
    5
    Post Likes

    Confused GFCI/RCD Main Breaker Tripping When Outside Inverter Compressor Starts

    I have a brand new LG LSU240HSV3 2 ton mini-split 20 SEER inverter unit that was just installed. When the unit is powered up using the remote control, the indoor unit fan turns on and everything is fine. There is a normal delay for this system before the outside inverter-driven compressor turns on. The issue is that within a second or two after the outside unit is commanded by the control logic to fire up, the main 50 AMP GFCI breaker (in main load center) feeding the 100 amp load center in the garage trips. The 20A breaker in the garage load center feeding the mini-split does not trip so this isn't an overcurrent trip but an upstream RCD/GFCI fault trip. I've replaced the upstream 50A General Electric RCD/GFCI circuit breaker w. a normal 2-pole 50A breaker, and the unit fires up and runs happily. All of the wiring is clean; all grounds are new and tight.

    The garage load center fed from the main panel DOES NOT have its earth and neutrals connected as per NEC. A 5 HP 240 VAC air compressor fed from the same panel DOES NOT trip the same upstream GFCI breaker.

    Any ideas why the GE 50A RCD/GFCI breaker is detecting a current imbalance in the 2 hot conductors feeding the mini-split system, and only when the inverter fires the compressor motor (and condenser fan)?

    I imagine (for a test only) that I can lift the earth ground wire at the outside condenser and see if the upstream RCD still trips. I am wondering if the inverter circuit board has suppressors connected from the inverter output to earth ground and this is causing more than 5 milliamps of hot to earth leakage current when the compressor drive is ON.
    Any ideas? I found one hit on Google w. someone else having an identical problem, but no solutions were given.

  2. #2
    Join Date
    Jan 2014
    Posts
    5
    Post Likes
    Thread Starter

    Exclamation This is apparently an known issue w. inverter driven units

    I believe that I've found my answer.

    In this case unfortunately:

    Removed link to other forum

    I did bury the 50A underground feeder cabling/conduit to the garage load center 18 inches below grade, so technically a RCD/GFCI breaker is NOT required by NEC (but would have been safer as the cable run AND all downstream devices would be GFCI protected) .

    It sure would be nice if the AC manufacturer (LG) would have put this information in any of the technical installation/engineering specifications documentation, all of which I thoroughly consumed prior to the actual inverter-driven unit's purchase and installation. Does anyone want to buy a 50A GE GFCI breaker for cheap :>) ?

    Also, it would be sweet if the RCD/GFCI breaker's "fault" current trip point were directly "tweakable" from 5 to 100 milliamps of fault current. Especially for a feed breaker like this that is supplying an entire load center as opposed to a human touchable outlet. In that way, I could have raised (hopefully just slightly) the leakage detection threshold of the feed breaker to just above that of the inverter's electronics maximum operational leakage current.

    I will measure the LG unit's operational earth leakage current w. my Fluke 87 true RMS meter when I have a moment and report back on that for those who are curious.

    I hope that this thread may help someone in the future w. the same issue.

    ** Thanks to both the HVAC forum and Google on this one.
    Last edited by jpsmith1cm; 01-12-2014 at 01:21 PM.

  3. #3
    Join Date
    Dec 2005
    Location
    Cincinnati, Oh
    Posts
    6,761
    Post Likes
    It truly amazes me how overcomplicating something results in failure in the HVAC world 100% of the time.
    "Better tell the sandman to stay away, because we're gonna be workin on this one all night."

    "Dude, you need more than 2 wires to a condenser to run a 2 stage heatpump."

    "Just get it done son."

    Dad adjusted

  4. #4
    Join Date
    Jan 2014
    Posts
    5
    Post Likes
    Thread Starter
    *

  5. #5
    Join Date
    Jan 2014
    Posts
    5
    Post Likes
    Thread Starter
    Sorry about the link to another forum... I did not know that was against the rules or why.

    Here was the gist of the other site's comments:

    ************************************************** *************
    Hi this is a really common problem with inverters see below
    In domestic and light commercial buildings it is common to have a device called an RCD residual current device, commonly called a trip switch which is used to protect the electrical circuits and the users of them.

    The RCD works by measuring the current in the Live wire and the Neutral wire and comparing them, in a healthy circuit the currents will be equal. In an unhealthy circuit some current leaks to Earth.

    If there is a difference of more than 30mA the switch trips and everything goes
    Off. This is not a fault with the aircon unit so don’t send out new parts.

    The problem with RCDs is that computers and Inverters all have some earth leakage so in some cases if the total earth leakage exceeds 30mA the RCDS can be nuisance tripped by the inverters.

    If this happens there are two courses of action available to you

    1- replace the RCD with a less sensitive model up to 100mA RCDs are ok in domestic installations
    2- Wiring the Aircon unit so it doesn’t pull current through the RCDs.
    ************************************************** *************

    Also w regard to HVACVEGAS's comment:

    Yes. A single speed compressor and contactor controlled by a dry contact closure at the thermostat is pretty bullet-proof and nice to troubleshoot, but I was dazed and drawn in by the 20 SEER rating of the LG inverter unit. I hope this is the last of my overcomplicated gizmo's issues .

  6. #6
    Join Date
    May 2006
    Location
    Philadelphia, PA
    Posts
    1,878
    Post Likes
    Your professional installer should have known this.
    Climate Control Solutions for your Home or Office

    Serving Northeast Philadelphia and Surrounding Areas

  7. #7
    Join Date
    Jan 2014
    Posts
    5
    Post Likes
    Thread Starter
    This admittedly unprofessional installer who is an electrical engineer specializing in power-electronics did not. I still believe this specific issue is w. the LG inverter's electrical design. There should be no net leakage current to the protective earth ground reference.

    Also, there should be an explicit reference to this in the published engineering spec docs available from the manufacturer.

    I would also question whether a professional installer would have looked at the upstream RCD/GFCI feed breaker for the garage load center before wiring the unit from the garage panel.

    ...could be wrong, and I truly do agree with there being few exceptions to having a lot of real-world field experience.

  8. #8
    Join Date
    Dec 2005
    Location
    Cincinnati, Oh
    Posts
    6,761
    Post Likes
    Quote Originally Posted by yotool View Post
    Also w regard to HVACVEGAS's comment:

    Yes. A single speed compressor and contactor controlled by a dry contact closure at the thermostat is pretty bullet-proof and nice to troubleshoot, but I was dazed and drawn in by the 20 SEER rating of the LG inverter unit. I hope this is the last of my overcomplicated gizmo's issues .
    There's nothing complicated about minisplits.

    I'm talking about this:
    "Also, it would be sweet if the RCD/GFCI breaker's "fault" current trip point were directly "tweakable" from 5 to 100 milliamps of fault current. Especially for a feed breaker like this that is supplying an entire load center as opposed to a human touchable outlet. In that way, I could have raised (hopefully just slightly) the leakage detection threshold of the feed breaker to just above that of the inverter's electronics maximum operational leakage current.

    I will measure the LG unit's operational earth leakage current w. my Fluke 87 true RMS meter when I have a moment and report back on that for those who are curious."

    "Also, there should be an explicit reference to this in the published engineering spec docs available from the manufacturer."

    "I would also question whether a professional installer would have looked at the upstream RCD/GFCI feed breaker for the garage load center before wiring the unit from the garage panel."

    "electrical engineer specializing in power-electronics did not"


    You overthought. Overthinking is not for the HVAC industry.
    "Better tell the sandman to stay away, because we're gonna be workin on this one all night."

    "Dude, you need more than 2 wires to a condenser to run a 2 stage heatpump."

    "Just get it done son."

    Dad adjusted

  9. #9
    Join Date
    Aug 2009
    Location
    Maine
    Posts
    8,077
    Post Likes
    So is this just an LG problem, or is it common with all HP's

  10. #10
    Join Date
    May 2006
    Location
    Philadelphia, PA
    Posts
    1,878
    Post Likes
    Quote Originally Posted by lytning View Post
    So is this just an LG problem, or is it common with all HP's
    Many furnaces have issues with gfci devices also. I'm no electrical engineer or anything , but I think I read it has something to do with induced currents. The why doesn't matter much. When experience says it doesn't work, and it is not required by code, to me it doesn't matter why.

    Inverters rarely play well with gfci devices.
    Climate Control Solutions for your Home or Office

    Serving Northeast Philadelphia and Surrounding Areas

Posting Permissions

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

Related Forums

Plumbing Talks | Contractor MagazineThe place where Electrical professionals meet.