RPC blows adjacent circuits
Close
Login to Your Account
Page 1 of 3 123 LastLast
Results 1 to 20 of 46
  1. #1
    Join Date
    Sep 2002
    Location
    Salinas, CA USA
    Posts
    4,151
    Post Thanks / Like
    Likes (Given)
    266
    Likes (Received)
    232

    Default RPC blows adjacent circuits

    New 15 HP RPC in a new shop with 100A service, everything new. This is attached to a subpanel fed from the original garage subpanel, with #2 copper wires run about 70'. I have two problems I would appeciate some help with.

    First, when the RPC is started, it blows two other circuit breakers (20A GFCI), even though there is no load on those breakers. The RPC is serviced by a 50A GFCI breaker, and that breaker has no problems. All 3 beakers are new Homeline tandem GFIC breakers (HOM 220GFIC and HOM 250GFIC). The RPC is basically a Fitch Williams design, and it evolved from a 5HP RPC that never had any issues.

    Second problem: I added 1500k ohm bleed resistors to each run cap circuit (AC and BC), and both resistors blew after about 10 minutes. They went with a loud bang and really smoked the garage. No other damage was done; after removing the remnants of the resistors and the AB and start cap resistors, the phase converter runs fine. AC voltage was 289 and BC voltage was 287 originally, but I removed one 45mfd cap from each leg and now AC is 269V and BC is 267V. I must have misunderstood the articles I read that recommended bleed resistors on each capacitor leg. Can someone set me straight?

    Dave

  2. #2
    Join Date
    Jul 2007
    Location
    Missoula Mt
    Posts
    345
    Post Thanks / Like
    Likes (Given)
    9
    Likes (Received)
    125

    Default

    A 100 amp service to too small for a 15 hp rpc, Is the wires from the 2 gfi,s in the same raceway as the feed for the rpc? If you blew 1500k(?) resistors you have major voltage spikes ( cant see how that is even possible you sure about the ohms?)...Phil

  3. #3
    Join Date
    Sep 2002
    Location
    Salinas, CA USA
    Posts
    4,151
    Post Thanks / Like
    Likes (Given)
    266
    Likes (Received)
    232

    Default

    The resistors were 15k not 1500k. Looking at Fitch Williams' design, he only put the bleed resistors on the start caps. Maybe there is a reason? The ones that blew were on AC and BC legs. Here is a photo:

    img_0521.jpg

  4. #4
    Join Date
    Sep 2002
    Location
    Salinas, CA USA
    Posts
    4,151
    Post Thanks / Like
    Likes (Given)
    266
    Likes (Received)
    232

    Default

    Quote Originally Posted by Phil in Montana View Post
    A 100 amp service to too small for a 15 hp rpc, Is the wires from the 2 gfi,s in the same raceway as the feed for the rpc? If you blew 1500k(?) resistors you have major voltage spikes ( cant see how that is even possible you sure about the ohms?)...Phil
    The conduit to the RPC runs direct from the breaker panel and is not shared by any other circuit. The breaker panel is rated at 125A and the ampacity of #2 copper is also 120-130A. The 50A breaker is further down towards the bottom of the breaker panel than the 20A breakers.

    When the RPC is started, the start caps are about 900mfd and the run caps are 180mfd on AC and 135mfd on BC.

  5. #5
    Join Date
    Nov 2018
    Country
    ALAND ISLANDS
    Posts
    76
    Post Thanks / Like
    Likes (Given)
    75
    Likes (Received)
    38

    Default

    I would say your light on starting capacitance. 900/15 = 60 mfd/hp i would add more so your start time is quicker, less amp draw my testing has seen 100- 120 mfd per hp works good, thy are disconnecting once started? I could run a 20 hp idler with 100 amp fused feed, but would trip breakers when starting. 100 amp was the service
    All the resistors were doing was bleeding down the start caps so they wouldn’t keep a charge as long. Shouldnt need them in there for it to work. If it don’t start within a a couple seconds something is not quite right yet.


    Sent from my iPhone using Tapatalk

  6. #6
    Join Date
    Jun 2001
    Location
    St Louis
    Posts
    18,415
    Post Thanks / Like
    Likes (Given)
    2040
    Likes (Received)
    3216

    Default

    Homeline is not the good stuff.

    The dual breakers pop when you frown hard at them, I do not like them.

    Then also, there is a fair chance of counterfeits. If they are not genuine, Lord only knows what if anything they have inside.

    15k is small, and your voltage is high. 5 watts per resistor. What is the rating?

  7. #7
    Join Date
    Sep 2009
    Location
    Apex, NC
    Posts
    1,535
    Post Thanks / Like
    Likes (Given)
    1
    Likes (Received)
    758

    Default

    For what it's worth, empirically, I have had a "15 HP American Rotary" converter for 10 years with no problems. It is fed from a 30 amp breaker (not GFCI), 200 amp panel. It barely dims the lights when switched on, and not at all when starting a 3 or 5 hp motor.

    GFCI breakers can be flaky to say the least;(I have one single phase that seems to trip randomly when there is a decent rain or thunderstorm). My understanding of codes (caveat here that I am not a code-expert by any means) are that GFCI's are not required in "finished spaces" above-grade having concrete floors. I guess there is some logic that if the space is "unfinished" then there is more possibility people are using tools and extension cords that might not be in the best shape, or would be dragging them outside. Most slab construction does not have every outlet GFCI-protected, such as my mother's house in SC with various tile floors, commercial buildings I have rented, etc). You might consider nixing the GFCI's.

  8. #8
    Join Date
    Feb 2004
    Location
    peekskill, NY
    Posts
    25,296
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    5145

    Default

    Quote Originally Posted by rimcanyon View Post
    ...
    First, when the RPC is started, it blows two other circuit breakers ...
    Dave
    I would think the key to solving this problem, is inside this question: do the two other branch circuit breakers trip from overcurrent, or does the GFI function trip off. Can you
    tell the difference between the two events, with this brand of breaker?

  9. #9
    Join Date
    Jun 2001
    Location
    St Louis
    Posts
    18,415
    Post Thanks / Like
    Likes (Given)
    2040
    Likes (Received)
    3216

    Default

    Since there is no load on the other two, and no load change on them from the RPC starting in any case, it seems that they have to be opening on the GFCI function.

  10. Likes Newman109 liked this post
  11. #10
    Join Date
    Feb 2004
    Location
    peekskill, NY
    Posts
    25,296
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    5145

    Default

    Quote Originally Posted by JST View Post
    Since there is no load on the other two, and no load change on them from the RPC starting in any case, it seems that they have to be opening on the GFCI function.
    The wise guess, but I want to hear it from him.

    Probably the adjacent GFI circuits are tripping because if induced current. The starting surge from a converter like that, started across the line, will be very large. Even though the converter circuit is in a separate conduit, the wires share space inside the panelboard. Strong confirmation would be if the other two GFI circuits can be reset and run fine, as soon as the phase converter is running.

    All the starting surge has to do is induce a few mA of current in either the neutral or the hot leg of the 20 amp branch circuits and it's all over. Slight chance the neutral for that
    panelboard has an unusually high impedance for some reason, and the neutral buss is riding up in voltage when the start surge happens.

  12. #11
    Join Date
    Jun 2001
    Location
    St Louis
    Posts
    18,415
    Post Thanks / Like
    Likes (Given)
    2040
    Likes (Received)
    3216

    Default

    The current has to exist, though, to do the classic current imbalance trip. That means it has to go to something and return somehow not on the hot or neutral. The classic GFCI has the hot and neutral running thru a toroid, so that current leaving on one and returning by the other does not cause any induction of magnetic field in the toroid.

    But the simple explanation is induction into the trip coil of the other breakers from the big breaker.

    I have seen cases where the parts were close enough together that heavy currents in one place could induce a current in a nearby toroid (in that case it was actually on the other side of a 16 ga plate of steel).

    It would not surprise me if the breaker induces current in toroids of adjacent breakers.

    Are these breakers on each side of the one with the RPC on it?

    If so, try moving them away by one space.

  13. #12
    Join Date
    Sep 2012
    Location
    Cincinnati
    Posts
    311
    Post Thanks / Like
    Likes (Given)
    5
    Likes (Received)
    62

    Default

    Quote Originally Posted by rimcanyon View Post
    New 15 HP RPC in a new shop with 100A service, everything new. This is attached to a subpanel fed from the original garage subpanel, with #2 copper wires run about 70'. I have two problems I would appeciate some help with.
    Subpanel - did you remove the bonding jumpers? How are the subpanels grounded?

  14. Likes Ohio Mike liked this post
  15. #13
    Join Date
    Feb 2004
    Location
    peekskill, NY
    Posts
    25,296
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    5145

    Default

    Quote Originally Posted by JST View Post
    The current has to exist, though, to do the classic current imbalance trip. That means it has to go to something and return somehow not on the hot or neutral. The classic GFCI has the hot and neutral running thru a toroid, so that current leaving on one and returning by the other does not cause any induction of magnetic field in the toroid.
    ....
    Maybe what's going on: The neutral wire for the converter circuit runs alongside the neutrals for the other GFI breakers. The converter draws 100 amps in a brief spike when it
    starts up, and that pulse of current is inductively or capacitively coupled to the other neutrals. Because the impedance of the hot leg for those other breakers, no current
    flows through the hot leg. The GFI sees current in the neutral and not in the hot leg, and trips.

    I've never seen any manufacturer's caveats about keeping high current breakers away from other ones. My guess is the currrent trip portion of those breakers has a
    pretty closed flux circuit to avoid just this problem.

    In any event, seeing if the other 20 amp GFIs reset after the converter is up and running, would be the next test.

  16. #14
    Join Date
    Jun 2001
    Location
    St Louis
    Posts
    18,415
    Post Thanks / Like
    Likes (Given)
    2040
    Likes (Received)
    3216

    Default

    As I said, I have seen the induction into toroids before. Not as hard as it might seem.

    And we are not talking about the higher current, we are talking about a fairly HUGE surge current as the motor starts. Possibly also the magnetizing surge.

  17. #15
    Join Date
    Sep 2002
    Location
    Salinas, CA USA
    Posts
    4,151
    Post Thanks / Like
    Likes (Given)
    266
    Likes (Received)
    232

    Default

    Thanks for the replies. Here are some answers to your questions:

    The 20A GFIC breakers flip when the start button on the RPC is pressed, and both can be reset when the RPC is running. The 20A GFIC breakers are in spaces 1,3 and 2,4. The 50A breaker is in spaces 9,11. The panel is a 12-space panel, so the 20A breakers are at the beginning and the 50A breaker is last. There is no load on the 20A breakers when they trip.

    The ground for the subpanel is #4 stranded and runs from the old garage subpanel along with the #2 copper feed; there is no earth ground or bonding screw at the new subpanel, but there is an earth ground at the old garage subpanel. There are 100A breakers for the new garage both in the old garage panel and the main breaker in the new panel. They do not trip when the RPC is started.

    I put a clamp on ammeter on the L1 line coming from panel to RPC and it spiked at about 150A, then subsided to below the range of the ammeter once the idler was running.

    The start capacitance stays in the circuit 1 second (there is a timer relay to drop it out). The 15HP idler motor starts immediately, so I could get away with a shorter time on the timer relay.

    I don't know what the wattage rating is on the 15k resistors, but they are not the small ones. The size is about right to fit between the terminals on the capacitors, as you can see from the photo.

    I may decrease the run capacitance again to lower the voltage from 269 to 240, but I have a 10HP Atlas Copco rotary screw compressor that is moving to the new garage, and was the primary motivation for the new RPC. I want to test the AC and BC voltage with that compressor running, and tune the capacitance accordingly.

    The outlets attached to the new RPC have the wrong phase orientation (compared to the Phase Perfect phase converter in the metal shop). i.e. if I take the same motor and plug it in to each system it runs in opposite directions. Since all the outlets are wired and all the leads are color coded, I don't want to swap leads inside the RPC. Will reversing the lines in the 50A circuit breaker solve the problem?

  18. #16
    Join Date
    Jun 2001
    Location
    St Louis
    Posts
    18,415
    Post Thanks / Like
    Likes (Given)
    2040
    Likes (Received)
    3216

    Default

    Well, that lets out the induction explanation.

    The most probable thing than is that the sudden drop in voltage when the RPC is started is the cause. It must cause a sudden unbalanced flow of current in those breakers that open.

    I assume there are machines connected on those circuits, and if so, try opening the disconnects to them. If the breakers no longer open then there is probably a difference between how much "stuff" is connected to the two lines when the machines are "off". That would affect the capacitance to ground, and could cause a difference in current flow when the voltage dops suddenly as the RPC start load is suddenly applied.

    Quote Originally Posted by rimcanyon View Post
    .......
    I don't know what the wattage rating is on the 15k resistors, but they are not the small ones. The size is about right to fit between the terminals on the capacitors, as you can see from the photo.

    ..........

    The outlets attached to the new RPC have the wrong phase orientation (compared to the Phase Perfect phase converter in the metal shop). i.e. if I take the same motor and plug it in to each system it runs in opposite directions. Since all the outlets are wired and all the leads are color coded, I don't want to swap leads inside the RPC. Will reversing the lines in the 50A circuit breaker solve the problem?
    That size resistor is good for a couple watts. As it sits, you have about 5W per resistor. You need a 7W or 10W resistor to be good indefinitely at 5W, and it seems they will have voltage across them as long as the thing is running.

    You need to connect the start capacitor wire to the opposite incoming wire. Reversing the wires at the breaker will not work.

  19. #17
    Join Date
    Feb 2004
    Location
    peekskill, NY
    Posts
    25,296
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    5145

    Default

    Quote Originally Posted by JST View Post
    Well, that lets out the induction explanation.
    ....
    Induction, direct from trip coil to trip coil.

    Induction or capacitive stay coupling, more likely. Right now the impedance to ground on the 20 amp circuits is vastly different, between hot and neutral. Another test is to put some reasonably sized load on one of the 20 amp branch circuits, say a few 100 watt incandescent lamps. Then start the converter. If the loaded circuit does not trip then that's definitive.

    One thing to inspect, is to open the sub-panal (and it is a sub-panel, right) and be sure the higher amp leads are dresses away from the other ones, and twist the hots and neutrals
    up to the point where the breaker is. Try to keep the high current leads tight against the side of the panel.

    This sub-panel should have four wires going back to the main panel. You are not sharing neutral and ground on that run, are you?

  20. #18
    Join Date
    Jun 2001
    Location
    St Louis
    Posts
    18,415
    Post Thanks / Like
    Likes (Given)
    2040
    Likes (Received)
    3216

    Default

    Installation instruction sheet: https://download.schneider-electric....f=48840-058-03

    Do we know if the other circuits are 240V or multiwire 120V? I did not see any description.

    If they are 240V, then there is no need for the neutral other than as a power connection for the GFCI function of the breaker (see install instructions). And in that case the source of the trip problem is not what it might be with the neutral connection.

  21. #19
    Join Date
    Feb 2004
    Location
    peekskill, NY
    Posts
    25,296
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    5145

    Default

    Quote Originally Posted by JST View Post
    Installation instruction sheet: https://download.schneider-electric....f=48840-058-03

    Do we know if the other circuits are 240V or multiwire 120V? I did not see any description.

    If they are 240V, then there is no need for the neutral other than as a power connection for the GFCI function of the breaker (see install instructions). And in that case the source of the trip problem is not what it might be with the neutral connection.
    It's there at the start: "All 3 breakers are new Homeline tandem GFIC breakers (HOM 220GFIC and HOM 250GFIC)"

    So there is no neutral involved. Sorta blows my theory out of the water. One thing for sure is there's going to be a pretty big spike from the start current.

    Altrernate test: disconnect wires from one of the 20 amp ckts, one at a time. When it stops tripping off, the last wire was it!

  22. #20
    Join Date
    Oct 2004
    Country
    UNITED STATES
    State/Province
    Iowa
    Posts
    3,216
    Post Thanks / Like
    Likes (Given)
    77
    Likes (Received)
    453

    Default

    Quote Originally Posted by jim rozen View Post
    Induction, direct from trip coil to trip coil.
    First off... HOMline breakers are HORRIBLE.

    HORRIBLE.

    I hate to say it, but HORRIBLE.

    They certainly won't fail to trip. I've had a 20A breaker trip every time I turned on a 40w light bulb... and the reason was because the electric panel was mounted inside a 24x24 unheated detatched garage, and it was -10F in January.

    The adjacent circuit is drawing enough current on startup, for the inductive trip mechanism in the adjacent breaker to kick off. About that same time, I discovered that my tenancy (which had been fitted with HOMLine panel and breakers during an upgrade) was having furnace problems, then garage door problems, then a bathroom light problem... (sigh)

    Did I mention that HOMLine breakers are HORRIBLE?

    After some research amidst professionals, I was assured universally, that the breakers were prone to nuisance trips of many different scenarios. Cold weather, and adjacent electromagnetic fields being two of the myriad of things that caused nuisance trips. Loud noises, frequent vibrations, or a solid whack to the box will do it too... so if you live near a quarry, and your dishwasher stops working every Tuesday at noon right after the siren and boom... or if you're near a railroad main-line...

    The best 'long term' solution is to replace the panel with a Square-D QO panel...

    The 'short term' experiement, will be to replace the offending HOMLine breaker(s) with a 'fittable' Siemens,Murray, or possibly a GE. I wound up replacing 20 HOMsick breakers 'tween that garage and rental unit, and ran them that way for rest of the winter, replaced one panel (garage)in the spring and the other (rental) in the fall (between tenants), and promptly threw the offending material in the trashcan.

    If a 'compatible' breaker solves the problem, you've found the answer...

  23. Likes JST liked this post

Tags for this Thread

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
  •