Phase failure detection using open delta connection

Discussion in 'General Electronics Chat' started by kishorc, Sep 4, 2010.

  1. kishorc

    Thread Starter New Member

    Sep 4, 2010
    2
    0
    Hello. Can someone guide me on the working of the open delta connection, shown in the attachement, for phase failure detection ?
    The requirement is, if a phase at primary winding fails, the corresponding LED on the secondary, should switch off.

    Thanks. Kishorc
     
  2. marshallf3

    Well-Known Member

    Jul 26, 2010
    2,358
    201
    I think you need to go over that circuit again, you don't hook a split phase transformer up to a three phase source.

    You're probably dealing with 240/120 high leg Delta?
    http://en.wikipedia.org/wiki/High_leg_delta
     
  3. kishorc

    Thread Starter New Member

    Sep 4, 2010
    2
    0
    Thanks, for your reply Marshallf3.
    I am not using high leg delta configuration.

    I missed mentioning that these are two individual single phase transformers, whose one leg is connected together. Instead using of three different PT's for phase failure detection, only two PT's are being used.
     
  4. marshallf3

    Well-Known Member

    Jul 26, 2010
    2,358
    201
    Not so hard to do with Neon lamp indicator assemblies, with LEDs it'll take a heck of a lot of resistance to limit the 100V down.

    Let's say you used a comon red LED, 1.9V @ 20 mA.

    Let's also remember that the 100V and the 200V are RMS values, their DC peaks are going to be 141.4V and 282.8V respectively.

    For the two 100V (141.4V) outputs you'd need to drop 139.5V @ 0.02A = 6,975 ohms at 3W minimum. A 6.8K 3 W resistor in series with each of those two LEDs would probably be close enough for the LED value I used to illustrate it.

    On the third LED which appears to have 200V (282.8V) across it, you're going to need to drop 280.9V @ 0.02A = 14,045 ohms at 6W minimum. A 15K 10 W would work fine in this case.

    It's far easier to pick some out of here, they're already panel mount to boot:
    http://www.mouser.com/Optoelectroni...Z1yzudpoZ1yzudosZ1z0x7qtZ1yzudpd&Ns=Pricing|0
     
  5. marshallf3

    Well-Known Member

    Jul 26, 2010
    2,358
    201
    Seriously forgot something when I made that post - you're going to need to put a simple rectifier in series with each of the LEDs or the reverse voltage is going to blast them. Anything good for an amp at 400V such as a 1N4004 would be fine on each.

    Yet another reason I prefer neon lights on higher voltage AC circuits, they just aren't so picky.
     
    Last edited: Sep 6, 2010
  6. Dyslexicbloke

    Active Member

    Sep 4, 2010
    420
    19
    Not withstanding all of the above responses which sound right to me I don see how that config will detect an L2 fail .....
    You would still have 415 across the pair of primary windings in series with each other.

    Whatever indicators you use I am fairly confident you will need them in a star config even if you dont have a neutral from the incoming supply a star config would create one for your circuit. If you do the maths you will see what I mean, nutral is simply earth back at the substation on a star wound transformer.

    WARNING .... This point (Virtual neutral) will float to some strange values if you remove a phase or your power factor is out.

    I would recommend Phase neutral connected indicators if at all possible.
    Most importantly there is no protection of any sort shown ....
    Please tell me this is for clarity or something, if not you are building an incendiary !!!!
     
  7. n1ist

    Active Member

    Mar 8, 2009
    171
    16
    I'd look into using capacitors instead of resistors to drop the LED voltage; power dissipation is a lot less.
    /mike
     
  8. marshallf3

    Well-Known Member

    Jul 26, 2010
    2,358
    201
    Not all three phase circuits have neutrals.

    If you look closely at the way he's doing it it will work and it saves him a transformer.

    BTW: The tolerance variations on caps is far too high to return the percentage of voltage drop tolerance you'd need to use LEDs, matter of fact it's going to be somewhat shaky just using resistors, LED life isn't going to be ideal.

    This is the reason I prefer Neon lamp assemblies. They tolerate a wide range of voltage variation and most of the panel mount ones already have the correct dropping resistor inside the assembly - which you can add to if desired.

    If you wanted to go neon you wouldn't even need the transformers, just an extra resistor for each indicator assembly.
     
  9. Dyslexicbloke

    Active Member

    Sep 4, 2010
    420
    19
    I get that the supply hasn't got a neutral .... But connecting the indicators in delta isnt going to work well.

    See attached.
    Assume discharge lamps with a current limiting resistor, neon's say
    that will breakdown and glow at around 70V. The resistors have no effect until this happens and current is flowing.

    In delta, on the left, all indicators are phase phase and will see 415 in normal circumstances.
    Ia is detecting L3 - L1,
    Ib is L1 - L2 and
    Ic is L2 - L3.

    If you were to loose L1 then Ic is unchanged but ....
    Ia and Ib are now in series with each other and also seeing L2 - L3

    Two 70V lamps in series will still light, although slightly less bright because far more than the 140V or so that is required to break down the pair is percent at L2-L3.

    However .. Using a star config, on the right avoids this.
    Even though there is no neutral precent the star point is, vectorially speaking, still null or neutral.

    Each indicator has a direct and discrete relationship to 1 phase. If the phases are in balance the central common connection will be neutral and 240V will exist across each indicator.
    Removing 1 phase will extinguish 1 indicator and place 415V across the remaining two in series ie 207 each.

    Actually the neutral point still exists it just that it has moved now because it isnt tied to ground.

    In practice the virtual neutral point and therefore the voltage across any particular indicator will float about somewhat because the three legs will be conducting current only when the lamp threshold in each is reached.

    Al
     
  10. marshallf3

    Well-Known Member

    Jul 26, 2010
    2,358
    201
    If his values are close enugh to the edge it's going to work, but I'll agree that the proper way would be to monitor each individual phase line even if a virtual neutral had to be created.

    I've got 480Y/277 feeding our building with an ampacity rating of 6,500 amps, stepped down through transformers where necessary. I love working with it and don't need any sort of indicator to tell me if I've lost a phase - all the big motors raise a racket you can hear throughout the building.
     
Loading...