Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

How are step-up/step-down transformers wired?

  1. Apr 28, 2012 #1
    I see a lot of these transformers that claim to do both step-up and step-down in the same unit. At least one had a photo of the back side and there was a switch labeled to set it according to the input voltage. What I am curious about is how these are wired. Are they isolation transformers or autotransformers. In particular, is the wiring grounded?

    What I need to do is step 208 volts that comes from two hot wires of a USA style three-phase 208/120 circuit ... where the neutral is NOT available because it is a 208 volt circuit ... down to either 104 volts, or if possible 120 volts.

    Since a 208 or 240 volt circuit in the USA is made from two hot (120 volts relative to ground) wires, either at 120 degree phase angle for 208 volts, or at 180 degree phase angle for 240 volts, neither wire is a grounded/neutral. This means an autotransformer could be unsafe as a means to step 208 volts down to 104 volts, since neither of those wires will be grounded.
  2. jcsd
  3. Apr 29, 2012 #2


    User Avatar

    Staff: Mentor

    As far as I'm aware, no side of any transformer winding is grounded during the manufacture, that would often defeat the purpose of using one! I expect all transformers have an insulation voltage rating, indicating the winding to winding or winding to core voltage that the device can withstand. Autotransformers are "inherently" unsafe because they offer a direct path to the mains. But if the autotransformer fed into an isolation transformer (rated for the 3 phase voltage) the arrangement should be fine and would allow adjustment without sacrifice of safety.
  4. Apr 30, 2012 #3
    Autotransformers are less safe for a couple reasons. One is that they depend on the including live conductors for ground reference, and in particular need to have the side that is common on "primary" and "secondary" (quotes because that's not exactly a correct terminology for an autotransformer) be grounded. They also are more risky for the step-down case than for the step-up case because a failed connection at the common point makes it a pass-through of the incoming voltage to the output.

    But autotransformers are cheaper due to less winding.

    In the case of an isolation transformer (separated primary and secondary), you either have an ungrounded floating secondary, or you have a ground referenced secondary by connecting the ground wire (not the neutral wire) to some point in the secondary winding. This is how power distribution transformers are wired to meet the NESC and NEC electrical codes for all cases but the special ones that need ungrounded for a reason (which also have added requirements like fault detector lights since a first ground fault won't cause system fault current).

    Most of these "travel transformers" are meant for the "one side is the neutral which is grounded at the source" type of electrical system. The American 240/208 volt system differs from that (both live conductors are 120 volts relative to ground and neutral is not usually connected but would be a third wire if it is).

    Using an autotransformer to step that American 240/208 volts down would mean the 120 volt "secondary" is effectively grounded in an odd way. The source is center grounded. But the only connections to the secondary loads don't involve a grounded conductor. At least it's not a wild float. But it could be an off-ground point with one wire at 120 volts relative to ground and the other at 60 volts relative to ground.

    An isolation transformer doing this could have the secondary grounded separately at one end or the other and thus be a normal 120 volt system. That's what I would want. So, basically, I'm looking for grounded isolation type transformer for the ideal case.

    I want to know how they build these products.
  5. Apr 30, 2012 #4


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Why bother with having the transformer specially 'grounded' on one winding? You can arrange that easily for yourself, surely.
  6. May 1, 2012 #5
    Really? How would one do that? These are supposed to be sealed boxes with outlets, not standalone transformer components.
  7. May 1, 2012 #6


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    But they have wires coming out of them, don't they? You can ground one end of the secondary winding easily enough.
    Or perhaps you want something that's more 'plug and play'? In which case you just have to go along with the available standard equipment.
  8. May 2, 2012 #7
    No wires coming out except for the cord with the plug. There is usually one of those universal sockets, and maybe one additional one for North America.
  9. May 2, 2012 #8


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I see. I misinterpreted your question, then.
    I might ask what you would want a non standard (i.e. not what the usual equipment supplies) supply for?
    Did you want to be a 'little bit naughty' but not 'very naughty'?
  10. May 2, 2012 #9
    In a colocated rack cabinet supplied with ONLY 208 volt power, and a requirement to follow electrical codes that include not permitting NEMA 5 type power outlets with 208 volts on them, we need to supply power to a device that has a "wall wart" transformer with NEMA 1 type (North American flat blade) power connectors. Since the device interconnects to other devices, and for other reasons, it must have ground referenced power.

    The idea is to use a 2:1 step down transformer that has NEMA 5 or NEMA 1 outlets on the secondary. I've seen them with Europlugs on the cord, and I've seen IEC to CEE adapters. So if I can confirm a correct wiring of the transformer, this would be good to go to get 104 volts output.

    Although the wall wart is rated for anywhere from 100 to 240 volts, the issue is that it has a fixed set of plug blades of the NEMA 1 type. So it is destined to get no more than 125 volts because we are not allowed to place more than 125 volts nominal on any output with NEMA 1 or NEMA 5.

    Power strip has IEC C13. Adapter with C14 plug and Europlug outlet would be plugged in. Cord from transformer with Europlug would be plugged in. Device power adapter would be plugged into NEMA 5 outlet on the transformer.

    If you have an alternate idea, it must comply with all electrical codes, UL safety, etc. This includes NOT modifying any power devices. FYI, I'm personally quite comfortable with wiring up transformers and outlets and such, as I have done so a few times, all the way up to a 50 kVA transformer. But in this case, it is not an option.

    I am also pursuing an alternate idea of having the ISP provide additional wiring for direct 120 volt service. But this is very expensive (the install cost is more than double the cost of the above setup). Another alternative being pursued is to find an alternate to the device which makes use of C13 inlet like a PC does, instead of its NEMA 1 based wall wart.

    I just need to find a properly wired step-down transformer and the project is on.
  11. May 2, 2012 #10

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    see if either of these links is any help...

    note 250 volt NEMA 6-15 plug on input side i think you probably have those ?
    not very expensive at hobby sites.


  12. May 3, 2012 #11
    They list only two ways to make such a transformer, yet there are three ways. The difference between these is that the grounded isolation method connects the neutral side (has the wider blade hole) of the 120 volt output to the ground wire itself, inside the transformer. It seems from their description they are using the floating isolation method.

    At least the greatest danger does not exist with this one (an autotransformer that would make the neutral output hot). But the floating output on this can still present a 60 volt coupled potential in the neutral (because it is 2 phases from a three phase system), which would create more noise. If it were the 2 phases of a split single phase system, then it would be OK because the float would be at a near zero potential (because both live wires would be at 180 degrees).

    If I could use a box with a wire-up transformer inside, I definitely could wire it right (if the correct winding voltages were available). Unfortunately, that is not an option.
    Last edited: May 3, 2012
  13. May 3, 2012 #12

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    that GISD series is a genuine isolating transformer with electrostatic shield.

    It is not clear to me whether the secondary 'floats'.
    Since it comprises a 'separately derived system' i'd think the output side neutral ought to tie internally to ground prong of input side, that'd be the internal 'neutral to frame bond' to satisfy NEC 250.30.a.2.a

    Good question for Stancor guys. Or find one and check with ohmmeter.

    If so would that work for your application?

    Be careful, Stancor makes autotransformer stepdowns that look real similar.

    EDIT finally found a drawing
    looks discouraging
    i wonder how they get away with selling that tranformer with floating secondary?

    old jim
    Last edited: May 3, 2012
  14. May 5, 2012 #13
    That would not be an issue on step-UP where the supply has an SGC/neutral. But clearly there is a complication if the supply has no SGC/neutral. I guess in such a case you don't get to use that exception since its requirement cannot be complied with.

    No. Since the need for the step-down is part of a three phase 208/120 volt system, with the part being just 2 hot wires in a 2-wire branch circuit, that would leave the result floating "off center" and thus with a relative (even of just capacitively coupled) voltage to ground. I don't want those issues. It needs to be a correctly grounded 120V-like (will be 104V) system.

    I am doing this to avoid having to wire an extra circuit or an extra neutral wire just to power one device. Fortunately, since starting this thread, I have found a version of the device (a KVM switch) that uses an IEC connector for power and handles 100-240. So the need no longer exists, although it's still an interesting issue to ponder.

    I could have just wired 208 volts to a NEMA 5-15R, but the colocation facility prohibits doing that (and I agree with them ... I would do exactly the same thing if in that role).

    There are uses for something to be fully isolated and floating. Eliminating ground loops in interconnected equipment could be one of those uses. if I had an electrical test bench, I'd want one of those to be handy.

    BTW, I also have a use, here at home, for a step-UP transformer. But it gets more complicated because there are TWO ways to wire an AUTOtransformer, and both of them have a valid use with 240V secondaries. Stepping up from 120 to 240 in an autotransformer involves a primary tap in the middle of the 240V winding (or between two 120V windings, depending on your perspective). The two ways are just reversing which input wire has the system grounded conductor.

    1. If the grounded neutral input connects to the shared end of the winding, then the output has one end grounded, too. The remaining output wire is 240V relative to ground. This is how you get power in virtually all of Europe and most of the world.

    2. If the grounded neutral input connects to the center point of the winding, then the output has both ends ungrounded, but 120V relative to ground in opposite phase angles. This is how you get 240V power in North America where "single phase" is provided.

    I would want #2 but travel transformers are likely wired like #1.

    FYI, the reason for wanting #2 at home would be satisfied by either a 120-0-120 system or a 60-0-60 system. I have a laptop with some metal parts. It is connected with a brick PSU that has an IEC C8 connector, and thus no ground. So the average voltage potiential across the two incoming wires is going to be coupled across to the DC. So the output is technically the sum of 60V AC and 12V DC. When I press hard on the metal, the 60V AC is pulled down since it has a very high impedance in the coupling. But a light touch gives that tingling effect. Unplugging the power and running from battery eliminates the tingling effect, so there is no doubt. My theory is if I get the input center referenced to real ground, then the resulting effect will eliminate the AC component in the DC, relative to ground (and air, and me) and I won't get the tingling effect from a light touch, anymore.

    I may just have to get a 6-20R circuit wired to my computer desk area for this.
  15. May 6, 2012 #14

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    That's really interesting. i think I'd try another brick.

    It'd be interesting to measure that leakage. Looks like it should be less than amilliamp.

    I wonder what's your threshold of perception? Again from Simpson:
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook