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

Voltage transformers and magnetic field leakage

  1. Oct 3, 2014 #1
    I’m looking to purchase a toroidal AUSTRALIA-to-US stepdown voltage transformer, and have narrowed my choice down to two models, one permitting a maximum rate of energy consumption of 100 Watts, and the other permitting a maximum rate of energy consumption of 250 Watts.

    Both of these maximum wattages (100 W and 250 W) are easily in excess of the wattages of the appliances that would be used with the stepdown transformer.

    My main purchasing criterion is the minimization of magnetic field leakage (i.e. stray magnetic fields), and this is what led me to decide on a toroidal design stepdown transformer.

    I know that I can use magnetic shielding, Mu metal, etc., to address concerns in relation to magnetic field leakage, but if the choice between using a smaller rated stepdown transformer (100 Watts, say) versus a larger rated stepdown transformer (250 Watts, say) could ALSO influence/affect the strength of the resultant magnetic field leakage, I’d like to go with (choose to purchase) the stepdown transformer that would give rise to a smaller magnetic field leakage for an appliance of a given wattage.

    For example, let’s say that the most energy-hungry appliance that would be used with the stepdown transformer, consumes energy at a rate of 50 Watts.

    If we let ‘MFL_100(50)’ denote the resultant magnetic field leakage when operating the 50 Watt appliance from the 100 Watt toroidal stepdown transformer, and ‘MFL_250(50)’ denote the resultant magnetic field leakage when operating the 50 Watt appliance from the 250 Watt toroidal stepdown transformer, then which of the following would most likely be true?:

    (a) MFL_100(50) < MFL_250(50)
    (b) MFL_100(50) = MFL_250(50)
    (c) MFL_100(50) > MFL_250(50)

    If anyone could answer this question for me, it would be greatly appreciated.
  2. jcsd
  3. Oct 3, 2014 #2


    User Avatar

    Staff: Mentor

    Welcome to the PF.

    The toroidal transformer is your best bet for minimizing leakage field.

    If you use a traditional construction transformer instead (your 100VA and 250VA examples?), then the larger transformer will likely have *slightly* less leakage given the same power throughput. But the difference will be small.

    Why are you interested in minimizing the leakage field?
  4. Oct 4, 2014 #3
    In an ultra-sensitive application, I saw to adjacent toroids cross-talk. I never found the mechanism, but the ones I've seen shielded were simply potted in a stamped cup made from mild steel. The same would hold true using a steel box - just space the transformer from the sides and allow that the steel has a moderately higher permeability than air to do the shielding.

    Mu metal is a bit much. It can cut workers for an application in which covers are sufficient

    - Mike.
  5. Oct 4, 2014 #4
    Thank you for that information, but if we were to compare the resultant magnetic field leakages for a 100VA toroidal stepdown transformer and a 250VA toroidal stepdown transformer, given the same power throughput, what would we likely observe? (i.e. in which of these two 'toroidal' scenarios would the magnetic field leakage be greater, if any?)
    The preferred location for the stepdown transformer, as stipulated by the user, would see it positioned immediately alongside IT equipment (workstation and peripherals). Although the said IT equipment may intrinsically have adequate magnetic shielding, I just don't want to take any chances ... I just want peace of mind that I've done absolutely everything to minimise magnetic field leakage!
  6. Oct 4, 2014 #5
    You're probably right, it's just that I always get a little 'paranoid' when stepdown voltage transformers are to be placed immediately alongside IT equipment (the user's preference, in this case).
  7. Oct 4, 2014 #6


    User Avatar
    Science Advisor

    Use the lowest power toroidal transformer that will do the job. That will minimise magnetisation current, which with a larger transformer can be greater than the power current.

    I believe you are worrying unduly. IT equipment is digital and so is immune to the very small stray magnetic fields from toroidal transformers. Some cathode ray tube displays had a problem with stray fields causing a cyclical drift in the image.
  8. Oct 4, 2014 #7

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Two thoughts:

    1. Transformers are rated for nominal full load conditions.
    Operating a core at lower than rated voltage lowers its flux level.
    That would help you minimize leakage.
    If you use a higher VA rated transformer with a tapped primary you can pick a higher voltage tap to lower flux;
    its voltage drop will be less at the light load it sees which gains you back some of the voltage you gave away by underexciting the primary.

    Look carefully at the transformer datasheets to see if that's an option....

    2. Back in my youth i roamed around the plant with a "Flux Detector", a small coil hooked to an oscilloscope.
    One of the things i noticed was the leakage flux adjacent a transformer was far from sinusoidal.
    Instead it was narrow peaks , as if the middle portion of the flux sine wave were flattened but the peaks exaggerated.
    That makes sense when you look at the B-H curve
    For economy cores are operated with flux pushed up almost to the knee of the curve.
    When flux is pushed up into the nonlinear region near knee , more of it will be pushed out into the surrounding air as leakage.
    That's why i said above that reducing the voltage will help reduce leakage flux.

    So - if you're having troubles with flux leakage, try derating your transformer by operating it at lower voltage.
    In US where we use both voltages, a 240 volt transformer operated at 120 is essentially leaving its core asleep . Flux never gets more than halfway to the knee.
    But you'd have to have a transformer rated for twice as many VA so that the copper is not overworked. Remember you halved voltage but not current.

    old jim
    Last edited: Oct 4, 2014
  9. Oct 4, 2014 #8
    @ Jim, Good observation regarding the non-sinusoidal nature of the leakage flux. Even worse is when you have a core saturate during turn on - ouch!

    The killer I've seen around digital equipment is the Electrically Fast Transient (EFT). Switches and relays arcing are a typical source, and an RC snubber or MOV across the contacts is usually a good cure. However, anything that has the CE mark on it has been through some degree of EFT testing and successfully passed it.

    It's very easy to slip into mysticism regarding electromagnetic compatibility because so few people have a the combination of working experience and a sound understanding of the theory. However, unless you're using homemade equipment, you're likely to have the European Union's CE mark which means your equipment is going to be tough enough to survive the normal spike's and surges and fields of day-to-day life.
  10. Oct 5, 2014 #9

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Amen , Mike.

    Statistically a sine wave spends more time near its peak than near zero, so with random switch closing most of the time your inrush isn't so bad.

    But randomly switching on AC power to a transformer will occasionally give a huge current transient.
    In the testing i did , it happens about one time out of ten .
    It's because you are starting from near zero flux .
    Since flux is integral of voltage, and a definite integral has an initial condition, the starting point is important. It's your initial condition.

    Here's normal operation. Focus on first half cycle of voltage.


    e = n d[phi]/dt
    observe that flux has positive slope over the whole interval where voltage is positive.
    It starts from negative phimax not zero, and it goes to positive phimax.
    phimax would be the knees of the B-H curve above in post 7.

    Now look at what happens when you have the misfortune to close your switch exactly on the voltage zero crossing.
    Flux now starts from zero instead of negative phimax.
    So flux goes to well past the knee of your B-H curve , and that's saturation. And it takes a LOT of current to push flux that high.
    I never measured leakage flux during that transient but suspect it too spikes. There's plenty of mmf. I did measure peak currents 10X transformer's full rated current.


    You can prevent that inrush by using a transformer at half its rated voltage. There's enough iron then to support twice your normal phimax.
    But that is wasteful of money, so is just a thought to keep in your "bag of tricks" for special situations like Darp is facing over in his 780 watt drill thread.
    Observe that had you closed the switch at the voltage peak your starting point would have been zero flux and zero magnetizing current, just as in normal operation, so there'd be no inrush.

    Design guys are aware that solid state relays come in a version that closes on the voltage peak instead of the zero crossing for just that application.
    We maintenance guys just stumble across them in good designs.

    old jim
    Last edited: Oct 5, 2014
  11. Oct 5, 2014 #10

    The Electrician

    User Avatar
    Gold Member

  12. Oct 5, 2014 #11

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

  13. Oct 5, 2014 #12

    The Electrician

    User Avatar
    Gold Member

    Maybe you have to be logged in?

    Anyway, here's the image for the 180 amp surge.

    I have a linear power supply with a 400 VA toroidal transformer, and I put a current shunt in series with the primary winding (120 VAC here in the U.S.), and disconnected the secondaries from the rest of the circuitry. I set up an oscilloscope to capture the current in the primary and then applied line voltage. I turned the supply on and off a lot of times until I had captured the current surge when the primary voltage was applied just as the sine was crossing zero volts. The peak surge current was about 180 amps.

    The grid voltage is applied to the primary of the transformer at about 2.2 cm (green). The current pulse (purple) doesn't even begin until about 5 mS later, at about 4.2 cm. This is because it takes that long for the flux to reach saturation (the flux is proportional to the integral of the applied voltage).

    Notice that the current surge is so large that it pulls the grid voltage down when it occurs, just after the peak of the applied grid sine wave.

  14. Oct 5, 2014 #13
    Transformer core saturation is an evil thing...
  15. Oct 5, 2014 #14

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Thanks Mr. Electrician !

    That peak would have been higher yet had your voltage source been stout enough to deliver more current.
    Where the purple trace leaps up is where the core saturated.
  16. Oct 5, 2014 #15
    Firstly, let me say thank you to everyone for your excellent posts in this topic so far.

    I've just realised that the pricing for the magnetic shielding and Mu metal options would result in the user's stipulated upper limit on the spend for the toroidal stepdown transformer being exceeded. Similarly, the supplier's pricing for the primary tapping option (an excellent idea, by the way, Jim!) would result in an overall spend that exceeds the user's budget for this purchase.

    Basically, the 100VA toroidal stepdown transformer and the 250VA toroidal stepdown transformer are each within the user's budget if purchased as an 'off-the-shelf' item in the default configuration, but the moment I request a modified configuration from the supplier along the lines of magnetic shielding or primary tapping, the overall price becomes cost prohibitive for the user.

    Given the user's budgetary constraint for the intended purchase, as well as the fact that I'm an equipment/parts purchaser rather than a 'hands-on' electrical whiz/guru, the only variable that I really have control over in this situation, is the choice between purchasing a 100VA toroidal stepdown transformer for the user, or a 250VA toroidal stepdown transformer instead.

    So, for the situation that I find myself in, the actual question that I'm really hoping someone can help me with, is:

    Which, if any, of the following two choices:

    (a) 100VA rating 240V-to-110V toroidal stepdown transformer
    (b) 250VA rating 240V-to-110V toroidal stepdown transformer​

    would be better in terms of minimizing magnetic field leakage?

    Note: Both of these maximum wattage ratings (100VA and 250VA) are easily in excess of the wattages of the various items of equipment that would be used with the stepdown transformer.​
  17. Oct 5, 2014 #16

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Fields is not my area of strength. I'm guessing from an intuitive feel.. See what other folks say..

    I'll cast my vote for the smaller of the two. Less surface area is my thinking.

    A steel fender washer about same od as transformer mounted over it on a 1/8th to 1/4 inch spacer i think should discourage leakage flux from straying afar. You might find it comes with one.
    It doesn't take a lot. I once shielded a loudspeaker magnet with a steel tin can from the pantry.

    old jim
  18. Oct 5, 2014 #17

    The Electrician

    User Avatar
    Gold Member

    Your best choice would probably be to buy two of the 100 VA transformers and wire the primaries and secondaries in series. Then you will have greatly reduced maximum flux a la Jim Hardy's suggestion. It's the near saturation of the core that "squirts" flux outside of the core. Two small transformers wired as I suggest will greatly reduce the flux density in the cores, and thus greatly reduce the external leakage flux. It's like using higher voltage taps, but without needing a special transformer.

    I don't know what the pricing is, but the 100 VA units are no doubt considerably less expensive than the 250 VA units, so maybe two of them will still be within your budget. :)
  19. Oct 5, 2014 #18
    The price of the 100VA unit is approximately 75% of that of the 250VA unit. Your suggestion is quite ingenious, but unfortunately it falls outside the user's budget (also, having to find space for two of the 100VA units on the user's desktop would likely be a big problem).

    If you had to guess which of the two units (100VA toroidal stepdown transformer versus 250VA toroidal stepdown transformer) would result in a smaller magnetic field leakage, what would your guess be? (and why?)
  20. Oct 5, 2014 #19

    The Electrician

    User Avatar
    Gold Member

    The problem is that my guess would just be a guess. Sometimes, smaller transformers are pushed further into saturation, which would lead to greater external leakage. The manufacturing process is fraught with considerable variation--laminations may not be stacked as well today as yesterday, they may slip during the impregnation process. Different batches of laminations may have different saturation characteristics.

    In the absence of any other information, I'd probably go with the small one.
  21. Oct 5, 2014 #20
    After spending over two decades going in and out of the EMC business, I can say this fear has never materialized.
    However, for those obsessed, I've found very slight leakage in the center of the toroid with the field traveling axially. I also found the field traveling parallel to the axis along the center on the outside.
    If ever the transformer manufacturers had a problem with this configuration, I'm sure they would have placed a copper "belly band" around the periphery to short the field as they do with E-I cores.

    Since the application requires some manner of packaging, a mild steel box with an overlapping mild steel cover can be applied as a low cost manner of shielding - especially if the transformer is spaced from the edges using non-magnetic stand offs.
  22. Oct 7, 2014 #21

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Mike is obviously better versed than i.

    One last thought....
    I assume you're 60 hz wherever you are.
    Is that transformer you're contemplating rated for 50 hz ?
    If so, then it has plenty of iron for 60 hz and peak flux will be ~5/6 of that for same voltage at 50hz and I'd agree it's a non worry.

    If it's NOT rated for 50 hz then buy their 50 hz version instead. It'll have either more iron or more turns, both of which lower flux density.

    Maybe you'd post a link to their catalog. If there's a 277 volt transformer available in 250 VA size , well, there's your answer.

    old jim
  23. Oct 9, 2014 #22
    I reside in Australia, where the grid power supply is 240V @ 50Hz.

    As the user wants to operate a US device here in Australia, a 240V-to-110V stepdown voltage transformer is required.

    Of course, such a transformer leaves the 50Hz frequency unchanged, but this is not likely to be a problem as the US device (which, technically speaking, would have been designed with a 60Hz frequency in mind) contains neither a motor, nor a frequency-reliant timing mechanism, and so should operate normally with a 50Hz power supply.

    All of the supplier's Australia-to-US stepdown voltage transformers are rated as producing a [110VAC, 50Hz] output from a [240VAC, 50Hz] input. Consequently, in terms of trying to achieve a lower flux density solely on the basis of a lower power supply frequency, no one model has an advantage over any other, and so I cannot discriminate between models on this basis.

    Nonetheless, the points you make Jim, are quite valid (and quite clever!).

    Just to update the overall situation, the user has now informed me that existing office equipment can be rearranged so that the stepdown transformer is at least one and a half metres away from any potentially vulnerable appliance/device.

    From memory, magnetic field strength is inversely related to the square or cube (at this moment, I forget just which!) of distance. If that is indeed true, then any worries I once had with regard to magnetic field leakage, have now all but disappeared (originally, the distance separating the stepdown transformer and adjacent I.T. equipment was going to be just a few centimetres!).

    I'm going to go with the 100VA stepdown transformer (it's sufficient for the task at hand) without any added magnetic shielding.

    I feel confident now, based on all of the posts to this topic, that there won't be any problems.

    Big thanks to everyone for helping me out ... I appreciate that you all gave of your time and expertise to do that.
  24. Oct 9, 2014 #23

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Great !

    If there's any question remaining in anybody's mind,
    invert a steel soup can or wastebasket over the transformer.

    Boring Anecdote Time:
    Toward the end of analog TV days I bought a cheap subwoofer and mounted it underneath my 26" Sony.
    It was unshielded and turned the corner of my picture bright purple. The purple stayed when i moved the woofer away, it had magnetized something inside the set.
    I opened the speaker and cut an empty Libby's peaches can to fit around the outside of the magnet. It required a small cut to make a snug fit so i was left with a narrow slit maybe 0.030 inch. A spot of hot glue held it in place and i oriented the slit down away from screen.
    I put the subwoofer back underneath the set then took my old fashioned soldering gun and waved it around in front of the TV screen. That removed the purple - old technician's trick, use a soldering gun for a degausser. The fix was successful .
    For curiosity i tested the new flatscreen TV that replaced my faithful old Sony . It seems unaffected by magnetic fields either AC or DC..

    Point of that anecdote -
    1. Shielding is pretty easy to do. If you do shield for an AC field, feel the material and if it's getting hot you have extreme leakage flux and need to increase clearance. Of course a non-alternating field from permanent magnet creates no heating.
    2. An old fashioned soldering gun makes a strong field at line frequency and could provide a forcefull sensitivity test for the client's equipment. Hold a tool adjacent a soldering gun and feel it vibrate. You can remagnetize a compass with one so it points the wrong way !
    3. Digital circuits have innate hysteresis that was built in for the purpose of reducing sensitivity to low level interference.
    I think you'll be fine.

    old jim
  25. Oct 11, 2014 #24
    Thanks heaps, Jim! ... You've been extremely generous with your help!

    As for your anecdote, I found it extremely interesting (and most informative)!

    Wow! ... This forum is very lucky to have a guru like you around to help people!
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook