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

Practical transformer question-3ph/single phase.

  1. Feb 13, 2012 #1
    I need a 10-15 : 1 step-down transformer for an electrolysis project. It needs to be capable of about 150A. I have an old ex-military transformer/rectifier, but it outputs 200A at upto 230V DC. It has a humungous transformer in it , and some banks of BIG diodes, so if I get a variac from ebay that will handle 0-240V ac at about 15-20A, and feed the output of that into the transformer, I'm thinking I should get 0-24V at 150A no prob. Obviously, the diodes and rest of it provide the most awesome DIY rectifier kit in the world. The one problem is that the transformer is 3 phase. It has inputs of 440,415 and 380 V, and outputs of 220,210 &200 V. There is also a 0 and 30V connection. There is a second 0 which has tabs for + & - 2% on either side. I guess this is for the neutral input.

    1) The transformer is in the form of 3 coils, all on the same laminated frame but with about 50mm air gap between them. The frame runs the full width top and bottom, but not down between the individual coils.
    What will happen if I connect one of the 3 coils to 240V supply. Will the others, which are on the same laminated core frame, be energised and cause problems? I only need to use one of the coils.

    2) I fear that the inrush current might suck the whole of the national grid into my workshop and turn the universe into a huge black hole! Could I feed the input through a 30A triac, and start with it turned off, to build it up slowly, or would the triac stand no chance?

    3) If it won't work well in its 3 phase form, would it be ok to cut the frame between the coils with a Stihl saw, and use the coils separately?
  2. jcsd
  3. Feb 13, 2012 #2

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    A three phase transformer is usually just three single phase transfomers joined together - imagine 'siamese triplets'.
    Can you post a picture?

    Look inside and see if you can separate some wires to power only one leg.
    It will work fine as it is. Just take your low voltage off the same leg as you apply the power. The other two legs will sit idly by, each making about half voltage. But if you try to load one of those legs it will collapse its voltage and raise voltage on other leg.

    Each leg is good for 1/3 of total transformer rating. Being military it's probably a conservative rating.
    If you use it at lower voltage than it's rated you will have no inrush problem.
    230 volts on a 415 winding will make life very easy for the iron core.
    But feel of the winding to make sure your copper is similarly underworked and staying cool.
  4. Feb 14, 2012 #3
    Hi Jim,

    Thanks for your input. ITLO what you said, it would probably be best to wire all 3 in parallel and thus divide the current equally, unless that in itself would cause a problem. I think I was obsessed with it trying to emulate 3 output phases from the one input.

    Do you think the triac idea would work for variable input, or would a nice meaty variac be the way forward?

    I've uploaded a couple of pics, ( I think! ). For scale, there is a 5L DI water bottle. I used to scrap main frame computer equipment for a living, so I've seen some pretty heavy duty power supplies, but the scale of this is quite different. I wonder how big it gets before they need to be oil cooled!

    PS point taken about feeling the coils up, it would be a shame to fry the old girl.

    Attached Files:

  5. Feb 14, 2012 #4


    User Avatar

    Staff: Mentor

    There should be no inrush of current at switch-on. An unloaded transformer is inductive, so current builds up 'slowly'.

    It's at switch off that inductive loads have an antisocial tendency. There can be a spark across the switch contacts due to the inductive kick as the magnetic field collapses.
  6. Feb 14, 2012 #5


    User Avatar

    Staff: Mentor

    There, fixed it for you. :smile:
  7. Feb 14, 2012 #6

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    """ unless that in itself would cause a problem. ""

    i think it would.

    Be aware that the magnetic flux flows around a closed loop in the iron, just as Kirchoff's current makes a loop in the copper.
    Now look at the center leg of your transformer.
    When flux is UP in that leg it is DOWN in the other two. Right hand rule and all.......

    So if you were to wire them in parallel you'd be trying to push flux UP in all three at once.
    Since flux couldnt then make a closed loop your current would go sky high.

    To get a feel for this, experiment with your core.
    Apply maybe twenty volts to center winding.
    Measure outside ones and see if you dont get ten volts each - that's because flux from center leg divided equally between the outside legs for its return trip.

    Now place one turn of big wire, like #10, around either outside leg. Or just short the winding.
    That leg will now have ~zero volts and the other one ~twenty because the shorted turn pushed all the flux over to the other leg.

    I reallly suggest you do that experiment - it will cement in your mind how a transformer works.
    Measuring the current in your shorted turn will give you a good idea of the amp-turns required to push the flux through your core, a measure of its reluctance.
    You could also wire all 3 phases in parallel and excite gently with your variac but monitor current and be cautious - use a fuse..

    It is great fun squishing single phase flux around a three phase core. You are lucky to have one!

    Have Fun !

    Maybe you can figure out a novel way to wire your coils and use them all, but i doubt you'll be able to do better than one at full voltage and two at half voltage.

    With three phase you can use all 3 windings at full voltage because of the 120 degree angle. Walk around a triangle and you have closed a loop, the three equal sides can be thought of as vectors that add to zero..
    But you can't make three vectors of equal length that are aligned 0 or 180 add to zero.

    please keep us posted!

    old jim
  8. Feb 14, 2012 #7

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    ps inrush to a transformer is possible.
    if you close in precisely at the zero crossing , a power transformer core will likely saturate on first cycle or two.
    But not if it's operated at considerably less than rated voltage.
    This is different than capacitor inrush.
  9. Feb 14, 2012 #8
    Why do they get it on about oil cooled transformers?
  10. Feb 14, 2012 #9
    Quite misleading.
    Transformers are known for their inrush currents on switching on. In-fact Protection circuits of heavy power transformers have some mechanisms to identify this inrush current from shot circuit current and prevent false tripping.
    http://www.opamp-electronics.com/tutorials/inrush_current_2_09_12.htm [Broken]
    this article describes about inrush current.
    Edit: Though, in this case inrush current won't be significant, because, as jim hardy told, we are operating 415V coils at 230 volts.
    Last edited by a moderator: May 5, 2017
  11. Feb 15, 2012 #10
    Thanks Jim,

    I will keep you posted, but it won't be immediate as I need to get it into my workshop from the farm next door where it's in storage, and that will take a bit of doing. Regarding your suggested experiments, I will do them because I don't like to not understand things, and my transformer knowledge obviously needs a bit of polish.
  12. Feb 15, 2012 #11

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    "" I will do them because I don't like to not understand things, ""

    My kind of fellow !!

    While thinking on this, it came to me
    i think you can get two windings at full power by connecting them parallel but with reversed(opposite) polarities, leave third one unconnected.
    That should push flux around a loop of two phases leaving third one idle. As Meatloaf says, "two out of three aint bad".

    Use your variac and come up slowly...i could be wrong.
  13. Feb 15, 2012 #12
    Just a thought, Jim.

    This might sound like awful butchery, but I did mention that as there is no core between the coils, it would be easy to cut the core, top and bottom in 2 places with a Stihl saw, and make 3 separate coils. Obviously, I can experiment with it in 3 phase mode first, to gain better understanding, but as there is no real advantage for me in having a 3 phase transformer, would this not be the best way to use all 3 ?
  14. Feb 15, 2012 #13

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    your magnetic flux must make a closed loop inside the iron
    if it goes UP through iron of one coil it must come back DOWN somewhere still through iron, so dont separate those iron legs !

    Think of flux as alignment of individual iron molecules like tiny compass needles.
    Air molecules dont align worth a darn - so keep that iron together.

    An energized three phase core is a marvelous thing to have around your workbench.
    Figure out how many volts per turn it makes and voila - instant adjustable AC source. Just wrap a few turns around a leg. I'm guessing from photos yours makes about 1/4 volt per turn.

    i dont think you will be able to use all three phases.

    What part of world are you in? You might be able to swap that for a huge single phase... keep an eye out for HAMFESTs , where amateur radio guys swap raw material for their projects.
  15. Feb 15, 2012 #14
    Point taken about the iron core. It's presumably the fact that it goes UP the core of one coil and down the core of the next coil that was fooling me. I guess that due to the 3 phases, if the flux of one phase is high, the flux of the next is low, so its core can be used for energising the other? It is the lack of iron between the coils that I didn't get- I assumed that it would therefore be ok to cut it. Shows what assumptions are worth. However, that is why I checked with someone who knows what they're doing. Always check before you cut!
    I'm in the middle of the U.K. There are ham meets, I'll see what's up. You're right that it might be better to swap it to someone who can use it properly.

    Thanks again.
  16. Feb 15, 2012 #15
    The three phases, each at the three 120 degree angles, will average out in the 3 legged core. The phase angle and amplitude of any phase angle is equal (and opposite for the core) to the phase angle and amplitude of the other two.

    At 200A 230V L-N per phase winding, this is a 138 kVA transformer. You could use 2 windings in single phase by removing a winding leg, for a 92 kVA transformer. The 400V side would be 115A. If you wire two 400V windings in series and two 230V windings in parallel, feed 230V to the (now 800V) series winding, you would get 66.125V on the paralleled 230V winding, with a capacity to go to 400 amps. To get your 150A at 66.125V, you'll draw 43.125A from 230V.

    The removed winding may be usable as a current limiting inductor.

    What voltage does your electrolysis project need? Is 66.125V usable? What current limiting means is involved? Or are you depending on finding a voltage that won't draw more than 150A at the lower voltage on the load?
    Last edited: Feb 15, 2012
  17. Mar 15, 2012 #16
    Skaperen, sorry for delay in answering, I'd put it on the back burner.

    This is a refining type project, so current limiting is by the size of the electrodes, it is inherent in the cell design. 66V is rather high, as in refining, with cell voltage at ~ 0.5 V, this represents a lot of cells in series. It is not out of the question though, as some kind of series/parallel arrangement is usually employed to be able to use real world voltages, and to avoid the inefficiency of the rectifier volt drop.
    My project is experimental, so I was hoping for the flexibility of something like a 0-24V transformer/rectifier. I was hoping to achieve this by using a 240V 15A variac or two and pulling the secondary off the 30V winding. Thus getting 0-30V at ~115A per winding.

    Just a question re your comment. You mention wiring 2x 400V windings in series, and 2x 230V in parallel. Which are the 400V ones? My understanding is that UK 3 phase is 3 x 240V at 120 phase angle which when multiplied by root3 gives the 415V. Mind you, I'm no great authority on this ! ( On second thoughts, and looking at my first post again, I guess you mean 400V as the primaries, and 230V I mentioned as the secondaries)

    Your series and parallel idea is interesting. My understanding of the way the tranny stands is 415 goes in as 3 x 240 V. Then, ~200V from each of the 3 cores is paralleled up to give the total current. Am I right in thinking that if, from the legends on the tranny, I get say 200V out from each of the 240's in, that if I wire in series the 2 I can use, that I will now get out 2x 100 V because each of the coils now drops 120V, thus halving the secondary voltage?

    Doing the figs., if it is 220V at 200A out, this is ~ 70A per core in parallel. So if it's 240 in and 220 out per core this is 220/240 x 70 = 64 A per phase in. So if I wire 2 primaries in series, and 2 secondaries in parallel I get 110 out at ~140A ?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook