3-phase "equalisation" transformer?

[Circlotron]

Is there some kind of three phase transformer that could perform the function of making the voltage of all three phases equal? Assuming the supply has a certain impedance and so is able to be pushed about a bit, e.g an alternator of say 5kVA, could a transformer of sorts take action if the sum of the instantaneous voltage of all three phases at any given moment was not zero, and either inject or draw current from the offending phase? I'm not talking about autotransformer action or a tap switcher, just something that would try and correct voltage imbalance by a corresponding and opposite current imbalance. As mentioned, the source would have to have some sort of impedance for this to work. Definitely not for connection to AC mains!

[NiHaoMike]

https://en.wikipedia.org/wiki/Zigzag_transformer

[soldar]

I do not think the zig-zag transformer does what the OP is asking but I would need more information on what exactly he is trying to do.

If the voltage of one phase, in a three phase system,  sags it means it is loaded more than the other phases. The only way I can think you could maintain voltage is by using (auto)transformer or re-arranging the loads of the other phases.

[Circlotron]

I do not think the zig-zag transformer does what the OP is asking but I would need more information on what exactly he is trying to do.

I'm just messing around on the bench doing experiments, nothing practical.
I have a 50 watt 2-channel audio power amplifier driven by two sine waves 60 deg apart. (yes, 60 deg, not 120 deg) The outputs of the amplifiers drive the primaries of three 28V-240V transformers, the primaries connected in delta with the third connection grounded. The secondaries are also delta connected and driving a ~1/4hp 6 pole induction motor. Rated motor speed is 1000rpm @ 50Hz. I can get it to run to almost 6000  rpm but the voltage is only 150VRMS, not nearly high enough, so raising the frequency further causes the motor to slow down.

The phase voltages are almost perfectly balanced and it is not a problem as such, but what I was looking for is something that would lock the phase voltages together by brute force somehow. If one phase voltage was a little low or high it would pull current from, or send current to the other phases somehow. Electrical engineering is one of those disciplines that appears simple until you start digging into it!

[Circlotron]

Is it possible to make a zig zag transformer using 3 single phase transformers? If so, what would the turns ratio be? 1:1?

[soldar]

Is it possible to make a zig zag transformer using 3 single phase transformers? If so, what would the turns ratio be? 1:1?

If by "single phase" you mean with only two windings then the answer is "no". 

With a zig-zag transformer you can (1) obtain phases which are offset in phase from the incoming phases, (2) rearrange loads to distribute them among phases. You can have a single phase load and distribute it between two or three phases.

As the wikipedia article says, this also affects the harmonics.

[duak]

You should be able to make a zig-zag transformer from three 1:1 transformers.  I don't think it will do what you want though.  Z-Z xformers are generally used to establish a local neutral connection for loads that require them.

Edited: I may have misremembered something here and was thinking of the Scott T connection.  To synthesize a corner grounded delta with two transformers use sine waves 90 degrees apart, not 60.  This should help your top speed.  Intuitively, it makes sense to use 60, but if you redraw the phasor diagram to correctly match the circuit, the angle becomes 90 deg.  ie., vectorially add  the opposite of one of the phasors to each of the existing phasors to determine what the two new phasors are.  I will whip up a sketch when I have a little more time.

If you want to run the motor at a high speed but you don't have enough drive voltage for it, try putting capacitors in series with it.  The capacitors will cancel out some of the inductive reactance and more effectively use the voltage you have.  Note, this will work over a fairly narrow range of frequencies so you will not be able to run the motor as slowly as before.  Also, the capacitance will resonate with the motor inductance at some frequency and voltage across them may build up to damaging levels.  Parallel capacitors can also be helpful, but will cause an increased voltage on the output of the power amplifiers.

[soldar]

You should be able to make a zig-zag transformer from three 1:1 transformers. 

Could you show a schematic?

[duak]

Circlotron, you are correct on the 60 degree phasing.  I had remembered the Scott T connection which has a 90 degree angle.  The attached file shows the phasor diagram I worked out - hopefully I've done it correctly.  The green phasor sprouting out of the center point isn't strictly correct but I couldn't get the CAD package to overlay the colors correctly so I cheated.

Soldar, l whipped up a sketch showing how three 1:1 transformers could be connected to implement a Zig-Zag connection.  These are apparently called Zig-Zag Grounding Banks.  see https://hvdc.ca/uploads/knowledge_base/zig_zag_transformer_for_grounding_system.pdf?t=1539717863   I don't know if they will solve Circlotron's problem though.

[soldar]

Soldar, l whipped up a sketch showing how three 1:1 transformers could be connected to implement a Zig-Zag connection.  These are apparently called Zig-Zag Grounding Banks.  see https://hvdc.ca/uploads/knowledge_base/zig_zag_transformer_for_grounding_system.pdf?t=1539717863   I don't know if they will solve Circlotron's problem though.

I am confused. The PDF shows an application for creating a virtual ground for three phases not an application where three phases come in and three (different) phases come out. That's what I am asking because I can't see how it can be done.

This graphic is from the Wikipedia page. To maintain output voltage you would need three transformers with two secondaries each and not 1:1 (or six transformers).

[filssavi]

What I think is happening is simply the OP is hitting the speed limit given by the combination of motor back EMF,  and maximum drive voltage.

if that is the case, usually two different routes are viable:
1)  usually with native three phase inverters flux weakening is used, where a part of the driving current is used to counteract (and thus lower) the rotor magnetic field, this allows to widen the operating speed range, of course thermal limits still apply and since the part of the current employed to lower the back emf  can't generate torque, the higher the speed in the FW region the lower the torque. However I'm not sure that with your setup (driving just 2 phases and not all three) its doable
2)  use a higher voltage to begin with (in this case higher transformer turn ratio or higher starting voltage)


in any case I strongly doubt the problem is a phase imbalance, also be aware of the mechanical speed limits of bearings, couplers, the rotor itself etc, as you are already 6 times higher than rated spead, and going higher the machine could suffer a rapid unsheduled disassembly

[ogden]

Is there some kind of three phase transformer that could perform the function of making the voltage of all three phases equal?

It is called active load balancing and is done using solid state technologies, not transformers. Just few examples of search results:

http://www.elcopower.com.my/ALB.html
https://www.schneider-electric.com/en/product-range-presentation/63426-accusine-pcs%2B/

[duak]

Soldar, you are correct that a more complicated arrangement is probably needed.  A few years ago I had looked into using a Zig-Zag Grounding Bank for a possible solution for an ungrounded delta circuit but never tried it out.  If my understanding of the zig-zag connection is correct, it wil also try to coerce a three phase circuit into balance because it has a different impedance to imbalance currents.   I wasn't sure that my idea would solve Circlotron's problem and said so.  The more complex circuit, as shown in the Wikipedia article, generates a better balanced set of phases rather than trying to coerce them into balance.

[Kleinstein]

The zig-zag transformer might improve the balance a little, but it is not 100% coupling.

Anyway the problem with the motor should have nothing todo with imbalance, more with insufficient voltage. Usually one needs to increase the voltage if a higher frequency is used - often a voltage about proportional to frequency is used.

If a higher speed is wanted, I would not start with a rare 1000 U/min motor, but with a more normal 2900 U/min (at 50Hz) motor, so already 3 times the speed. A US motor for the lower 208 V and 60 Hz could also be a good idea. There are also special motors for higher frequencies: something like 200 V at 300 Hz is sometimes used.

[NiHaoMike]

There are a lot of BLDC motors that will operate that those range of speeds.

[Circlotron]

The zig-zag transformer might improve the balance a little, but it is not 100% coupling.

Anyway the problem with the motor should have nothing todo with imbalance, more with insufficient voltage. Usually one needs to increase the voltage if a higher frequency is used - often a voltage about proportional to frequency is used.

If a higher speed is wanted, I would not start with a rare 1000 U/min motor, but with a more normal 2900 U/min (at 50Hz) motor, so already 3 times the speed. A US motor for the lower 208 V and 60 Hz could also be a good idea. There are also special motors for higher frequencies: something like 200 V at 300 Hz is sometimes used.

Two things - I’m just messing around with a junk motor I pulled out of the rubbish at work. Trying to see how fast I can make it go, not actually trying to do anything useful with it. I know that it isn’t being fed enough voltage, especially at high speed. Second, I know a zig zag transformer won’t make the motor run any better with my particular test setup. I was simply thinking about 3-phase stuff and I wondered if there was a transformer of sorts that would oppose any unbalanced voltages applied to it. Similar in principle to a transformer that has two 120v windings and you connect each winding to a separate 120vac source, both sources in phase. The transformer will draw current from the source that is slightly higher in voltage and supply current to the source that is slightly lower in voltage. Just wondering if there was a transformer that could do the same thing with 3 different sources of approximately the same voltage but this time separated by 120 degrees.