Favourite connection method for two phase mains?

[Faringdon]

Hi,
Which of the attached is your favourite connection method for 2 phase mains? (LTspice and jpeg attached)
The 6 diode rectifier gives you a much lower peak mains current, much smaller output capacitor, and much less ripple current in the capacitor.

[TimFox]

If you have all three phase wires available, the 6-diode rectifier gives you much less ripple for a given filter capacitance than the 4-diode rectifier (full-wave bridge) operating from two phases.
Not all customers for power supplies have all three phases available, so you need to check with the customer.
I once needed the customer's electricians to install a three-phase outlet, since my boss hadn't asked me and told the customer that single-phase mains sufficed for the high-power installation.

[Faringdon]

Thanks, and as you know, the above uses only two phases with a 6 diode rectifier. Which is best, right , or left?

[TimFox]

Your drawing on the left is a conventional three-phase rectifier.
That on the right is a conventional full-wave bridge.

[Faringdon]

Thanks, yes, but the 3 phase rect only has two phases going into it.....and its better than the one on the right.
Wondering why its not used for 2 phase systems?

[gnuarm]

Thanks, yes, but the 3 phase rect only has two phases going into it.....and its better than the one on the right.
Wondering why its not used for 2 phase systems?

Which phase is not going into the device in the left diagram???  I'm not following you at all on this.

[TimFox]

The three nodes "A", "B", and "C" feeding the rectifiers in the left diagram are three phases of a three-phase distribution.
(The neutral node is not needed for this rectifier.)

[gnuarm]

The three nodes "A", "B", and "C" feeding the rectifiers in the left diagram are three phases of a three-phase distribution.
(The neutral node is not needed for this rectifier.)

I believe the OP thinks that because he is only showing two voltage sources, that he is using two phase, rather than three phase. 

[TimFox]

In real life, his left drawing would be fed from a "delta" or "wye" transformer secondary from three-phase mains.
For the full-wave rectifier circuit as drawn, with a grounded load, no connection to the center of the wye is required, but of course the secondary windings are isolated from ground.
If the feed is not isolated from "ground", then the load must be galvanically isolated, such as a motor drive where the motor windings are insulated from ground.
Often, the "bus" voltage across the filter capacitor is not grounded, but coupled to a grounded load through a high-frequency transformer as part of the switching power supply.
Hint:  120^o is a clue for the existence of three-phase voltage.

[uer166]

"2 phase" should be same as single phase, you should get the same ripple and issues as a 1-phase supply input stage..

[nctnico]

Thanks, yes, but the 3 phase rect only has two phases going into it.....and its better than the one on the right.
Wondering why its not used for 2 phase systems?

Which phase is not going into the device in the left diagram???  I'm not following you at all on this.

The schematic is perfectly clear about this! Hint: in a 3 phase mains system, the phases are shifted by 120 degrees

The problem I see is that the device will see an odd variety of pulses and it will load the grid in a bad way (especially when the device draws a lot of power). IMHO you shouldn't design equipment like this. Either use single phase or 3 phase.

[gnuarm]

Thanks, yes, but the 3 phase rect only has two phases going into it.....and its better than the one on the right.
Wondering why its not used for 2 phase systems?

Which phase is not going into the device in the left diagram???  I'm not following you at all on this.

The schematic is perfectly clear about this! Hint: in a 3 phase mains system, the phases are shifted by 120 degrees

I don't think you understand what I asked.  It doesn't matter.  It was not directed to you.

[IanB]

Thanks, yes, but the 3 phase rect only has two phases going into it.....and its better than the one on the right.
Wondering why its not used for 2 phase systems?

The one on the left has three phases going into it (as said by other respondents). If you have three wires, each spaced 120 degrees apart, then you have three phases. There is no way to get these three wires from a three phase mains supply without having three phases present (unless you do something dumb like letting the three wires be N, L1 and L2).

[multipla]

My favourite connection method for 2 phase mains is as attached, it gives you even less ripple and because 8 it a multiple of 2 and 4 (as opposed to 6) you can do the layout even more symmetrical (better).
Would you agree?

[nctnico]

That is very clever! 

[gnuarm]

My favourite connection method for 2 phase mains is as attached, it gives you even less ripple and because 8 it a multiple of 2 and 4 (as opposed to 6) you can do the layout even more symmetrical (better).
Would you agree?

I can't tell what parts of this thread are serious and which are jokes. 

Where do you get the two AC phases aligned this way?  The original approach was using 3 phase power and calling it two phase.  If you really want to use two independent AC lines like this to minimize ripple, wouldn't they need to be 90° out of phase, not 180°? 

This way the four angles produce peaks of conducted current at 0°, 120°, 180° and 300°.  Clearly not optimal.  Better is 0°, 90°, 180° and 270°. 

Or am I missing a sarcastic post?

[nctnico]

Or am I missing a sarcastic post?

Definitely! 

[ConKbot]

Feed the two phase plus neutral into a 3 phase transformer (oversize to handle the dropped phase) and use phase shifted windings to get both a star and a delta output, and feed that into a 12 pulse rectifier. This reduces ripple currents even further.

[multipla]

Or am I missing a sarcastic post?

Yeah... And I extra signed up for this post! But I've been reading for quite some time already

[Faringdon]

I believe the OP thinks that because he is only showing two voltage sources, that he is using two phase, rather than three phase.

Thanks, yes thats right...if there are only two voltage sources, then it cant possibly be 3 called phase...even though there might actually be a third phase somewhere back in the system.

The attached LTspice (& jpeg)  shows a three phase delta supply into a 6 diode rectifier...at t = 100ms, one of the three phases gets totally switched out...leaving only the other 2 phases supplying current......as can be seen, the currents in the rectifier remain exactly the same  before and after t = 100ms....the load "knows nothing" of the supply system calamity that has just occurred.....and there is no way (inside the device with the rectifier)  to detect that a phase just got dropped.

This is definetely a 2 phase system after t = 100ms, since there are only two voltage sources remaining after the third gets switched out.

The problem I see is that the device will see an odd variety of pulses and it will load the grid in a bad way (especially when the device draws a lot of power). IMHO you shouldn't design equipment like this. Either use single phase or 3 phase.

Thanks yes, but 2-phase is a bona fide system....its used widely in the USA. USA streetlighting i think uses 2 phase (?)...multiple runs of streetlights share out the pairs of phases such that "overall", the overall 3 phase system is balanced.....even though each individual streetlight only gets 2 phases.....my betting is that 2-phase USA streetlights use the one on the left in the top post?

[langwadt]

Thanks, yes, but the 3 phase rect only has two phases going into it.....and its better than the one on the right.
Wondering why its not used for 2 phase systems?

The one on the left has three phases going into it (as said by other respondents). If you have three wires, each spaced 120 degrees apart, then you have three phases. There is no way to get these three wires from a three phase mains supply without having three phases present (unless you do something dumb like letting the three wires be N, L1 and L2).

https://youtu.be/quABfe4Ev3s

[nctnico]

Thanks yes, but 2-phase is a bona fide system....its used widely in the USA. USA streetlighting i think uses 2 phase (?)...multiple runs of streetlights share out the pairs of phases such that "overall", the overall 3 phase system is balanced.....even though each individual streetlight only gets 2 phases.....my betting is that 2-phase USA streetlights use the one on the left in the top post?

But ask yourself: how many wires go to each lamp post? And how is that different from a phase + neutral system?

[Faringdon]

Thanks, but suppose you have two 3 phase generators, and 3 runs of streetlights to power.......then its best to use three two phase "doublets" and supply a phase pair  to each streetlight run...rather than one 3 phase genny for one streetlight run, and the other 3 phase genny doing the other two streetlight runs....the load is better shared with the method of the top post.....the  "2 phase method"...the load on the gennys would be more shared.

[uer166]

I believe the OP thinks that because he is only showing two voltage sources, that he is using two phase, rather than three phase.

Thanks, yes thats right...if there are only two voltage sources, then it cant possibly be 3 called phase...even though there might actually be a third phase somewhere back in the system.

The attached LTspice (& jpeg)  shows a three phase delta supply into a 6 diode rectifier...at t = 100ms, one of the three phases gets totally switched out...leaving only the other 2 phases supplying current......as can be seen, the currents in the rectifier remain exactly the same  before and after t = 100ms....the load "knows nothing" of the supply system calamity that has just occurred.....and there is no way (inside the device with the rectifier)  to detect that a phase just got dropped.

This is definetely a 2 phase system after t = 100ms, since there are only two voltage sources remaining after the third gets switched out.

The problem I see is that the device will see an odd variety of pulses and it will load the grid in a bad way (especially when the device draws a lot of power). IMHO you shouldn't design equipment like this. Either use single phase or 3 phase.

Thanks yes, but 2-phase is a bona fide system....its used widely in the USA. USA streetlighting i think uses 2 phase (?)...multiple runs of streetlights share out the pairs of phases such that "overall", the overall 3 phase system is balanced.....even though each individual streetlight only gets 2 phases.....my betting is that 2-phase USA streetlights use the one on the left in the top post?

Then show me how you can reduce ripple with the 2-phase American system (whether it's split phase or 2-out-of-3). Hint: you can't because in those systems you don't have neutral access, and in split phase it wouldn't help anyway. If you're running 3 wires it's no longer a "2 phase system" by the usual definition. "2 phase plus Neutral", sure, but at that point just friggin use all 3 available phases.

[langwadt]

I believe the OP thinks that because he is only showing two voltage sources, that he is using two phase, rather than three phase.

Thanks, yes thats right...if there are only two voltage sources, then it cant possibly be 3 called phase...even though there might actually be a third phase somewhere back in the system.

The attached LTspice (& jpeg)  shows a three phase delta supply into a 6 diode rectifier...at t = 100ms, one of the three phases gets totally switched out...leaving only the other 2 phases supplying current......as can be seen, the currents in the rectifier remain exactly the same  before and after t = 100ms....the load "knows nothing" of the supply system calamity that has just occurred.....and there is no way (inside the device with the rectifier)  to detect that a phase just got dropped.

This is definetely a 2 phase system after t = 100ms, since there are only two voltage sources remaining after the third gets switched out.

The problem I see is that the device will see an odd variety of pulses and it will load the grid in a bad way (especially when the device draws a lot of power). IMHO you shouldn't design equipment like this. Either use single phase or 3 phase.

Thanks yes, but 2-phase is a bona fide system....its used widely in the USA. USA streetlighting i think uses 2 phase (?)...multiple runs of streetlights share out the pairs of phases such that "overall", the overall 3 phase system is balanced.....even though each individual streetlight only gets 2 phases.....my betting is that 2-phase USA streetlights use the one on the left in the top post?

Then show me how you can reduce ripple with the 2-phase American system (whether it's split phase or 2-out-of-3). Hint: you can't because in those systems you don't have neutral access, and in split phase it wouldn't help anyway. If you're running 3 wires it's no longer a "2 phase system" by the usual definition. "2 phase plus Neutral", sure, but at that point just friggin use all 3 available phases.

you need a minimum of 3 wires for two phases, and split phase is not two phases

[nctnico]

Thanks, but suppose you have two 3 phase generators, and 3 runs of streetlights to power.......then its best to use three two phase "doublets" and supply a phase pair  to each streetlight run...rather than one 3 phase genny for one streetlight run, and the other 3 phase genny doing the other two streetlight runs....the load is better shared with the method of the top post.....the  "2 phase method"...the load on the gennys would be more shared.

How many wires to each lamp?

[IanB]

Thanks, but suppose you have two 3 phase generators, and 3 runs of streetlights to power.......then its best to use three two phase "doublets" and supply a phase pair  to each streetlight run...rather than one 3 phase genny for one streetlight run, and the other 3 phase genny doing the other two streetlight runs....the load is better shared with the method of the top post.....the  "2 phase method"...the load on the gennys would be more shared.

Where you apparently live (in the UK), three phase power distribution at low voltage to consumers is the common system.

What would commonly happen with streetlighting is that they would be on a three phase substation transformer (possibly the same one supplying nearby houses). The distribution is on four wires: three phases plus neutral (star or Y arrangement). All consumers get the neutral, then the three phases are distributed evenly between the consumers to balance out the load.

In the case of streetlights, you could label each light alternately A, B and C. All the A lights would get the first phase, the B lights would get the second phase, and the C lights would get the third phase. All the lights A, B and C would get the common neutral.

Then you would have four wires in total rather than six wires or more. This is the reason three phase distribution exists.

Why on earth would you imagine streetlights being powered by a generator? But even if it were a three phase generator, the four wire distribution would still be used.

[johansen]

This topic has come up on a certain machinist forum where its  not entirely uncommon for folks in NZ to be  served with 2 phases of 240/415.. but not the third phase.

As such you can sort of run a Y motor with its  normally hidden Y point connected to neutral on 2 out of 3 phases. Some people say they do self start. The one test i was able  to perform with a motor on 120/208.. it would not self start on two phases.

Two 240/240v single phase transformers with each primary connected line to neutral , and secondaries in series, one end connected to neutral can be used to create the missing third phase.



As for all the diagrams in this thread. No one has isolated 2 phases of 240v 120 degrees apart. Ether you have open delta 240v or you have 240/415 2 out of 3 phases. As such you have either a 3 phase rectifier or you have a single phase 415v rectifier.

There is no magical way to get around this problem

[IanB]

I think it is only sensible to label the conductors N, L1, L2 and L3 before talking about them.

Then if you have any two of those four conductors, you have a single phase supply. If it's a pair of line conductors you have line voltage, if it is one live and the neutral you have the phase voltage.

If you have the three line conductors (L1, L2, L3) then you have a three phase supply at line voltage.

If you have two of the three line conductors and the neutral (say L1, L2 and N), then you have two phases at phase voltage and one phase at line voltage. This arrangement would not be ideal for any significant loads as it would tend to unbalance the supply. If you absolutely needed to get three equal phases from this then it could be done, but it would always be a sub-optimal solution compared to having the original L3 available to you.

[MF-jockey]

Even if only two phases are shown, the connection of the two phases on the left results in a 3-phase system.



https://americas.hammondpowersolutions.com/en/resources/faq/definition/open-delta-transformer

[johansen]

If the 2 out of 3 phases are delta, i already explained you have 3 phaae.

If its y, you will have 415v single phase unless you use transformers to load the neutral and produce the missing third phase.

Many american farm lands were fes 3 phase delta with just 2 wires and a neutral, and 2 transformers open delta on the pole.

[Faringdon]

Many american farm lands were fed 3 phase delta with just 2 wires and a neutral, and 2 transformers open delta on the pole.

Presumably you're saying "2 single phase transformers connected open delta"

...so it no doubt results in savings as there's only two single phase  transformers...giving an albeit reduced power " 3 phase system".
So i wonder  if this is widespread in the US?....because you could do multiple lots of them so overall, the originating 3 phase back at the power station sees a balanced load.

I bet two single phase distribution transformers connected open delta is a lot cheaper than a 3 phase delta distribution transformer

[amyk]

https://en.wikipedia.org/wiki/Two-phase_electric_power

This is a weird thread. Did no one notice that two-phase power is almost extinct?

[langwadt]

https://en.wikipedia.org/wiki/Two-phase_electric_power

This is a weird thread. Did no one notice that two-phase power is almost extinct?

90 degrees two-phase is almost extinct,  two out of three 120 degree phases not

[Faringdon]

The real 2 phase power system, which lost out to 3 phase years ago, was two phases 180 degree apart. You just couldnt use it phase to phase...as it was zero.
But it was convenient to wind generators to produce it.

[langwadt]

The real 2 phase power system, which lost out to 3 phase years ago, was two phases 180 degree apart. You just couldnt use it phase to phase...as it was zero.
But it was convenient to wind generators to produce it.

180 degrees apart is not two phases, it is split phase and hot-to-hot it is twice the voltage hot-to-neutral

[MF-jockey]

Supplementary consideration of the right circuit:
In the drawing on the right, the two phases are connected in such a way that only the simple phase voltage results. This gives no advantage over just one phase, just double the resistance and reactance of the voltage source. Only the power is distributed between the two phases with the disadvantage of a poor power factor (current not in phase with voltage).
If you connect the two phases differently, as in the second picture below, then the voltage is increased by a factor of about 1.732.

You can also rotate the connections if you define V3 in the first picture with SINE(0 141 60 0 0 300), the phase is shifted by 180 degrees. Then the graphic of the arrows will also be conclusive.

[Faringdon]

Thanks,
So presumably, if we just want a few watts delta three phase supply, as a test input to some kit, then we should do the "2phase 3 phase" version...as it will be the same and a lot simpler.?....
LTspice and jpeg attached

[dmills]

FFS just draw the phasor diagrams all becomes instantly clear.

A small three phase supply is easily fabricated from ebay junk, here is how I got 1.5kVA for my lathe:

240V (1p) -> 1.5kVA control transformer wired backwards to get 400V (1p), it buzzes a little due to me pushing the core slightly hard, but seems fine given the use case.
400V (1p) -> 400V motor inverter set to 50Hz, makes a rather rough PWM three phase delta.
Finally a 400V three phase 'sine wave filter' from an airconditioning place makes a reasonably nice sine.

Bit clunky but it gets it done. Id does deliberately drop frequency on overload as the output voltage collapses during a hot start on the lathe motor in high speed mode, but does recover once the speed picks up, a peculiarity of the drive I am using.

I would note that hooking an impulse generator to the output of this thing to investigate things like transient immunity is strongly contra indicated, I would expect smoke from the drive if attempted.   

Oh on the subject of impulses, much of the three phase you will encounter is MUCH stiffer then most domestic power, arc flash is a real consideration if close to a substation transformer, size your circuit protection to suit the PSC available, and don't guess.

[djsb]

I've not read all the posts in this topic, but my first impression was that "Favourite" should not be the first criteria when choosing something as important as mains power connection. Surely safety and legality should be the first consideration. Then I saw who the original author was. However, I must confess my ignorance on this topic as I mostly work with small signal stuff. So I shall read the rest of the posts anyway and maybe learn something.

[Faringdon]

400V (1p) -> 400V motor inverter set to 50Hz, makes a rather rough PWM three phase delta.

Thanks , i googled this, and came up with this..
https://www.approvedhydraulics.co.uk/products/tec-3-phase-electric-motors-2-pole-3000rpm-ie2-high-efficiency?pr_prod_strat=use_description&pr_rec_id=522d513e6&pr_rec_pid=12527820819&pr_ref_pid=12518329235&pr_seq=uniform&variant=52665033491

Though i confess i still dont see where the 3 phase delta can be gotten?

I can imagine something like a three phase induction motor getting spun round at the right speed, and then its coils, if connected in delta, would give the 3 phase delta supply at 50Hz?

Simulating 3 phase delta supplies, shows them to be pretty awful...if the phase angles, and amplitudes, are not really well matched, then very high circulating currents occur in the delta. I find it difficult to see why anyone would want to use them, instead of the star , or Y type?

[dmills]

You don't need the motor, just the inverter drive.

There is actually a trick you can do with a motor alone, connect the mains between two of the terminals, and a 22uF or so 'motor run cap between one of the supply terminals and the third terminal, apply power, the motor will spin up but produce very little torque.

What it does do is make reasonably good three phase delta, effectively operating as a rotary transformer.

Yea, acoustically noisy, and not pretty but plenty of home machine shops run off such things, key phrase is 'Rotary phase converter', personally I prefer electronically commutated motor drives, quieter, more efficient and much, much lighter, but whatever works.

FWIW my lathe in the garage runs off a 1500VA 240V-400V 'control transformer' wired backwards to make 400V single phase, which I stuff into a small IMO controls VFD to make three phase, with a suitable line reactor and some filter caps you don't even get much switching residual on the waveform, and the drive has some limited ability to operate in regen with a dump resistor.

[Faringdon]

There is actually a trick you can do with a motor alone, connect the mains between two of the terminals, and a 22uF or so 'motor run cap between one of the supply terminals and the third terminal, apply power, the motor will spin up but produce very little torque.

Thanks, we could just buy a cheap 3 phase induction motor and do this.
We  would still need to generate a single phase mains at 60Hz, but from then onwards it gets easier.
We need 3 phase delta at 60Hz,  85-125VAC.