Paralleling of transformers
- Thread starter electrics
- Start date
- Status
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
110606-0633 EDT
electrics:
The question is simple, but don't consider doing it.
Even if you think the two source voltages would track each other in amplitude they won't.
Consider two essentially identical transformers in parallel. There will be minor differences so the output voltages will not be perfectly identical. Unloaded there will be some small circulating current. Not desirable, but probably of no great consequence. Loaded this disappears, but there will be some minor difference in load distribution. Again probably of no real consequence.
The big problem in what you proposed is that the two source voltages won't precisely track each other compounded by the inherent transformer differences seen on the output side relative to coupling coefficient, and other factors that influence the equivalent internal impedance of the transformers as viewed from the secondary side.
Transformers may have an approximate open circuit secondary output voltage that is based on the turns ratio, but it is not precisely this ratio because all flux lines coupling the primary do not couple the secondary. The lost flux lines are called leakage flux and contribute to a modification of the coupling coefficient, and the series equivalent leakage inductance between primary and secondary.
.
electrics:
The question is simple, but don't consider doing it.
Even if you think the two source voltages would track each other in amplitude they won't.
Consider two essentially identical transformers in parallel. There will be minor differences so the output voltages will not be perfectly identical. Unloaded there will be some small circulating current. Not desirable, but probably of no great consequence. Loaded this disappears, but there will be some minor difference in load distribution. Again probably of no real consequence.
The big problem in what you proposed is that the two source voltages won't precisely track each other compounded by the inherent transformer differences seen on the output side relative to coupling coefficient, and other factors that influence the equivalent internal impedance of the transformers as viewed from the secondary side.
Transformers may have an approximate open circuit secondary output voltage that is based on the turns ratio, but it is not precisely this ratio because all flux lines coupling the primary do not couple the secondary. The lost flux lines are called leakage flux and contribute to a modification of the coupling coefficient, and the series equivalent leakage inductance between primary and secondary.
.
Last edited by a moderator:
what do you mean with coupling coefficient here? I know what this term is but please let me understand what you mean here. I think you mean transformers must have the same characteristics alltogether? Or what is the relation between coupling coefficient different than 1 and the fact that different turn ratios cause the paralleling to be impossible (if so).
kingpb
Senior Member
- Location
- SE USA as far as you can go
- Occupation
- Engineer, Registered
Man, that was a lot of fancy words in one post.........:roll:
I would agree it is not a good idea, however, it can be done, maybe.
If the impedance of the transformers are close to the same, the better. Circulating currents will occur which will cause heating in the transformers. So, if they are lightly loaded than should not be a significant cause for alarm. However, as the loading increases, you may encounter undesirable operating temperatures.
Since the reason for the question is unknown, the answer is vague.
I would agree it is not a good idea, however, it can be done, maybe.
If the impedance of the transformers are close to the same, the better. Circulating currents will occur which will cause heating in the transformers. So, if they are lightly loaded than should not be a significant cause for alarm. However, as the loading increases, you may encounter undesirable operating temperatures.
Since the reason for the question is unknown, the answer is vague.
I think it is a fancy way of thinking that you think I dont know the reason of my asking this question bro.
I have two transformer which has different input voltages but have the same output (intrinsically different turn ratio they have) and I want them to operate paralleled, so strange that you can think it can not have reason, in practical I have such a situation and I wonder if we can operate them paralleled so that one transformer feeds the other transformers' loads if one of them is off and if the loads in total are light enough. I think you could understand a bit...
By the way I am asking what is the relatiob between loading ratio (two guy said "as the load increases .... ") and the drawback which causes the paralleling to be impossible, still wondering and waiting for the right answer.
I have two transformer which has different input voltages but have the same output (intrinsically different turn ratio they have) and I want them to operate paralleled, so strange that you can think it can not have reason, in practical I have such a situation and I wonder if we can operate them paralleled so that one transformer feeds the other transformers' loads if one of them is off and if the loads in total are light enough. I think you could understand a bit...
By the way I am asking what is the relatiob between loading ratio (two guy said "as the load increases .... ") and the drawback which causes the paralleling to be impossible, still wondering and waiting for the right answer.
Last edited:
hurk27
Senior Member
- Location
- Portage, Indiana NEC: 2008
First we must keep in mind that know one knows how much one knows or can we read ones mind to know what one is thinking.
With that said, it seems you are trying to design a back up power system to take the load if one system goes down by supplying the load through another transformer that is fed from a different system, the problem is a transformer will transform in both directions, and if you were to parallel the two secondaries together you will create a dangerous situation, as the powered transformer will now try to supply the system that is down, the only safe and code compliant way to do this would be to use a transfer switch, some automatic transfer switch's are fast enough to transfer without dropping out power.
kingpb
Senior Member
- Location
- SE USA as far as you can go
- Occupation
- Engineer, Registered
Just as any impedance in a parallel circuit the current will be shared proportionately. The closer the impedance is the same in both, the more closely load will be shared equally. Any difference in impedance will account for circulating current, which will increase your upstream current draw, but not do any work, e.g. lost in heat.
Possibly of more concern, is that you can significantly increase the fault current on the bus because you have two transformers connected to it in parallel. A fault study should be done. I would also recommend verifying your protection is still coordinated.
Possibly of more concern, is that you can significantly increase the fault current on the bus because you have two transformers connected to it in parallel. A fault study should be done. I would also recommend verifying your protection is still coordinated.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
110606-1021 EDT
electrics:
In a transformer with a magnetic core the primary coil creates a magnetic field most of which is contained within the core, but some is outside the core. Same with respect to the secondary if the secondary was excited. The magnetic flux that is not common to both both coils does not contribute to induction of voltage in the other coil.
The induced voltage in a coil is v = N * df/dt
where
v is the instantaneous voltage at the coil terminals
N is the number of turns on the coil
df/dt is a calculus term but means the rate of change of flux with respect to time.
For a transformer we can write two equations
vp = Np * dfp/dt
vs = Ns * dfs/dt
An ideal transformer has dfp/dt = dfs/dt and thus
vp/vs = Np/Ns or
vp/vs = turns ratio
No real transformer has all of the primary flux lines coupling the secondary and thus dfs/dt is some amount smaller than dfp/dt.
So
vp/vs = K*Np/Ns where K is somewhat larger than 1 because dfs/dt is slightly smaller than dfp/dt. In special cases with a toroid core of very high permeability in a limited frequency range it is possible to come quite close to 1.
None of this says anything about loaded conditions, coil resistance, and capacitance.
Coupling coefficient is essentially the value of 1/K. In other words the percentage of primary flux that links the secondary. Buying off the shelf transformers with different turns ratios will make it very difficult to get transformers which will balance. Many transformers may not be defined by turns ratio, but by voltage ratio at full load.
kingpb:
If you buy two identical transformers of the same manufacturing batch and correctly phase and parallel them, then there will be some small circulating current at no load. As soon as a small load is applied there is no more circulating current. Rather all the current from each transformer goes to the load.
Where a problem occurs is when full load is applied and the internal transformer impedances are not close to the same value, then unequal current is drawn from the transformers. Thus, unequal heating. If the maximum load is adjusted so that neither transformer has its maximum current exceeded, then no problem.
Probably the greatest problem with what electrics wants to do is the likely unbalance in the relationship of the two different voltage sources.
.
electrics:
In a transformer with a magnetic core the primary coil creates a magnetic field most of which is contained within the core, but some is outside the core. Same with respect to the secondary if the secondary was excited. The magnetic flux that is not common to both both coils does not contribute to induction of voltage in the other coil.
The induced voltage in a coil is v = N * df/dt
where
v is the instantaneous voltage at the coil terminals
N is the number of turns on the coil
df/dt is a calculus term but means the rate of change of flux with respect to time.
For a transformer we can write two equations
vp = Np * dfp/dt
vs = Ns * dfs/dt
An ideal transformer has dfp/dt = dfs/dt and thus
vp/vs = Np/Ns or
vp/vs = turns ratio
No real transformer has all of the primary flux lines coupling the secondary and thus dfs/dt is some amount smaller than dfp/dt.
So
vp/vs = K*Np/Ns where K is somewhat larger than 1 because dfs/dt is slightly smaller than dfp/dt. In special cases with a toroid core of very high permeability in a limited frequency range it is possible to come quite close to 1.
None of this says anything about loaded conditions, coil resistance, and capacitance.
Coupling coefficient is essentially the value of 1/K. In other words the percentage of primary flux that links the secondary. Buying off the shelf transformers with different turns ratios will make it very difficult to get transformers which will balance. Many transformers may not be defined by turns ratio, but by voltage ratio at full load.
kingpb:
If you buy two identical transformers of the same manufacturing batch and correctly phase and parallel them, then there will be some small circulating current at no load. As soon as a small load is applied there is no more circulating current. Rather all the current from each transformer goes to the load.
Where a problem occurs is when full load is applied and the internal transformer impedances are not close to the same value, then unequal current is drawn from the transformers. Thus, unequal heating. If the maximum load is adjusted so that neither transformer has its maximum current exceeded, then no problem.
Probably the greatest problem with what electrics wants to do is the likely unbalance in the relationship of the two different voltage sources.
.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
110606-1126 EDT
electrics:
Now that you have somewhat defined your purpose, then hurk27's answer is probably your best solution, a fast transfer switch.
Another technique could be:
An output generator driven by two motors, one from each voltage source. The motors would be coupled to the generator via one-way clutches, or if all were mechanically locked together, then a way to disconnect a motor input if its supply was lost.
.
electrics:
Now that you have somewhat defined your purpose, then hurk27's answer is probably your best solution, a fast transfer switch.
Another technique could be:
An output generator driven by two motors, one from each voltage source. The motors would be coupled to the generator via one-way clutches, or if all were mechanically locked together, then a way to disconnect a motor input if its supply was lost.
.
a transfer switch ? we just need to make the breaker of the one of which energy is gone be off so that no backfeeding occurs, isnt it? transfer switch makes the two of the xformers operate at the same time impossible and in the case you tell there is no such a situ. So I think there is no any drawback for the operation of two transformers which have different input voltages together so that we can maintain a realiability level in the system.
kingpb
Senior Member
- Location
- SE USA as far as you can go
- Occupation
- Engineer, Registered
electrics, I'm trying to understand your operating situation more clearly.
Are you wanting to operate in parallel all the time, or only momentarily when you switch from one transformer to the other?
I will reserve further comments until I can understand your situation better.
Are you wanting to operate in parallel all the time, or only momentarily when you switch from one transformer to the other?
I will reserve further comments until I can understand your situation better.
- Status