# Transformer Question

• posted

Someone asked me a question the other day and gave me these specs about a transformer and I did not know the answer.

They gave me the scenario of a 1000VA transformer with losses of 5% core losses and 3% copper losses. I told them that the losses in the core are due to voltage and frequency and that the losses in the copper are due to amperage. I also told them that eddy currents cause core losses, but that these are minimized by the the laminations of the transformer.

Since these losses are related to the VA of the transformer I just took

8% of 1000 and assumed the total losses to be 80VA. I assume that this is correct.

He then asked about what if the transformer was loaded to its maximum. This is the part I was unsure of, but going back to my memories of AC circuit analysis I told him that I thought that the VA losses remained the same for the core and copper losses since the input to the primary does not change. However, the voltages change depending on whether the transformer is a step up or down and that has to do with the number of turns. I also told him that the amperage will be different in each side of the transformer depending on if it is a step up or down but that this did not effect the core and copper losses.

Was I completely incorrect in this assumption?

Thanks, Steve

Loading thread data ...
• posted

Core losses are a function of core material, core design, and frequency. With certain exceptions, core loss tends to be fairly constant regardless of the load on the transformer. That is why an energized but unloaded transformer still gets quite warm. Copper loss is simply a function of Ohm's law (I squared R). Copper loss would be near zero on an unloaded transformer and at its maximum with a fully loaded transformer.

A transformer designed for step-down will typically have heavier copper in the secondary to efficiently accommodate the larger current, while a step-up transformer would normally have the reverse design.

Vaughn

• posted

Without cracking a book I will venture this opinion:

Transformers are speck'ed fully loaded. That is to say the stated losses occur at rated load. Eddy currents and magnetic hysteresis account for the

5% core loss; 3% copper loss is the combined loss of both primary and secondary.

Bob Swinney

• posted

Of course it changes. More output current draws more input current.

• posted

Actually, core losses don't change much until the core undergoes saturation, then they go sky high.

But copper loss increases more or less directly with current, but exacerbated by temperature rise.

LLoyd

• posted

In article , Don Foreman wrote: : :Copper loss will increase as the square of load current.

Faster than that, actually. As the transformer heats up, the winding resistance rises, further increasing the losses. That can cause an overloaded transformer to go "over the hill" pretty rapidly.

• posted

Copper loss increases as square of current, further axacerbated by increased resistance due to temp rise as noted.

Peak core flux density changes very little with load; it actually goes down a bit with load due to IR drop in the primary. It is determined by primary voltage and frequency. It it isn't saturated under noload condx then it won't saturate under load. Core loss may change slightly with core temperature, not sure how that works with silicon steel.

• posted

Don.

1) I know the I^2R law. I said "more or less directly" because there are several things that can change it, among them being the frequency of AC being transformed, and interwinding/inter-turn capacitance. Even though it's usually lumped together as "copper loss", it's not usually _only_ the DCR of the wire they're accounting for in power transformers. They break all that stuff out when spec'ing audio and RF transformers, but seldom 50-400Hz stuff. 2) Many transformers operate in rectifier circuits, and may have some DC component in the secondary, as well as AC. Increasing the DC can indeed induce saturation of the core. You know that.

(Prototypeing lab supervisor, Florida Transformer corp., div of Transitron,

1968)

LLoyd

• posted

Good point about rectification. I did forget about that, and that would indeed be load-dependent. Thanks.

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.