transformer for low frequency power

[electrofelon]

Say I want to use a 60 hertz transformer on a 15 hertz source. My understanding is that the magnetic flux must reach higher levels (for the 15 hertz source compared to the 60 hertz source) which will likely result in core saturation. Am I correct that, for the 15 hz source, using a tranformer with four times the turns on both windings will keep me at the same saturation limits?

Example: 120 volt source both cases
60 hz source, transformer rated for 120 primary, 12 volt secondary, 1kva
15 hz source, transformer rated for 480 primary, 48 volt secondary, 1kva
same saturation limits?

Since for a given transformer lowering the frequency increased the magnetizing current, does this philosphy used in the above example also keep the magnetizing currents equal?

 

[steve066]

I don't think that will work right. 1st of all, I'm not sure if the equations for inductance are linear for the number of turns. I remember Xl=2*pi*f*L which is linear. But I don't remember if "L" is directly proportional to the number of turns on a coil.

So since you are cutting f by 1/4, you want to multiply L by 4. But I'm not sure multiplying the turns by 4 is the same thing.

2nd problem: You are wanting to apply 120V to a 480V transformer. The 480V transformer has smaller wire than the 120V transformer. (Same KVA rating, higher voltage means less current). So by using a 480V transformer, you are probably getting one with smaller wire and an even lower core saturation than the 120V transformer.

3rd problem: I'm not sure the 480V transformer has more turns on it. About all we can say about the 2 transformers you listed is that they have the same turns ratio between primary and secondary, and we can also figure out the current rating for each winding. But nothing says the 480V transformer has more turns than the 120V transformer.

4th problem: more minor and maybe not a real problem: even the iron laminations in the core are designed for a certain frequency (per their size and thickness). You may get more (or less, I'm not really sure off hand) losses from the transformer with a different input frequency.

Think about this: Instead of using a higher voltage transformer, why not use a transformer with the right voltage ratio, but one rated for 4x the KVA. That way you will get more iron, and it won't saturate as fast. I'm not sure it will work right, but it might.

Steve

 

[winnie]

I am giving this answer based upon my strong knowledge of AC induction motors, not direct installation experience with fixed transformers. The underlying physics is the same.

To a very good approximation, core flux is directly proportional to applied voltage and inversely proportional to frequency. This means that if you keep voltage proportional to frequency you can maintain desired core saturation.

If you have a transformer rated at 60Hz, 480V, 1KVA, and you apply 15Hz and 120V, then you will get proper core flux density.

Note however that while you have adjusted the supply voltage to match frequency, and thus are not saturating the core, you have not in any way changed the _current_ rating of the transformer. Since the KVA rating is the product of voltage and current, and you are operating at 1/4 the rated voltage, the 're-rated' power handling capability will only be 1/4 the original rating.

Thus a 60Hz, 480V, 1KVA transformer, applied at 15Hz and 120V will only be capable of operating at 0.25KVA power.

In general, transformer cores are 'good' up to their design frequency, and only have _lower_ losses below the design frequency. So operating the core below its design frequency won't cause excessive losses as long as you reduce the supply voltage appropriately.

Steve, you are on the right track. If you want a transformer rated at 15Hz, 120V, 1KVA, then you need a transformer that would be rated 4KVA at 60Hz. But to get the voltage right you need a 60Hz, 480V, 4KVA device.

-Jon

 

[rattus]

I agree with Winnie, mostly:

I agree with Winnie, mostly:

The primary inductance at 15 Hz will be 1/4 of the primary inductance at 60Hz, therefore the magnetizing current will be the same for 1/4 primary voltage.

However, eddy current loss is proportional to the square of frequency, and hysteresis loss is proportional to frequency, so iron loss would be less at a lower frequency.

Copper loss would be the same.

 

[electrofelon]

Thanks for the replies guys, and sorry for the long delay in a response - I just noticed the thread was replied to. Yes I definitly agree on having to derate the KVA in this situation because the transformer is 'expecting' current based on the KVA and designed voltage, not the derated voltage.

Steve,

My theory is that, all other things being equal, a transformer with a 2X voltage rating, will have twice as many turns as a transformer with an X voltage rating, thus allowing us to use voltage is inversly proportional to frequency to prevent core saturation. I recently obtained an article on transformer design that supports this theory:

Your next step is to find out how many turns of wire will be required for the primary. The formula is: "N = 100,000,000 x E / 4.44 x f x A x Bm"
N is number of turns, E is line volts, 4.44 is a multiplying factor, f is frequency in cycles per second, Bm is maximum flux density in lines per sqin, A is area of core in sqin.
Bm is the value of the flux or magnetic lines of force set up in the core by the primary exciting current. If the density is too high, the transformer will heat excessively and waste power. Various values are selected by a designer according to the use. In some electronic transformers it may be as low as 20,000; in some cases a density of 80,000 has been used, especially for intermittent duty. A value of 60,000 lines is a good average for small transformers.

So If you plug into the formula 60 Hz at 480v, you get four times the turns as 60 hz at 120v - if all other things are equal, which Im not sure is always the case but.... The formula also supports that Hz and voltage are inversly proportional. I think I said all that right - I'm confusing myself again.

 

[brian john]

From the Sola Hevi duty Web page. just from 60hz to 50hz

Can 60 Hz transformers be used on 50 Hz?

A. Yes. 60 Hz transformers can be used on 50 Hz if special precautions are taken. The change in frequency will impact the flux density of the transformer causing it to run hot, as if it were overloaded. To offset this effect, you must decrease the input voltage by approximately 17% (1/6th). This means that a transformer rated for a 480 Volt, 60 Hz input could run at 50 Hz but with a maximum input voltage of 398 volts. On the other hand, 50 Hz transformers can be run on 60 Hz with no ill effects.

http://www.sola-hevi-duty.com/support/faq.html#Q16

 

[steve66]

Steve,

My theory is that, all other things being equal, a transformer with a 2X voltage rating, will have twice as many turns as a transformer with an X voltage rating, thus allowing us to use voltage is inversly proportional to frequency to prevent core saturation. I recently obtained an article on transformer design that supports this theory:

So If you plug into the formula 60 Hz at 480v, you get four times the turns as 60 hz at 120v - if all other things are equal, which Im not sure is always the case but.... The formula also supports that Hz and voltage are inversly proportional. I think I said all that right - I'm confusing myself again.

All that sounds right, as long as you don't forget that for all else being equal, if you change the primary voltage, the area "A" will probably also change.

For example, for the same KVA rating, a transformer will have 4x the current rating at 120V as at 480V. So at 120V, the core probably needs 4x the area.

I'm not sure where that takes you

Steve