500 kHz-500 MHz current transformer

[JBE]

I am designing a current transformer that maybe cannot be done using a single core without switching. I am considering a single- turn primary winding with the secondary winding loaded with a low resistance. I request some advice regarding its design.

I have references indicating that in order to have a maximum accuracy in the ratio and phase angle of the output current the exciting current must be as small as possible. It looks like using a core material with a high permeability and low core loss can do it.

I have observed that some ferrite cores with high permeability (i.e. 300) do not look feasible above maybe 100 MHz. Complex permeability real and imaginary parts cross over at relatively low frequency. Apparently then, the phase error becomes around 45 degrees. It looks like the higher the complex permeability crossing point the lower the operating frequency.

Even though I have references of NiZn ferrite used for RF tuning wideband and balun transformers with permeability around 18 with permeability crossing around 250 MHz.

On the other hand I have seen some iron powder cores ranged to work even up to 500 MHz, from 50 MHz with a permeability of 1.

My question could be if cores with permeability below 300 would work properly as a current transformer (trying to reach the ideal no phase error output current).

May somebody please give me some advices, notes or simulation tools for this 500 kHz-500 MHz current transformer?

Thank you in advance.

 

[GOTWW]

Are You designing for maximum power transfer or just a linear sample? The flux linkage increases with the square of F and you are going over 3-octives of bandwidth. The core will be easily saturated at the higher end. If you find a ferrite with Ur that drops off inverse with the frequency it may help.

 

[JBE]

It is trying to be an ideal linear sample, to measure the primery current phase and a multiple of its magnitude. May you please be more specific about your "Ur" suggestion? What about my question about the low permeability at UHF frequencies vs. current measurement accuarcy?
Thanks

 

[GOTWW]

I notice that mini-circuits TC1-IT is an RF transformer/balum 1:1 from 0.4-500MHz. So It would appear what you want to do is feasible.

In general the high permeability material will push up the self inductance, combined with residual capacitance will drive your resonant frequency down. parallel resonance on the primary side will give you an oscillator. any other resonances will prevent your device from operating as designed.

These materials have real and complex permiabilities and permitivites as you have mentioned. I can not remember if the touchtone allows for this in thier XFER model, I know that it was allowed in the PERM statement. This would allow you to model the device somewhat, if in fact you have good characterization data of the prospective materials. I once did ferrite characterization and modeling from 50M to 50G to extract these u,u*, e,e*, and wideband baluns, but it was mainly cut and try.

In summary, I am not much help

 

[JBE]

Would be a single amorphous cobalt based core able to cover the requested bandwidth (500 kHz-520 MHz? What about the nanocrystaline Vitroperm 500F?
Please let me know

 

[GOTWW]

Why not get some free samples of a variety of prospect material and try them. Eventually you will need to test your product anyway.

 

[JBE]

I set up a PCB for testing different materials. I have already tested several powered iron cores and ferrites that we have available.
I have ordered some other materials. I have been trying to get samples from the company manufacturing the convenient for current transformers nanocrystaline Vitroperm material.
On the other hand I do not have of any reference (independently of the french Bergoz Indtrumentation)to get amorphous cobalt cores information or samples.
May somebody provide me with information?

 

[olandoski]

JBE
I believe that permalloy is the correct material for current transformer.
Marcos

 

[JBE]

Although I have not field experience, I do not think that the Permalloy nickel/iron alloy would work in the considered bandwidth (0.5- 520 MHz). Am I wrong?

Any reference about the nanocrystaline and amorphous cobalt based materials?
Thanks

 

[BenoitV]

Concerning coblat based amorphous and nanocrystalline, here are few things to be known :

These two materials are indeed the best for high accuracy current transformers.

For high frequencies, you may need coblat based amorphous (like Vitrovac 6150F) with a permeability around 1500.

You may also use nanocrystalline material like Nanophy (made by the company Mecagis) which propose high permeability (from 20 000 to 200 000) cores for high accurancy current transformers.

These two materials have good characteristics (low losses, high saturation induction) but their use depend on the type of transformer. If you have a DC current thourgh your core, then coblat based amorphous is better. Otherwise, nanocrystalline is the most suitable.

 

[GraviMan]

I have sometimes wondered if transformers in these frequency ranges actually need a core. If you twist the input and output wires together (assuming they are well insulated), you could wind a pretty good torroid. One idea is to use a flexible nylon core, that could then be wrapped around end to end.

Any thoughts?

Mart