Goodchild wrote ...

nzoomed,

When looking for toroids for your GDT core you will want to consider a couple key parameters. Lets start by looking at the universal transformer EMF equation:



This equation is technically for sinwave inputs and outputs, but it gives an okay approximation for GDTs with a square drive.

f - being your frequency of operation
N - number of turns on the core (for the winding of interest)
a- cross sectional area of core
Bpk - the peak saturating flux density your core will support
Erms - being the RMS voltage the core will support at the given frequency

If you rearrange the equation and play with it a bit you will notice a couple of interesting things.

1. For any given transformer configuration (ie fixed - core Bpk, turns, and a) you will get a fixed volt/second coefficient that the transformer will be able to handle. Exceeding the V/S ratio and by proxy Bpk the transformer will saturate.

2. Increasing turns or cross-sectional area with decrease Bpk. Bpk is flux density or flux per unit area.

Now on to the fun part selecting your GDT core. First select a material. T65 if a good choice I also use T65 cores for a lot of the things I do. Look for your self at the datasheet:

http://de.tdk.eu/blob/187202/download/4/pdf-t65.pdf

It has a saturating flux density (Bpk) of 0.46T at 25*C. The initial permeability is a little less critical in this application but something in the range of 3000u to 6000u will work fine.

TDK and other companies have lots of flavors of ferrite and I encourage you to shop around.

Go shop around for a larger core with T65 materiel (Digikey also sells a lot of them). Once you select a core plug it's specs into the above equation and you will be able to see if it will work in your application.

I hope this helps!

nzoomed wrote ...

...
I tried some calculations, but my Erms comes out as 498080587.848

Which is obviously ridiculous!

Is the frequency in the equation supposed to be written in Hz or KHz?

Cross section of the core is 256mm2 and my coil will be running at 68000Hz

Transformer has 14 turns on the primary, so these are the figures i used in my calculation.

hen918 wrote ...


you will need everything in SI base units. i.e. meters squared and Hz

Goodchild wrote ...

hen918 is correct you need to use all SI base units in any kind of electrical engineering.

Those cores you linked are crazy expensive, you can go much smaller and cheaper.

Here is an example, take this core on digikey:

It's T37 ferrite, Bpk of .38T at 25C.

It has a cross sectional area of 51.26mm^2 or 5.226e-5m^2 So plug these numbers into the formula:

4.44 * 68e3Hz * 14 * 5.226e-5m^2 * 0.38T = ~84V


This as you can see is a rather overbuilt configuration. The UD typically puts out 24V so this would work fine. Remember though this case is for 25*C the saturation changes with temperature and it's up to you as the engineer to plan for the worst case your design would see in operation.

Hope this helps.


EDIT for further reading on magnetics check out this page:

Goodchild wrote ...

I only picked that core as an example. You will need to shop around to find what work for you. My goal was only to give you the means to correctly select a core for your use.

nzoomed wrote ...


OK, i understand now where you are coming at.
But the only problem is that all the large cores are rather expensive anyway, so i was just going to go with one of those larger cores.

Anyway, since ive ordered everything i thought i needed from mouser, its going to cost much more to order another single toroid from them, since its going to cost me at least $50 including shipping
Im limited pretty much to what RS can supply me.

Im interested how you say i can go much smaller and cheaper.
I cant really go much smaller if im trying to turn 14 turns of cat 5 on the core.
It seems most people are using them this size for their GDT also.

The weird thing is that Mouser was showing this as discontinued the other day, but when searching by size and dimensions, i find the very toroid listed with the same part number

Its my fault for not checking the dimensions i guess, but i thought the part number for my toroid was a larger one like this.

Lesson learned.