The physical dimensions table is important since we must know how big the toroid is to allow sufficient space in our project. The cross sectional area and mean length (magnetic path) are included so that flux density and magnetising force calculations can be made. Iron powder toroids use a type number which begins with "T" followed by a figure which denotes the outside diameter in hundredths of an inch and a second figure denoting the type of material.

Another table lists the A_{L} values
for all core sizes and material
types, and these values remain substantially constant for each iron powder
material up to VHF frequencies. The different materials have different values
of permeability (µ), which also determines the useful range of frequencies
for each type of material. In addition, iron powder cores are colour coded for
easy recognition. In general, the smaller cores are used at the high end of the
specified useful frequency range and the larger ones at the low end, so that a
better Q factor is obtained. Typical unloaded Q factors of 100 - 200 are
achieved.

A table lists the maximum turns vs wiregauge for each core size using single layer windings. This saves a lot of cut and try so that we can make a good first approximation for our winding.

Suppose we wish to wind a toroid coil for use in a receiver
front-end tuned to 7.0MHz. Since there is no rf power involved we
do not have to worry about saturation of the core and can choose
almost any size core. Without going overboard, it is best to choose
on the larger side so that we can achieve a better Q. If we want to
tune this coil with a 100pF capacitor we will require an inductance
of approximately 5uH. With a frequency of 7MHz we have a choice of
2 or 6 material, but let us decide on a T-50-2 as our core. The
A_{L} value of this core is given as 4.9 and so we can
calculate the required turns to produce 5uH by using the
formula
listed earlier. This works out to 31.9 which we will round up to 32
since we cannot have fractional turns on a toroid. (A wire passing
straight through the core is defined as 1 turn). Next we look at
the table showing turns vs core size for different wire gauges and
see that we can fit 39 turns of 24g enamelled wire on a T-50-2
core. For any toroid winding it is not good practice to fill the
circumference with the winding since the distributed winding
capacitance causes a "self resonance" in the winding which lowers
the effective Q. A correctly designed winding should cover about
300 degrees of circumference but not more then 330. Slight
adjustment to the inductance can then be made by compressing or
spreading the turns to either increase or decrease the winding
inductance. Our toroidal inductor should have an unloaded Q of
about 200, which is quite a respectable figure for this
application. The finished winding can be secured with some
polystyrene cement (Q-dope). If a coupling link is required on the
toroid, it should be put on at the cold (earthy) end of the
winding, and in this case, could consist of 2 or 3 turns. It is
important to know how to measure the resonance of a tuned circuit
containing a toroidal inductor with a "dipper". With the lack of an
appreciable external field we are unable to inductively couple the
dipper coil to the toroidal coil directly, so we must use a link to
couple between the toroid and dipper coil. Use the minimum number
of link turns (at the cold end) which gives a noticeable dip.

Three data tables list the dimensions,
A_{L} values and
wire table for ferrite toroids as for iron
powder toroids.

Unlike iron powder material, the permeability of ferites changes with
frequency and therefore the listed A_{L} values for ferrite cannot
be used directly for calculating inductance. Excel
programs incorporating all the necessary parameters are provided for
calculations.

One important table lists the ferrite material type
and shows the useful range of frequencies for resonant circuits (narrow band)
and wideband circuits. This table indicates the low loss frequency limits for
each type of ferrite, and you can see that for some frequencies there is more
than one type available. Notice that the lower permeability materials
operate with low loss at the higher frequencies. This initial permeability
(µ_{i}) is the permeability measured at about 10kHz.

Suppose that we want to design a wideband transformer for use
between 3.5MHz and 30MHz and it has to match a 200 ohm source to a
50 ohm load. Since this is an RF application we must use the program
FT_calc_1.xls, located under Program on the homepage.
A common *Rule of Thumb* states that the winding should have an inductive
reactance of not less than (4 x load) i.e.800 ohms at 3.5MHz. Calculating the
required inductance to get 800 ohms at 3.5MHz gives a result of 37uH. Similarly
the secondary winding would need an inductive reactance of 200 ohms which works
out to be 9.2uH. Now let's work out the turns required for a ferrite core
of suitable material for the frequency range. Assuming this a low
power application (e.g. receiver or very low power interstage
transformer in a transmitter) we will use an FT-50-43, which is listed as
suitable for 1 - 50MHz.

Opening program FT_calc_1.xls we first have to place the physical data for the FT-50-43 under the Manufacturers Data in the Program sheet. This data is listed in the MyData or Data worksheet and can be copied into the Program sheet. The #43 ferrite data is located in the Data sheet and can be copied and pasted if necessary into the Program sheet. We now have all the manufacturers data inserted into the Program sheet, and need to insert our particular User Data. This application is for a low level transformer so we will enter 1 for Power, 200 for Load impedance and 1 for Stack. The result shows Winding voltage = 14.1. For a temperature rise of 30C we see Pmax is 0.7W.

We now come to Winding Calculations which will provide a lot of data. In Table 3.1 we can enter values for Pcore and Turns for each listed frequency. On the line for 4MHz we will enter 0.7 under Pcore and intially 10 under Turns. This produces an XL value of 621, which is less than our required 800 ohms. By increasing Turns to 12 the XL value increases to 895. The value at 3.5MHz will still be near enough to 800 ohms. It is interesting to see that at 30MHz XL is 1342 -- well above 800 ohms, and the values for L are close to our initial calulations. Since this transformer has a 2:1 turns ratio the windings are made up of 12 turns and 6 turns and would best be made as a trifilar winding of 3 x 6 turns.

If we were designing a current (Guanella) balun for use with a coax feeder, the design should also be checked with program Balun_cmr_1.2*.xls referenced in Program on the homepage.

28/12/13homepage