Some input fields and some calculation results that need comments have tooltips. The contents of these tooltips repeats the aid stations. 1. The losses in the core depend on the combination of induction and frequency. At high frequency and high induction losses will be so large that the heating of the core will exceed reasonable limits. Therefore, the program itself limits the maximum induction depending on the entered frequency value. 2. For PWM controllers set the frequency equal to half the frequency of the master oscillator chip. The pulses of the master oscillator are fed to the outputs in turn, so the frequency at each output (and transformer) is 2 times lower than the frequency of the master oscillator. IR2153 chips, and similar to this family of chips, are not PWM controllers, and the frequency at their outputs is equal to the frequency of the master oscillator. Do not chase the high frequency. At high frequency switching losses in transistors and diodes increase. Also, at a high frequency due to the small number of turns, the magnetization current is too large, which leads to a large idling current and, accordingly, a low efficiency. 3. The fill factor of the window describes what part of the Windows core will take all the windings. For rings can be set more to 0.4, as they have a large window area. In W-shaped cores window area is limited by the core and frame. For them, the window fill factor can be in the range of 0.15-0.35, depending on the size of the core, smaller values correspond to smaller sizes. 4. The current density depends on the cooling conditions and the size of the core. At natural cooling it is necessary to choose 4 - 6 A/mm2. When ventilation current density can be selected more, up to 8-10 A / mm2. Large current densities correspond to small cores. When forced cooling, the permissible current density depends on the cooling intensity. 5. If output voltage stabilization is selected, the first output is the master. And it is necessary to assign the output with the highest consumption. The rest of the outputs are considered to be the first. For real stabilization of all outputs it is necessary to apply the throttle of group stabilization. 6. With unipolar rectification, despite the greater consumption of copper, the midpoint rectification circuit has the advantage, since the losses on the two diodes will be 2 times less than on the four diodes in the bridge circuit. 7. Setting the minimum current consumption is necessary to calculate the inductance of the throttle. Too small a value should not be set, because it can get a large value of inductance, and the throttle to a large inductance and a small minimum current will be difficult to make, because you still need to ensure that the core of the throttle is not included in the saturation. 8. For the correct operation of the throttle in the rectifier after the diodes before the throttle should not be any capacitors! Even a small denomination. 9. When you add a core to the database, the name of the new core is entered in the combo box. There is no need to assign to the name of the core the prefix of its shape and the name of the core material. The prefix of the core shape and the name of the material are added automatically by the program. 10. When adding a new core to the database in any case, it is better to use the "Add to database (data entry)"checkbox. The data should be taken from the manufacturer's documentation. Decoding of data designations: Ae-core area, mm2 An - the area of the bobbin window under winding, mm2 le - average length of the magnetic line, mm Ve is the volume in cm3 In case of absence in the documentation of the manufacturer of the area of the window bobbin under winding, calculate the area of the window yourself. In the absence of information from the manufacturer of the volume of the core, the volume is easily considered independently. The volume is equal to the product of the area of the core section by the length of the average magnetic line. Although I did the calculation of the core data on the geometric dimensions, the results of this calculation are not accurate, since the data from the manufacturers are somewhat different from the direct calculation of the geometric dimensions. 11. When choosing the shape of the core "Other" data taken from the manufacturer's documentation.