TDK Corporation has announced the intention to discontinue the PC50 power material. The factory is replacing the PC50 material with PC95 material on all quotations. Therefore it is imperative that we understand the differences between the PC50 and PC95.
PC50 History – The PC50 material was announced in 1988 after years of developing new processes for powder preparation and Sintering. The PC50 material was the first ferrite material design for the operating frequency of 500 KHz – 1MHz. A review of the power material can be found in the paper entitled Ferrite power material for high-frequency applications by Tadashi Mitsui and Gary Van Schaick. This document can be found at:
A couple of key points can be found in this paper.
- Relationship of Bs and Br – As temperature increases, Bs and Br of a power ferrite material changes. It is desirable to have the difference between Bs and Br try and remain constant, i.e. ΔB is constant. In some power materials, (TDK and our competitor), Br decreases and then starts to increase
- Core Loss – Comprises three components, Hysteresis, Eddy Current and residual. As operating frequency increases from 100 KHz through 250 KHz to 500 KHz and beyond, the Eddy Current losses become the dominant loss of the total loss. Therefore if a core material is to work at 500 KHz and above the material must control the Eddy Current losses, unfortunately this could have an effect of increasing the Hysteresis losses. The goal in developing a material for high frequency is to balance the Eddy Current losses and the Hysteresis losses
Let’s look at the specification of the two materials.
|Core loss volume density||Pcv kW/m3||100KHz 200mT||500KHz 50mT|
|Saturation magnetic flux density at 1194A/m||βs(mT)||25°C||530||470|
|Remanent flux density||βr (mT)||25°C||85||140|
|Curie temperature||Tc (°C)||min.||215||240|
|Electrical Resistivity||ρv (Ω·m)||6.0||30|
So let’s look at the graphs that show this at high frequencies. Figure 1 shows the core loss for varies materials at low flux levels, 50mT. At these lower flux levels, the PC50 has the lower cores losses across the temperature spectrum with the PC95 being the next best material. Now look at the 100mT graph, figure 2. PC50 and PC95 demonstrate the same core losses until the actual operating temperature reaches a normal operating temperature of 80°C – 100°C.
Flux Density Considerations – From the paper referenced above, it is desirable to have Bs remain as high as possible over temperature. Bm, the maximum flux of the design has to be below Bs at the worst case. In a forward converter topology, the flux swings from Br – Bm where Bm<Bs. The goal is to have ΔB be as high as possible to give the design maximum power through put. The ideal power material maintains the same ΔB from room temperature up to the operating temperature. In reality there is a decreasing ΔB so the practical goal is to keep the ΔB from changing too much. In comparing the PC95 and PC50, there are a couple of advantages.
Forward Converters: The PC95 has the distinct advantage of a higher ΔB (Bs – Br) across the entire temperature range.
Push-Pull/Full Bridge: The PC95 has the advantage of higher overall Bs allowing for a larger flux swing from -Bm – +Bm
Inductance Consideration: This may be the larger concern to some designers. The permeability of the PC95 is over twice the permeability of the PC50. If the cores are going to be gapped, there is no problem as the gap length becomes the dominate characteristic of determining the inductance of the device. If there is no air gap, the cores cannot be a drop in replacement. However this may be an advantage in that the numbers of turns would have to be reduced and this could result in lower cost, perhaps lower copper losses as well.
Overall the switch from PC50 to PC95 has distinct advantages. For initial design in, PC95 is the best way to go.