Hello,
I am using the saturable core in PLECS.
What exactly is meant with µ_r,unsat and µ_r,sat?
I am using Ferrxocube cores, in the datasheet I can find values for the AL value of the specific core or the µ_i initial permeability of the core material. How can I calculate or find the values needed for the saturable core in PLECS? Or can I use the µ_i as µ_r,unsat?
Thank you.
The Saturable Core component documentation shows the how the µ_unsat and µ_sat relates to the BH curve. Ferroxcube application notes define a clear difference between initial permeability μ_i (pg. 7 eq. 4) and effective permeability μ_e (pg. 7 eq. 5). One is a purely material property while the other takes into account gapping. The inductance factor (AL) is defined in the same document on page 9 / eq 27 and is a function of μ_e, which takes into account gapping.
So in PLECS, you can have a Saturable Core component that models the material based on μ_i, with the gap seperately modeled, or of the core based on μ_e which includes the gapping effects in the Saturable Core component. You should be able to define a consistent relationship between all three variables, although if they are inconsistent it could be due to fringing effects for larger air gaps that Ferroxcube might include in AL/μ_e.
.Hello Bryan,
thank you for your answer. This clarifies it.
I looked at the “flyback_converter_with_magnetics” demo model. The model description says, a Ferroxcube 2616 pot-core made of 3C91 material is used. The core dimensional parameters in PLECS are exactly like in the core datasheet, but µ_r,unsat = 9550.
When I look in the Ferroxcube datasheet for the 3C91 material, I can’t find this value, only µ_i = 3000.
Where does this value 9550 come from?
There is also a leakage flux path included. Can I get this value also from the datasheet? Or do I have to take real measurements to include it in the simulation?
Thank you
Where does this value 9550 come from?
The derivation of this value is not part of the demo model documentation. However, the 3000 value of µ_i from the 3C91 datasheet is actually at 0.25 mT and not at 0 mT as the Ferroxcube material I linked would have suggested. Looking at the 3C91 data close to 0 mT, as would be the relevant conditions in the PLECS model, the slopes are noticeably steeper and the value was closer to 10k. I would say that explains the discrepancy.
Can I get this value also from the datasheet? Or do I have to take real measurements to include it in the simulation?
Leakage inductance is effected by winding techniques (i.e. interleaving) and so it isn’t something you can find on a datasheet. Measurements are ideal, although in the absence of a prototype I’ve come across a rule of thumb, that the leakage is typically 3-5% of the magnetizing inductance for a flyback transformer.
What data figure in the datasheet do you mean here?
Fig.3 Typical B-H loops in the 3C91 material sheet.
Are there any additional information/documents about this topic? I am not really sure what value for µr_unsat I should use for my specific core. In the other demo models where the saturable core is used, the µr_unsat is also different to datasheet values and also not documented in the model description.
Sorry, that was a typo in my original post - it’s 25 mT as shown in the datasheet.
Fundamentally you’re trying to fit a square peg (hysteretic BH curve) into a round hole (Saturable Core component). To use the Saturable Core you need to make some engineering assumptions and pick a point on the curve or average the slopes of the BH curve.
The Hysteretic core might be a better fit. The component documentation shows how the BH curve relates to the various parameters.
Thank you for the explanation, I will take a look at the hysteretic core.
Actually, the 0.25mT in your previous answer were right, the ferroxcube datasheet gives µi for 0.25mT.
Hi,
I really do not understand from where , the value of the unsaturated releative permeability is used, and also the saturated releative permeability is 1 , is it ok the unsaturated releative permeabiltiy is 9950 and the saturated releative permeability is 1
Thanks
The 9550 value is an approximation.
Looking at the BH curves in the material datasheet, you will need to determine the slope of the curve around H=0. Since neither temperature curve actually goes through the origin with these being the hysteresis curves, you are basically needing to eyeball the midpoint and do your own rise-over-run determination.
As an example, if you draw a straight line from B=0 to B=100mT, this would bring us close to H=8A/m. Then, to determine mu_r,unsat, you need to divide this slope by mu_0:
(0.1 T / 8 A/m) / (4pi10^-7) = 9947, which was then rounded up to 9550.
Thanks for your reply, now it is very clear,
I would like to verify this thing, you have taken into account the BH curve at 100 degree Celsius.
Regards,
That is correct. So you can feel free to adjust for temperature in your application as needed. Using the 25 C curves doesn’t result in a big difference in H using the slope around H=0 (~5 A/m), but I understand that leads to a pretty significantly different value of mu_r,unsat.
Thank you indeed, really thank you
