Powered iron inductor for DER-472....Value calculation
Hi,
On page 33 of DER-472, the inductance of a kool mu torroid wound with 45 turns, and with a DC Bias of 4.6A, is stated as being 144.9uH.
Using the Torroid datasheet and the Mag-Inc website, the value is 167uH. How does Power Integrations calculate 144.9uH?
The torroid is 77930-A7.
The following Mag-Inc website allows us to clearly calculate that the percentage of initial permeability with the relevant value of H is 51% (of the initial value of ur).
https://www.mag-inc.com/Products/Powder-Cores/Kool-Mu-Cores/Kool-Mu-Material-Curves#kmpermvsdc
From this the inductance is clearly 167uH. How does Power integraions calculate 144.9uH?
回應
Hi treez,
thank you for your interest in PI products.
I am currently looking into the calculation details for the toroid inductance and losses. I will get back to you once I have the answers.
Thank you so much
Hi treez,
we investigated the spreadsheet used in DER-472 and compared it to your numbers.
First of all, when I use the formulas from the Mag-Inc website for 4.6A, I am getting 162uH. Could you please verify that you are getting 167uH in this case?
162 uH is still significantly different from the 144.9 uH in the spreadsheet. You could also see this difference on the second line of page 34, with a MUR_RATIO of 0.46 instead of 0.51. There are two reasons for this discrepancy.
1. The spreadsheet accounts for current ripple in the inductor and computes the inductance at 4.9A current. Using 4.9A gives a ratio of 48.4%.
2. The formulas for H on the manufacturer's website are curve fits based on their internal data. They provide different formulas for other measured quantities. These are not always fully consistent with each other, which you can see when attempting to combine multiple fitted curves together. This is an understandable limitation of equation fitting using limited real data. In addition, there are some issues at the high/low edges of these fits. For these reasons, PI has developed its own curve fits for all of these materials. The curve fits stay close to the manufacturer's values throughout the range, but never perfectly follow them. This accounts for the remaining 2-3% difference in the permeability ratio.
For core losses, I had trouble reproducing your value of 0.69W using the manufacturer's formula. What is your value of B in this calculation?
We discovered that the spreadsheet is using an old, less accurate version of the core loss formula. We will update the HiperTFS-2 spreadsheet to use the more accurate formula from Mag-Inc's website.
Also, on page 34 of DER-472 it states that the core loss for the output inductor of the two transistor forward is 1.28W. However, using the Fit functions on the Mag-Inc website the loss comes out as 0.69W. Please can you show your workings?