Continue to Site

# Calculating peak B in PFC inductor (Kool Mu)

Status
Not open for further replies.

#### cupoftea

Hi,
By stepping H and continuously finding the delta B from H to “next H”, I have calculated the peak B of a PFC inductor as 0.25T. (as in excel attached)
Its 50 turns on a 77072A7 torroid from Mag-Inc.
I can’t find on the website where I can check this value….is it on this page?...

https://www.mag-inc.com/Products/Powder-Cores/Kool-Mu-Cores/Kool-Mu-Material-Curves

Peak current is 12.3A and switching frequency is 60kHz (it’s a 600W PFC with 100VAC input)

77072A7 torroid

#### Attachments

• PFC inductor_forum.zip
17.5 KB · Views: 79

Peak B is proportional to peak I ( same for min )

you need to calc the flux ripple ( 60kHz ) in order to calc the losses in the core - the 100Hz flux ripple losses can be largely neglected.
--- Updated ---

for your core you can have 50 - 70mT ( on average ) each way @ 60kHz before the losses rise too greatly - per the date sheets you reference

Last edited:
for your core you can have 50 - 70mT ( on average ) each way @ 60kHz before the losses rise too greatly - per the date sheets you reference
Thanks, the attached (please ignore volts on scale) is the B waveform for the core.
I reckon it is on average <60-70mT
--- Updated ---

I must admit i am baffled as to why Mag-Inc dont subscribe to the method of calculating peak B as shown in the excel doc of the top post here...(ie, stepping H in small steps, calculating delta B from H to "next H", and then summing all the delta B's along the way....i mean, what could be clearer than that?....its the obvious, intuitive way to do this........looking through their app notes looks nowhere near as clear as this...moan over
--- Updated ---

Peak B is proportional to peak I ( same for min )
Thanks, thats actually a great reminder, that even though this is non linear magnetic material, B is still proportional to current, no matter what the current is.

#### Attachments

• B waveform.jpg
102.7 KB · Views: 68
Last edited:

I must admit i am baffled as to why Mag-Inc dont subscribe to the method of calculating peak B as shown in the excel doc of the top post here...(ie, stepping H in small steps, calculating delta B from H to "next H", and then summing all the delta B's along the way....i mean, what could be clearer than that?....its the obvious, intuitive way to do this........looking through their app notes looks nowhere near as clear as this...moan over
I believe there was discussion about this point some time ago. Your longwinded method can be expected to give correct results, but most manufacturers, also magnetics, have curves and approximation functions that alow to calculate B versus H respectively I directly.

Thanks, for the core in the top post, i have now calculated core loss of 1.14W using their equations.
I took the average dB as 0.06mT.
Then i believe this has to be divided by two for the equations given in the Mag-Inc docs?
Also, the result of their equations is in mW/cm^3 i presume?

...I am guessing this as the Mag-Inc docs dont make it clear.

Thanks, ive worked out the problem now with kool mu torroids.
Its necessary to do homework before diving into it...and produce the attached type of "kool mu torroid" template.
I think its a waste of time starting to do any calcs until that has been done.

#### Attachments

• Mag inc coeffs _koolmu.zip
132.6 KB · Views: 86

I have also contacted Mag-Inc yesterday to ask how to buy their kool mu torroids in qtys of 1 or 2....and if this is even possible.......i hope they get back. The way component shortages are gong now, it woudlnt be surprising if its just not possible to buy the kool mu torroids at all?

Hi,
Pg 20 of this..

states that H (in oersteds) = [N.i/(path length)] * (4.pi)

.....but surely its [N.i/(path length)] * (1/79.58) ???

Hi,
I calculated an inductor of 250uH at 12.3A, by using two pieces of Mag-Inc 77620A7 with 21 turns.
(For a PFC inductor)
At mains peak, Imax=12.3A and Imin=6.2A.
Also calculated total core loss = 4.2W and total winding loss = 1.8W for two parallel strands of TEX-ELZ 7/0.3mm.

(As in the orange tab of the attached)
Do these ring true?

#### Attachments

• Mag inc coeffs _koolmu_1.zip
240.3 KB · Views: 64

Status
Not open for further replies.