Big core for SMPS Steve Ward syle / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Big core for SMPS Steve Ward syle

1 2 next

Move Thread

LAN_403

Robert2

Fri Jan 07 2011, 12:29PM

Registered Member #1773 Joined: Tue Oct 21 2008, 06:56PM
Location: Poland
Posts: 93

Hello,

i have problem. I watn to buy big ferrite core and i found nice U-93/76/30.

The problem is that I cant find anywhere information on the material from which the core is made becouse since a lot depends on material. I know it might be a dumb question, but maybe someone somewhere has seen these cores and knows what they are material, or where they are used? I know I'd find the corresponding cores on ebay, but shipping to my country comes dearly, perhaps even more expensive than the core.

Greetings from Polish! Robert

Sulaiman

Fri Jan 07 2011, 01:19PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

That core looks to be made of a ferrite grade meant for power throughput.
I don't know the actual grade (it doesn't matter much) but look up 3C90
for a material that is probably similar.
The material is important but the airgap is more important.

Ash Small

Fri Jan 07 2011, 02:51PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Sulaiman wrote ...

That core looks to be made of a ferrite grade meant for power throughput.
I don't know the actual grade (it doesn't matter much) but look up 3C90
for a material that is probably similar.
The material is important but the airgap is more important.

I think 3c90 is pretty similar to N27 (someone please correct me if I'm mistaken.) IE good for up to around 100kHz.

These cores (in the link) look damaged. This will affect their performance. Also, any hairline cracks will act as a gap.

I have some similar sized N27 cores I'm planning to use for a supply. Can anyone recommend any relatively simple method of calculating what the gap should be?

I know the gap is pretty critical, as I once took a job for a while in a coil winding factory to learn about them, I know we used different gaps for different applications, but I never found out the details of the various applications that the different coils were used for. Only that the ones we supplied to Raytheon were a lot more expensive than most of the others. I did end up designing a new vacuum impregnation system for them while I was there, though, using a custom made steel belljar with a perspex window in the top as the production workers kept breaking the glass ones.

Sulaiman

Fri Jan 07 2011, 06:05PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

One of the basic formulas for inductance ... L=U.A.N.N/le
L :Inductance (H)
U :Permeability (H/m)
A : cross-sectional area of flux path (m.m)
N.N :Number of turns, squared (#)
le :effective path length. (m)

Since the ferrites used in this example have a relative permeability 1000's of times higher than air,
le is effectively the length of the air gap(s).

P.S. if the core is meant to have an airgap (energy storage)
if any cracks (or repairs) are narrower than the intanded airgap there should be no problem.
If the cores are to be used without an airgap then cracks/repairs will cause significant flux leakage / leakage inductance.

IamSmooth

Fri Jan 07 2011, 09:09PM

Registered Member #190 Joined: Fri Feb 17 2006, 12:00AM
Location:
Posts: 1567

Sulaiman wrote ...

Since the ferrites used in this example have a relative permeability 1000's of times higher than air,
le is effectively the length of the air gap(s).

If one has a solenoid with N turns, and the length is le, how does le become the length of the air gap? What if the solenoid is solid and has no air gap?

Electra

Fri Jan 07 2011, 10:18PM

Registered Member #816 Joined: Sun Jun 03 2007, 07:29PM
Location:
Posts: 156

I have noticed older ferriteâ€™s have more a coarse matt look, where as new modern ones have a much more of a smooth glossy look. I am still not sure if this is a difference in the actual grade of the ferrite or just down to improvements in the moulding process though.

If they are old stock possibly could be something similar to 3C80, which was common. I think one of the main differences is that the newer materials are less lossly, as modern psu designs tend to work at higher frequencies. But you could find the winding area available may limit the power you can use it for, more than the core itself.

Sulaiman

Fri Jan 07 2011, 10:32PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Iamsmooth, for a solenoid the airgap is the outside of the coil,
think of the magnetic field pattern of a solenoid (or rod magnet)
(try saturating the core of a solenoid, sine or dc...not pulse)

Electra, I think 3C85 was the popular ferrite in Europe (Phillips...NXP)
In USA I think type R,S or 73,77 were common though not so sure.

The MnZn ferrites (<1 MHz, dark grey, high Ur) seem quite 'grainy'
the high frequency, lower permeability, lighter grey NiZn ferrites do seem smoother to the touch, though as you say it may just be modern manufacturing techniques/formulations.
(I recently bought a few sizes of NiZn binocular cores and I find myself fondling them due to the texture ...)
I have some 3C85 1"x 1" x 4" blocks that are machined on all faces except one - which has the texture of cornflakes.

Ash Small

Fri Jan 07 2011, 11:41PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

"One of the basic formulas for inductance ... L=U.A.N.N/le
L :Inductance (H)
U :Permeability (H/m)
A : cross-sectional area of flux path (m.m)
N.N :Number of turns, squared (#)
le :effective path length. (m)"

I understand the permeability can be measured by winding one complete turn around the core, but would this be done with two 'C' cores and no gap?

Also is N number of turns on the primary or secondary, or both?

Sulaiman

Sat Jan 08 2011, 08:53AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Just did a quick bit of Googlinf,
the cores are probably Epcos (Siemens) N27 or N87 material,
Ur around 1800-2000 ...
Design for 100mT (1,000 gauss) flux density for about 10W total core loss at 25 kHz, that's about 4.5 volts-per-turn.

Hurry up ... 5min. left

Robert2

Sat Jan 08 2011, 09:40AM

Registered Member #1773 Joined: Tue Oct 21 2008, 06:56PM
Location: Poland
Posts: 93

Sulaiman wrote ...

Just did a quick bit of Googlinf,
the cores are probably Epcos (Siemens) N27 or N87 material,
Ur around 1800-2000 ...
Design for 100mT (1,000 gauss) flux density for about 10W total core loss at 25 kHz, that's about 4.5 volts-per-turn.

Hurry up ... 5min. left

You're right. This is a B67345-B1-X27 . A bit of me it does not fit, not enough space in the "window" on the high voltage winding. To tell the truth for now I have to play with the X-ray transformer 3c90, B1:

The primary winding of copper tape 2*5 turns. Thank you for your replies in this topic. Now, hold on to the purchase of the core, may find a better, bigger. I would like to build a

Regards warmly :)

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob