What's the difference between an E-I core and toroid core?

[nicleo]

Re: Ferrite cores help

1. 4+4 pri, 16T+16T
2. 12V in, out should be 12V x 4 = 48V
3. 20khZ
4. the magnet wire did not get hot
5. 1 pc 1.25" toriod only

i will try to increase the turns ratio.. maybe 6T+6T pri and 24T + 24T on secondary...

one more thing, are E-I cores and toriods frequency dependent? i mean does driving a toriod @ 20khz, make it more "heating up" ?? or should i increase frequency ??

Yup ... there will be 'eddy current' loss, which is frequency dependent (higher frequency, higher loss). However, ferrite power core is designed for high frequency application, so the loss at high frequency will be lower compared to soft-iron laminated core.

Based on your explanation, the driving mosfet is HOT when driving the toroidal core. At mosfet site, power loss (and therefore heat) is from switching loss and conduction loss. What is the duty cycle of your 20kHz switching signal?

 

[rx5]

Re: Ferrite cores help

duty cycle?? i dont know..all i know is im using SG3525 at 5% deadtime... does this mean 95% duty cycle??

someone told me iron powder cant be used on high frequency cause this will just heat? up..
--------------------------
Yup ... there will be 'eddy current' loss, which is frequency dependent (higher frequency, higher loss). However, ferrite power core is designed for high frequency application, so the loss at high frequency will be lower compared to soft-iron laminated core.
--------------------------

yup, your righht, higher frequency means lhigher loss of output voltage... i used variable resistor for Rtiming of SG3525, an i get a "variable" output SMPS.. but when you set the frequency too low, say near 20khz, youl hear an audible whine at the core...

 

[nicleo]

Re: Ferrite cores help

duty cycle?? i dont know..all i know is im using SG3525 at 5% deadtime... does this mean 95% duty cycle??

someone told me iron powder cant be used on high frequency cause this will just heat? up..
--------------------------
Yup ... there will be 'eddy current' loss, which is frequency dependent (higher frequency, higher loss). However, ferrite power core is designed for high frequency application, so the loss at high frequency will be lower compared to soft-iron laminated core.
--------------------------

yup, your righht, higher frequency means lhigher loss of output voltage... i used variable resistor for Rtiming of SG3525, an i get a "variable" output SMPS.. but when you set the frequency too low, say near 20khz, youl hear an audible whine at the core...

Deadtime and duty-cycle are two different things. Deadtime is required to prevent 'shoot-through' problem between upper and lower switches at the same leg. Duty-cycle can be varied from 0~100% and it tells how long the switch will be turned on in one cycle or period. When the duty-cycle is higher, it means the the switch will be turned longer and therefore more power will be delivered to the wound toroidal core (in your case).

In my last post, the eddy current loss of iron powder (ferrite) core at high frequency is lower compared to that of laminated iron core (like those iron sheet based cores used in (50Hz/60Hz) power transformer or inductor design). Therefore, iron powder core is preferred in switching mode power supply design (SMPS). As the eddy current loss is proportional to f², the heat will become a problem again if the frequency is too high. However, iron powder core is common in 20kHz, 40kHz, 100kHz, 200kHz, etc. SMPS design.

Usually designer will set the frequency higher than 20kHz. For your information, the audible range of human ear is 20Hz ~ 20kHz.

 

[rx5]

Ferrite cores help

i think i just better use an E-I core... to keep the mosfets from heating .... besides, its very hard to get toroids in my area.

 

[strabush]

Re: Ferrite cores help

All your discussion is futile. The first step to understand the problem is to understand the functioning of the SMPS. Someone asked what is the toplogy, and the answer was just SMPS, well this is no answer. There are topologies like Flyback that the functioning of the power supply depends on BOTH the inductance and the saturation characteristics of the transformer, so if you do not address both matters you just are shooting in the dark, without understanding what is really happening. The shape of the core itself is not the matter, but the electrical characteristics very much matters. So more exact description of the circuit is needed.

 

[TECHLABS]

Re: Ferrite cores help

you can download the software for designing them for free. here is link to site. use the free software button for complete list. www.tech-systems-labs.com