Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Selecting the core for PUSH PULL Converter

Status
Not open for further replies.

samhtcandroid

Newbie level 5
Joined
May 16, 2012
Messages
8
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,344
Hi everyone,
First of all, i would like to say"Do not thrash me..".
I have some basic things to ask.

I am tryingto design a PUSH PULL converter.
I have gotten a few papers on studying how it works.
However look at the schematic i just cant visaulise how it works.
Q1) is there any idiot proof video for beginners like me to understand?
Q2) I am trying to select a inductor core for building the transformer but not knowing how the cct works i cant decide on one. Can anyone giv me some guidiance?

thank you
 

Dear samhtcandroid
Hi
There are some tables and application notes about how to select core for or desired power and frequency , did you search for them ever?
Best Wishes
Goldsmith
 

I have a small problem. I cannot really visualise how physically to even wind the core.

Reference below: Pix1

**broken link removed**

Am i right to physically wind the core as shown in the pix2 below?
**broken link removed**


Dear samhtcandroid
Hi
There are some tables and application notes about how to select core for or desired power and frequency , did you search for them ever?
Best Wishes
Goldsmith
 

Hi again
Did you think that you can simply add two switch and then build a push pull converter ? of course not . you'll need a dead time creator and a PWM modulator and a compensation network and ......... in the other hand , you'll need a driver IC , and for push pull , SG3526N can be good choice .
Anywhere : the winding seems good . ( don't forget that points that are symbols of start of the winding , are pretty important , and in your picture according to the winding , those are correct )
Best Luck
Goldsmith

---------- Post added at 16:37 ---------- Previous post was at 16:36 ----------

BTW : for secondary , the start point isn't correct
 

hi Sir,
Thanks alot, I do have some light on the part of the Mosfet, PWM and gate driver.

"BTW : for secondary , the start point isn't correct"
May i know which is the start point? Do you mean that the centre is tapped wrong at
the centre?

Do i physically get a block core that look something like the one i've drawn?
Any specific CORE that you would recommend?(from FARNELL/Element14/ RS?)

Am using:
1) 2 N-channel Mosfet for switching
2) Inductor core for Isolation.(the concept of winding)
3) PWM controller.(still searching for suitable ones)
4) Gate driver & etc....

best regards,
Samuel Foo

Hi again
Did you think that you can simply add two switch and then build a push pull converter ? of course not . you'll need a dead time creator and a PWM modulator and a compensation network and ......... in the other hand , you'll need a driver IC , and for push pull , SG3526N can be good choice .
Anywhere : the winding seems good . ( don't forget that points that are symbols of start of the winding , are pretty important , and in your picture according to the winding , those are correct )
Best Luck
Goldsmith

---------- Post added at 16:37 ---------- Previous post was at 16:36 ----------

BTW : for secondary , the start point isn't correct
 
Last edited:

Regarding pix2 in post #3, if you compare with the polarity indicators in pix1, you'll notice that the secondary winding is drawn incorrectly. Simply wind the center tapped winding in same direction from start to end, for both primary and secondary.
 

Hi Samuel
May i know which is the start point? Do you mean that the centre is tapped wrong at
the centre?
See the FvM's post , he correctly said , what you should do to improve it .
Any specific CORE that you would recommend
for how much power and what frequency ?
2 N-channel Mosfet for switching
it depends on th frequency of operation and current and voltage .
3) PWM controller.(still searching for suitable ones)
as i told you at past , try to use a simple SG3526N . ( it is pretty good ) .
If you use this IC , because it has a totem pole at it's out put , you won't need any thing else for your mosfet driving process . and for feed back , you can use a simple opto coupler and zener diode .
Good luck
Goldsmith
 

1)
The spec for my circuit is
Pout=400W
Vout=350V
Is a simple opto coupler and zener diode suitable?

2) I am using 2 N channel mosfet for switching, that means i need 2 driver gate?

3) Reference below. For Feritecore...
i) is all the Primary and Secondary coil wounded on the centre of the core.
ii)Does the pins has any special purpose?
**broken link removed**


best regards,
Samuel Foo
 
Last edited:

Hi Samuel
Is a simple opto coupler and zener diode suitable?
of course it is suitable . you can use a voltage divider ( consists two resistor ) and then use a TL431 ( variable zener diode ) when the reference voltage be about 2.5 volts , it will conduct . thus you can use a simple opto coupler in series with this zener . simply . vout = 2.5(1+(R1/R2)) .
2) I am using 2 N channel mosfet for switching, that means i need 2 driver gate?
Of course not ! the SG3526N , has two out put . it designed for this aim .
And about your core , it seems good , but don't forget that this core should be able to handle this power .
By the way : did you decided about switching frequency ?
And an advice : if it is your first time , with SMPS design , i suggest you , to start with lower voltages , and then decide for 350 volts . ( because this voltage is DC , it will be more dangerous than AC , so , be careful . )
Good luck
Goldsmith
 

Hi Samuel
of course it is suitable . you can use a voltage divider ( consists two resistor ) and then use a TL431 ( variable zener diode ) when the reference voltage be about 2.5 volts , it will conduct . thus you can use a simple opto coupler in series with this zener . simply . vout = 2.5(1+(R1/R2)) .

Of course not ! the SG3526N , has two out put . it designed for this aim .
And about your core , it seems good , but don't forget that this core should be able to handle this power .
By the way : did you decided about switching frequency ?
And an advice : if it is your first time , with SMPS design , i suggest you , to start with lower voltages , and then decide for 350 volts . ( because this voltage is DC , it will be more dangerous than AC , so , be careful . )
Good luck
Goldsmith

I) I am using 20Khz for my Mosfet.
2) Am using this Mosfet(a little costly, S$10 each):VISHAY FORMERLY I.R
. - IRFP460APBF - MOSFET, N, 500V, 20A, TO-247AC
I beleive SG3526N should be able to drive it.

3) for the SG3526N , the output current is 100mA. I read from some papers, it says the output current
is used to drive to MOSFET, therfefore when PWM controller cant produce high enough current, we need a gate
driver?

4) but i come across some articles that says MOSFET is Volt driver, but why the concern over the riving current?
 

20 KHZ seems good , but if you increase the frequency of your PWM , it will be better . because your magnetic elements will be smaller and cheaper . for example try 100 KHZ .
And about your mosfets , with a simple approximation , at push pull arrangement we can say : the maximum voltage that these mosfets should be able to tolerate , should be upper than 2vin . so if your input is 311 volts , it should be able to work at voltages more than 600 volts ( e.g 700 volts is enough ) so the mosfets such as 2SK2611 .
Here is it's datasheet : **broken link removed**
And about SG3526N , yes , it will be able to drive them as well .
4) but i come across some articles that says MOSFET is Volt driver, but why the concern over the riving current?
I know , exactly , what you mean ! but don't forget that each PN junction has a capacitor . for mosfet we have some capacitors , but most important one is GS capacitor . at transient time it is short circuit because it is discharge , and you should give it high value of inrush current . after the time that your capacitor is full of charge , you have a voltage across the GS junction , so your mosfet can be turn on . hence the drive current is most important parameter in each mosfet . don't forget that when you increase the frequency , naturally , you should increase inrush current , or select a mosfet with lower CISS .
Best Wishes
Goldsmith
 

Thank you very for the advice.

Reference to the
**broken link removed**

The "X" axis represents Magnetic Field Intensity (Coercive Force) which is proportional to the Ampere·Turns.
The "Y" axis represents Flux Density which is proportional to the Core Area and the Volt·Seconds applied to the primary winding when active.


I cant seem to locate this BH curves from CORE specification.

**broken link removed**
 

Can anyone share a light?

Thank you very for the advice.

Reference to the
**broken link removed**

The "X" axis represents Magnetic Field Intensity (Coercive Force) which is proportional to the Ampere·Turns.
The "Y" axis represents Flux Density which is proportional to the Core Area and the Volt·Seconds applied to the primary winding when active.


I cant seem to locate this BH curves from CORE specification.

**broken link removed**
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top