36v to 400 push pull boost converter

[krish2487]

hello all..

before i went ahead with building the boost converter i just wanted to confirm that i am on the right path and that my calculations are correct so far.

Heres the idea
i wanted to build a 36v to 400 v boost converter capable of delivering 100w
the supply voltage is 36 v
here is the schematic i arrived at.

The transformer has primary turns of 6 (swg 21) and secondary turns of 84 (swg 35)
It is a EE28 transformer

i wanted some light on whether the method of isolated feedback i used here is correct or not and whether the values of the components i arrived at for the specfic case are correct.

i have not yet given the values of the gate resistors as i first wanted to confirm that my PWM modulator ckt is acceptable.

attached are the eagle files and the bmp of the same

any help is highly appreciated

 

[eem2am]

you should put a safety sense resistor in the sources of the fets.

because also you with push pull have problems with flux walking.

also, how do you deal with the magnetising current....it must be given a path to discharge........also the leakage inductance

 

[FvM]

The feedback circuit can't work, because the shunt regulator's maximum voltage rating is exceeded. An additional zener diode is needed.

 

[krish2487]

@eem2am
that is something i havent thought of until u mentioned it.
i ll have to modify the ckt and get back on that.

@fvm
yes it is.

i did modify the ckt post posting it online
i did not however upload the updated ckt pics online

 

[espirit]

The topology is not clear to me. It looks like a Push-Pull (forward) buck converter. But then the rectification is better if the sec. of the xform has 2 windings, each with 1 diode to the inductor. That saves 2 diodes, and 2x voltage (and power) drop.
The magnetizing current reset can be solved by using a current mode controller.

 

[E-design]

MUR810 is a 100V diode. The 9mH choke seems high compared to similar power units I designed that operated well above the audible range. I would expect values much lower. What is the rest of you design specs, ripple, Vin min, core material type etc?

Added after 34 minutes:

But then the rectification is better if the sec. of the xform has 2 windings, each with 1 diode to the inductor. That saves 2 diodes, and 2x voltage (and power) drop.

2 windings also mean 2X the winding area and may be a bigger core to accomodate this.

You should use the SD input (pulse by pulse current limit) on the 3525A to protect your devices.

On further inspection it looks like you are switching around 40-50kHz which explain the 9mH choke. If you can keep min output current high the inductor value can be much smaller.
Some quick calculations on core data of the net I get:

Max thru power with good tranformer construction @ 50kHz (127W)
With 30V Vin min, 8T per 1/2 primary, at least 1643 Circular mils wire

114T secondary, 79 min circular mils

 

[Tahmid]

Hi,
A few suggestions:
1. Use snubbers across the transformer primaries for transient suppression.
2. Increase the value of the output capacitor for lower output ripple. In a 12v to 290v push pull converter (300W), I used 150uF 400v capacitor.
3. Connect pin 10(SD) to ground.
4. You left pin 15(VCC) open. It should be connected to V+.
5. Provide a shunt and short-circuit/overcurrent protection using an op-amp/comparator.
6. Apply gate to source resistors and zeners for gate protection.
7. You can use zener-opto for feedback or high value resistor divider.
8. Have an input inductor (to prevent input current inrush) and bulk capacitance (required in SMPS) in the range of 2000-4000uF.
9. Use a diode of higher rating (400V*1.6) = 640V rated minimum. So you can use MUR480 or MUR4100.

 

[deepak.k]

Hi All

I am also doing the same thing. In my case battery voltage is 12V and Fsw is 50 Khz.
My question is what will happen if kept the secondary winding open? in this condition i could not get the proper wave form of primary winding, there is to much spikes on the primary side also there is no effect of snubber in this contition.
Spikes is more then 100V is present.
Is the core is getting saturating? should i put some load on secondary winding.

Valuable reply will be appreciated.
Thanks
deepak

 

[haibin]

Hi All

I am also doing the same thing. In my case battery voltage is 12V and Fsw is 50 Khz.
My question is what will happen if kept the secondary winding open? in this condition i could not get the proper wave form of primary winding, there is to much spikes on the primary side also there is no effect of snubber in this contition.
Spikes is more then 100V is present.
Is the core is getting saturating? should i put some load on secondary winding.

Valuable reply will be appreciated.
Thanks
deepak

if you open the second wind, the transfomer be change to two connect inductor , when on of FET is off , the magnetize current must charge thecapacitor of FET Cds, if your transformer have no leakage inductor between two primary winds, the Voltage of one FET Cds must be clamped up to 2*Vin by body diode of another fet , if your transformer leakage inductor is too much , and your magnetize current is bigger enough , the energy of leakage inductor can charge the voltage of fet's Cds too high maybe broken your fet .
you can decrease your leakage inductor and increase your snubber to decrease the peak voltage of fet.