Designing a boost converter

[Mercury]

boost converter

Hello,
I'm designing a boost converter and I'm trying to use a mosfet instead of the diode. Here's the basic idea:

Well, this doesn't work, as the M3 mosfter conducts current in both directions regardless of the drive signal. I suppose this is because of the integrated diode. I have seen mosfets used instead of diodes, but this doesn't seem to work in this case. Is there anything that I could do?

Best regards!

 

[max0412]

boost converter diode

Flip your high side mosfet around so the drain is at the output.For n-fets the source must be more negative then the drain or the body diode conducts. You could also use a p-fet for your high side switch. See picture schematic below.

You know you will need a boostrap drive for the high side n-fet?. There are controllers that have this integrated or you can purchase one separately like FAN7382:

https://www.fairchildsemi.com/ds/FA/FAN7382.pdf

 

[Mercury]

200w boost converter

Thanks, that did the trick. Yes, I know how I will have to drive the mosfets. This application note from TI was really helpfull for that:
**broken link removed**

Thanks again!

 

[max0412]

how does boost converter work

One thing you may want to do is put a Diode suitably rated from input to output. The diode will conduct at start-up charging your output capacitors. This will minimize the stress on your HS Fet (M3). If the diode isn't used the charging current will go through the body diode.

You can see from the schematic I uploaded they are using a Schottky to do this. I usually put a rectifier anode at the input cathode at the output caps. The diode will cease conducting when the output becomes more positive then the input by about a volt.

 

[eem2am]

zero cross detector boost converter

i am sure you can get dedicated IC's for use with your fet in place of diode in boost. They will in turn have app notes giving you good help.

 

[eem2am]

power loss in diode boost converter

i think there is a problem in disontinuous mode when you are doing boost with fet in place of diode

..this then thus means that you should be careful with a ccm boost at very light load as it will be in dcm.

you see.....

think what happens when the boost transistor stops conducting in dcm......
..the inductor starts discharging through your high-side fet...........

BUT

How does your high side FET know when to turn off?

if its using a diode then its ok cuzz the diode turns off as it gets reverse biased by the output voltage.

your fet does not have this ability....so it will stay ON..........and your output caps will start conducting back towards your input.....bad news...

so you will need to put in some extra circuitry to detect when your inductor current has gone to zero.....and then command the fet to turn off.......the detection will involve an extra detector winding on your boost inductor........and some circuitry or an IC that can use it wisely.

......are you sure you dont want to just put a few nice, low loss, schottky diodes in parallel?

becasue after all your power is only 10 watts

 

[chatwitdebo]

designing boost converter

hi all,
please i want to design a boost converter
Vin=100Vdc;
Vout=150Vdc~400Vdc;
output power is 200W, max
please does anyone have circuit ideas, especially how to choose an appropriate inductor, seems to be a very important factor.

 

[culho]

diode for boost converter

i agree with eem2am.
use a zero crossing detector.
you can either measure the voltage across the Nfet V= rds*Isense or put a resistor in series with the Nfet and measure its current.

Cheers