Synchronous Boost Converter

[RuneHansen]

Hey everybody

i am now trying to make a synchronous boost converter, but i have no experience with them...?
i can't really find detailed information on the internet about them.

what kind of mosfet is needed (to replace the diode) ?? n-channel or p-channel? and why?

and will this work, with the right controller?(schematic in picture).



thanks you in advance

 

[BradtheRad]

There is a problem with making a synchronous boost converter.

It is easy to bias the first transistor on and off. (This is the transistor which connects the inductor to ground for the first half of the cycle.) It can be an N-device. The clock pulses can be low voltage.

It is not so easy with the second (output) transistor, whether it is an N or P device. You must make the bias voltage at least as high as your output voltage. This is required to shut off a P-device.

OTOH, you can use an N-device. However to turn it on, you must bias it at an even greater voltage than your output voltage.

These factors must be solved when designing a control circuit for the boost converter.

 

[red_alert]

Actually, it could be as simple as this:



It's a half-bridge design, so you could use any monolithic low-side/high-side MOSFET driver (IR2110) and a cheap MCU/arduino to generate the PWM signal and to manage the output voltage feedback control.

The MOSFETs are commonly used N-channel type.

 

[treez]

also LT8705 can do it.........and others you will see at the list at end of its datasheet.

 

[RuneHansen]

thanks for the response.

the boost converter is powered by a 3.7v (maks 4.2v) li-ion battery. (load from 0.10-2.00 ohm)

is it possible to drive the boost converter mosfets with a normal mosfet driver, because i cant find any H-bridge ic with a operation between 3-12v

What about the LM5110? could this mosfet driver be used for this?

 

[red_alert]

The output stages of LM5110 have common Vcc and Vee connections thus you can't use one of the output as a high-side driver and the other one as a low-side driver.

You could use optocouplers and discrete drivers (BJT totem pole configuration) though. And, of course, use bootstrap technique for the high-side driver.

 

[treez]

you might need to have a housekeeping boost converter to give you a bias power rail...eg one of the ones by micrel. Then use a standard bootstrap high side drive sync boost converter

 

[red_alert]

There is low gate-source threshold (1.8V) MOSFETs which could be used with the actual 3.7V rail voltage.

 

[RuneHansen]

thanks guys.

what do you all think about this circuit? will it work?

 

[RuneHansen]

PS: my Primary boost converters volt out is 10v

 

[schmitt trigger]

Your idea is good and has merit.
I would make sure to study the start-up behavior though. The time where your boosted output is building up.

 

[RuneHansen]

how do i test the start up, without building the circuit?

 

[schmitt trigger]

Find the model for the component(s) and insert them in the simulator of your choice.
At one point of time you'll have to bite the bullet and actually build a prototype.

 

[RuneHansen]

Okay thanks.

One last thing: how much higher volt, must have bias voltage be for the high-side drive, if the primary boost output is about 10v

Thanks in advance ��

- - - Updated - - -

Okay thanks.

One last thing: at what volt, must The bias voltage be for the high-side drive, if the primary boost output is about 10v

Thanks in advance

- - - Updated - - -

Bad grammar ��

 

[schmitt trigger]

Depending on the Mosfet type, for full channel enhancement the gate will require anything from 3 to 10 volts above your source voltage.
This would mean from 13 to 20 volts with respect to ground.

But this function is precisely what the charge pump does.

 

[RuneHansen]

I'm using n-channel enhanced mosfet, so would 10v from the secondary boost converter to the driver, be fine?

Can't remember? But does NCP5901 have charge pumps? And if it does, would 10v driving it be optimal �� or just good practise?

Thanks for all your info ��

 

[RuneHansen]

Would This circuit work?

hallo.

Im working on a synchronous boost converter, and this is what i have come up with:
it is for my FIRST product, so please help



but, what do you guys think?

will it work?