Synchronous Boost Converter - High Side PMOS Switch

[aabundle]

Hello,

I was just wondering about the orientation of the High side switch in a synchronous boost converter.

For PMOSFETs the source terminal is usally tied high, so would the following configuration be correct?

**broken link removed**

Should the PMOSFET source and drain be switched around?? Because a higher voltage will be seen on the output of the boost converter??

Cheers

Michael

 

[aabundle]

**broken link removed**

 

[crutschow]

What type of picture are you posting? I can't see it as it's showing on my computer as a red X.

 

[RCinFLA]

Don't forget body diode of MOSFET from D to S. P-Ch should have source to output capacitor on boost but don't rely on body diode for initial conduction. Put a parallel fast diode as if you have a non-syncronous boost. You can turn on MOSFET to reduce parallel diode drop. Parallel diode will also establish initial output filter cap charge to give MOSFET G-S turn on voltage.

Syncronous boost are a little trickly on timing. You need coil current sense to know how to control MOSFET rectifier. Be conservative on conduction timing of MOSFET to be sure to avoid cross conduction with boost N-ch MOSFET. You have the parallel diode to take the initial conduction and end conduction period leaving safety margin for turning on and off rectifier MOSFET.

 

[aabundle]

https://imageshack.us/photo/my-images/638/notdone.jpg/

 

[aabundle]

What type of picture are you posting? I can't see it as it's showing on my computer as a red X.

Sorry crutschow here is a link to the pic

https://imageshack.us/photo/my-images/638/notdone.jpg/

 

[aabundle]

Don't forget body diode of MOSFET from D to S. P-Ch should have source to output capacitor on boost but don't rely on body diode for initial conduction. Put a parallel fast diode as if you have a non-syncronous boost. You can turn on MOSFET to reduce parallel diode drop. Parallel diode will also establish initial output filter cap charge to give MOSFET G-S turn on voltage.

Syncronous boost are a little trickly on timing. You need coil current sense to know how to control MOSFET rectifier. Be conservative on conduction timing of MOSFET to be sure to avoid cross conduction with boost N-ch MOSFET. You have the parallel diode to take the initial conduction and end conduction period leaving safety margin for turning on and off rectifier MOSFET.

Hey RC,

Here is a pic of the CCT

https://imageshack.us/photo/my-images/638/notdone.jpg/

I have a shotkky across the PMOS as the parallel diode and have allowed dead time between the switching signals of each of the MOSFET's as to avoid any shoot-through.

 

[simonbramble]

Try the LTC3786: http://cds.linear.com/docs/Datasheet/3786f.pdf

They use an N channel FET and high side drive it. They use the body diode as the standard rectifier diode, then short this out with the main FET channel once this diode has started to conduct to obtain better effiency. However, with this architecture (and the one described above with the Schottky diode) you cannot disconnect the input from the output. If this is a desirable feature, use a buck-boost controller (LTC3780/3789)

 

[aabundle]

Try the LTC3786: http://cds.linear.com/docs/Datasheet/3786f.pdf

They use an N channel FET and high side drive it. They use the body diode as the standard rectifier diode, then short this out with the main FET channel once this diode has started to conduct to obtain better effiency. However, with this architecture (and the one described above with the Schottky diode) you cannot disconnect the input from the output. If this is a desirable feature, use a buck-boost controller (LTC3780/3789)

Hey Simon,

I have actualy looked at this chip before, but my input voltage is from a Li-ion battery (4.2 - 2.75) volts out, so it rules alot of the integrated chips that are available (and would be much easier). The intended output is a regualted 5V at a maximum 1.7 Amps. I have looked at some low input voltage Boost converter IC's from Texas Instruments but they have maximum current outputs of around 1A.

In my original design i used a NMOS for the High side switch but lacked the voltage from the controller or the battery to use a high side driver, so instead i picked a PMOS and inverter the control siganl from the uC.

 

[simonbramble]

Pah! Forget that Texan lot. Stay in the Valley. What about the LT3112: https://cds.linear.com/docs/Datasheet/3112f.pdf

 

[aabundle]

Pah! Forget that Texan lot. Stay in the Valley. What about the LT3112: https://cds.linear.com/docs/Datasheet/3112f.pdf

Hahahaha, Alrighty then. That chip is spot on its $9.78 from Farnell too (plus the 100 externl components ). Thank you sir.

So is it safe to assume you prefer Linear over TI?

 

[aabundle]

nice website BTW. Ahhh i remember doing OP-AMP noise calcs in uni.