Buck-Boost-Converter

Thread Starter

TowerDefender

Joined Mar 15, 2023
7
I am currently designing an buck-boost-converter that has an input voltage range from 10V to 50V and an output voltage range from 10V to 50V. It should deliver a output power of 500W over the output voltage range. So I decided to use this coil to suit my needs.

Coil: https://www.vishay.com/docs/59061/ipla.pdf

My converter should use an buck-boost converter IC with four switching mosfets and so I decided to use this one:

Converter IC: https://www.analog.com/media/en/technical-documentation/data-sheets/ltc7878.pdf

So my problem is to find the right mosfet for my application. I already know that my mosfet should withstand an dc current of more then 60A (80A minimum) and a voltage of around 50V (75V as a save zone) but it should work with the converter IC so the gate threshold voltage should as low as it gets for a power mosfet. Also the values of Qg, RDS(on), tr, tf and coss should also be as minimal as possible for better efficiency (less switching losses).
I dont know if its possible to find a mosfet with those requirements also it should move in the save operating area.

example mosfet: https://www.mouser.at/datasheet/2/196/Infineon_IRFP4468_DataSheet_v01_01_EN-1732903.pdf

I hope sombody can tell me if there is a better mosfet as my example mosfet and maybe somebody can tell me what gate voltage the IC can deliver for the mosfet.
 

Papabravo

Joined Feb 24, 2006
21,158
I think your ability to achieve your goal is dubious at best. A boost ratio of 5, going from 10V to 50V will be nearly impossible to control for reasons articulated in other threads. There is also the input power requirement. This breaks down as follows:

Output Power = 500 Watts​
Presumed efficiency(estimate) = 85%​
Input Power (required): ≈ 589 Watts​
Input Current from 10V Supply 58.9 Amperes​

What do you suppose the maximum inductor current will be? The answer might surprise you. At those current levels can you be sure the inductor will not saturate?

A schematic, if you have one, would be most helpful. Without one we are just poking around in the dark.
 
Last edited:

ronsimpson

Joined Oct 7, 2019
2,985
Down load LTSPICE and run the attached file. It is set at 12V in 12V out at 10A.
Play with SPICE for a while before you start making component changes.
Here is a picture of the startup. The output starts at 0V and ramps to 12V @ 10A. The blue trace is the inductor current.
1678885622757.png
 

Attachments

Thread Starter

TowerDefender

Joined Mar 15, 2023
7
I think your ability to achieve your goal is dubious at best. A boost ratio of 5, going from 10V to 50V will be nearly impossible to control for reasons articulated in other threads. There is also the input power requirement. This breaks down as follows:

Output Power = 500 Watts​
Presumed efficiency(estimate) = 85%​
Input Power (required): ≈ 589 Watts​
Input Current from 10V Supply 58.9 Amperes​

What do you suppose the maximum inductor current will be? The answer might surprise you. At those current levels can you be sure the inductor will not saturate?

A schematic, if you have one, would be most helpful. Without one we are just poking around in the dark.
The inductor current DC will be around 50A (DC) and with ripple current it would go up to around 60A. The ripple current was set to be 20A (+/- 10A). Most inductors that I found with my requirements had a saturation current of more then 130A so no problem and sadly curently I dont have a schematic but I am at it to make one but it will take some time because I use Altium and I have to implement all parts into my library. As a side note its a free time hobby to make this so it can take a little longer till I have the schematic.
 

Thread Starter

TowerDefender

Joined Mar 15, 2023
7
The attached file is my current state of the ciruit I tested to implement a other mosfet but I dont get it to work and I have to implement two coils that I want to test but currently its seems to work for the simulation.
 

Attachments

Papabravo

Joined Feb 24, 2006
21,158
I have no idea what problems you are having. The canonical process of importing a library worked fine for me and took all of 5 minutes.
1679062483264.png
 

Papabravo

Joined Feb 24, 2006
21,158
The attached file is my current state of the ciruit I tested to implement a other mosfet but I dont get it to work and I have to implement two coils that I want to test but currently its seems to work for the simulation.
I don't understand what you are trying to do. You mention testing coils, but I'm not clear on the connection to the library in your post #10. Do you want to substitute a MOSFET from the library into your LTC7878 design?

ETA The symbol that I used is one that is automatically generated by LTspice. You can of course edit the graphic to be anything you want. I just wanted to make sure there was no problem with the library.
 

ronsimpson

Joined Oct 7, 2019
2,985
I am currently stuck at including this library because I want to use the IRF150P220
I am not good at adding new parts to the library. I know there is a way.
I would choose a new MOSFET, click on Rdson to order the parts. There are several parts with an on resistance in the 2.5mOhm range with 100V voltage rating. Pick one of them for now.
 

Thread Starter

TowerDefender

Joined Mar 15, 2023
7
I am not good at adding new parts to the library. I know there is a way.
I would choose a new MOSFET, click on Rdson to order the parts. There are several parts with an on resistance in the 2.5mOhm range with 100V voltage rating. Pick one of them for now.
The problem is not the Rdson I only wanted to implement all components that I want to use and testit with them. I also implemented the mosfet that @Papabravo already tested but with that mosfet the circuit needs ages till it ran through the simulation.
 

Papabravo

Joined Feb 24, 2006
21,158
The problem is not the Rdson I only wanted to implement all components that I want to use and testit with them. I also implemented the mosfet that @Papabravo already tested but with that mosfet the circuit needs ages till it ran through the simulation.
There are multiple ways to implement a MOSFET. As a subcircuit I would expect the simulation to take longer than an implementation with a .model card. My mistake I grabbed the 221 part implemented as a subcircuit. Let me see if there is an alternate choice.

This one looks better in terms of the lower Vgs(th) and the more gradual increase in drain current.

1679074648774.png
 
Last edited:

Papabravo

Joined Feb 24, 2006
21,158
The next thing you should do is check the system response to a load transient. This will tell you something about the stability of the system assuming you don't plan to determine this by other means.
 
Top