zero-voltage switching (ZVS) in boost converter

[NewBuilder]

Hi, I am trying to implement zero voltage switching in a boost converter but I'm not quite sure how to especially in spice. I think I get the concept about it creating a resonant voltage oscillation and switching at the zero crossing but I'm not sure how to actually implement it. Any help would be appreciated.

 

[BradtheRad]

Do you mean that coil current should drop to zero before a new cycle begins? Find the proper point where you can monitor voltage (or current). Tap at that point, and
it can serve as a feedback signal.

This simple boost converter is a good example how to do that. It takes feedback from the node between inductor and switch. It applies it as a control signal to the lefthand transistor.

Notice that coil current falls to zero between cycles.



This design isn't mine. I first saw it posted by a member at Edaboard.

 

[FvM]

I don't think that it's primarily a SPICE related question. You are asking about switcher topologies.

A Google search for "zvs boost" gives pretty much hits. You'll find a large variety of implemented switcher topologies. To realize zvs over a wide duty cycle range, you'll most likely need an auxiliary switch.

I believe that single switch boost converters (e.g. PFC) can achieve acceptable low switching losses without zvs techniques, that's the reason why they are rarely used in this context.

 

[NewBuilder]

A Google search for "zvs boost" gives pretty much hits. You'll find a large variety of implemented switcher topologies. To realize zvs over a wide duty cycle range, you'll most likely need an auxiliary switch.

I believe that single switch boost converters (e.g. PFC) can achieve acceptable low switching losses without zvs techniques, that's the reason why they are rarely used in this context.

You are right, it is more of a design/topology question. I found a design which uses an auxiliary switch. For me, even increasing the efficency 1-2% is a big deal. Plus it is supposed to remove a majority of the high-frequency content that I see when switching. I'll be using GaN FETS and a SiC diode. I'm trying to push switching as high as possible (500 kHz-MHz range). So I think ZVS will be the way to go. I read an IEEE paper where they got almost 99% efficiency using it!

Do you mean that coil current should drop to zero before a new cycle begins?

Thats an interesting design too. That would be zero-current switching (ZCS). I was looking into that as well but I've read a few places that ZVS gives better efficiency when also considering the gate signal. Using ZVS, you don't have to discharge the FET capacitance. Below is a link to a thesis I found. They outline everything that I think I would need but I cant seem to get it to work.

https://www.collectionscanada.gc.ca/obj/s4/f2/dsk3/ftp04/MQ59306.pdf

 

[FvM]

The quoted paper with the auxilay switch topology is under the first "zvs boost" Goggle hits. It's from 2000 and thus not particularly considering the requirements of SiC or even GaN switches. But this or a similar topology might work. The general point is that you need to trigger an auxiliary switch to prepare the zvs condition for the main switch, because with largely varying duty cycle you can't expect that a resonant circuit achieves the condition for all operation points.

The paper is detailed enough that you should be able to confirm or falsify the usability of the topolgy for your problem.

 

[NewBuilder]

The quoted paper with the auxilay switch topology is under the first "zvs boost" Goggle hits. It's from 2000 and thus not particularly considering the requirements of SiC or even GaN switches. But this or a similar topology might work. The general point is that you need to trigger an auxiliary switch to prepare the zvs condition for the main switch, because with largely varying duty cycle you can't expect that a resonant circuit achieves the condition for all operation points.

The paper is detailed enough that you should be able to confirm or falsify the usability of the topolgy for your problem.

I agree. I'm a newbie for real and I cant validate it ? sorry for the emoti but I really can't. I know it's not for WBG and that's why I was trying to get some help. I know that if I'm trying to do something novel I should do it on my own but I was just hoping for some help. I'm completely lost and I can't make it work. I really just want to do away with my high-frequency content. I can't figure out how other than this or other methods for ZVS. I really am desperate. I've been doing research and haven't got it yet. I also read a paper from pilowa which is an expert in the field and he used it too. I just don't know what to do to get rid if it. I will keep trying other things though

- - - Updated - - -

I don't want to seem like I'm just looking for answers. I really am trying. Everything in the paper makes sense though. I can see the benefit in zvs over vcs. I know both are better. I just don't understand how to do it. I really am trying though.

 

[asdf44]

Well given that context you'd be much better off sticking with a hard switched PFC since it's the dominant solution. You can find far more resources on a 'regular' PFC or find a consultant much more easily. Hard switched PFC's are used 'everywhere' so unless you have an extremely exotic application there are ways to make it work for you.

There are also many 'PFC Modules' which are off-the-shelf bricks. They'll provide application notes or EMI filter modules that should provide a complete solution.

**broken link removed**

 

[NewBuilder]

It's dc-dc. Would you use a pic for that?