High Side Drive For 2 Switch Forward Converter?

[codex653]

ok so i've been looking at a forward converter and i'm loving the looks of things so far. problem is tho, i've read about a problem with the high side gate driving...apparently it's a floating gate and is harder to turn on and off? how would i solve this? i want to use a direct gate drive from the PWM IC, but i know that's going to make the mosfet sluggish...i really don't want to use an IC for the gate driver (after the PWM) and would like to go the discrete way. All a gate driver's purpose is is to provide a higher current to the gate of the mosfet than what the PWM IC can produce right? The increased current = a faster charge time which = faster switch on time correct? So on the high side gate driver, does it just provide the gate with more current than the low side gate? or am i getting things mixed up here.

 

[codex653]

i found some examples of concepts for a discrete gate driver with transistors to quickly turn on and off the device...look on page 14..would any of these work for a 2 switch forward converter?

https://www.electro-tech-online.com/custompdfs/2010/12/slup169-1.pdf

my circuit i would want to build is this right here...the mosfets are IRF640

https://i.cmpnet.com/powermanagementdesignline/2008/01/nationaltws2.jpg

 

[ronv]

Assuming you are using N channel fets on the high side you need to get the gate 10 volts above the source to keep the fet on. What this means is you need a voltage 10 volts above your supply voltage. In addition to doing the level shifting from logic to the higher voltage the high side driver will generate this voltage either by charge pump or bootstrap or both. Keep reading past page 14. If I were you I would start looking for high side drivers.

 

[crutschow]

I don't believe your referenced circuit will work properly. You have two transistors in series with the winding to turn the voltage on and off, but that will not work unless you are building a flyback type inverter. And for a flyback you only need one transistor. The transformer polarity must also be reversed for flyback operation and you don't need Lo or D2.

To use the transformer in the normal transformer mode, the primary and secondary can have no average DC current. The transformer signal must be equal plus and minus voltage, such as a normal sine-wave generates. A bridge circuit driving the primary would do that. A center tapped primary with the center-tap going to Vin, and alternately grounding the two winding ends with transistors will also work.

 

[codex653]

oh wow...this has gotten way more complex than i thought it would come out to be....hmmm i'll have to think about going forward with this type of converter...pun intended heheh
wait a sec...if the gate voltage would have to be atleast 10 volts higher than the source, then that would put my gate voltage at 170v roughly!!!! yikes....my mosfets can only withstand +/- 20v on the gate...crap this is not going to work....i guess i'm back to my buck-boost converter...

 

[ronv]

oh wow...this has gotten way more complex than i thought it would come out to be....hmmm i'll have to think about going forward with this type of converter...pun intended heheh
wait a sec...if the gate voltage would have to be atleast 10 volts higher than the source, then that would put my gate voltage at 170v roughly!!!! yikes....my mosfets can only withstand +/- 20v on the gate...crap this is not going to work....i guess i'm back to my buck-boost converter...

The gate voltage will "float" 10 volts above the source voltage. So yes 170 volts, but only 10 volts above the source when the fet is on.

 

[codex653]

so....if the output is meant to be 30vdc...then the gate will be at 40 volts?? even still that would be too much..my mosfet would still blow

 

[ronv]

Are you trying to do something like this?

International Rectifier - Product Information Power Integrated Circuits

 

[codex653]

wow! yah actually that would be perfect! but definitely not in that pin configuration, i can't do that small soldering stuff but alas, i still run into the problem with the gate voltage on my mosfet...haha i guess i would just have to get some different ones! any suggestions on mosfets that can handle gate voltages of around +/-200v? i've never actually seen those before ...but granted i've never needed a mosfet like this either haha they would also have to be able to handle 5A continuous and around 8-10A peak

 

[crutschow]

No MOSFETs will handle that amount of gate voltage.

But it's rather pointless, if you circuit design doesn't work.

 

[codex653]

ok.... i'm a bit confused then.... so if the voltage at the gate is referenced to the source voltage and the source voltage is 30, then if you have a driver that outputs 10v to the gate, will the gate voltage be 10 or 40???

 

[crutschow]

ok.... i'm a bit confused then.... so if the voltage at the gate is referenced to the source voltage and the source voltage is 30, then if you have a driver that outputs 10v to the gate, will the gate voltage be 10 or 40???

The gate to common (gnd) is 40V. The gate to source voltage is 10V. It's the gate to source voltage that is specified in the data sheets, and that is the voltage you are concerned about.

 

[codex653]

The gate to common (gnd) is 40V. The gate to source voltage is 10V. It's the gate to source voltage that is specified in the data sheets, and that is the voltage you are concerned about.

ahhhhh ok! i get it now! so since my mosfet has a gate voltage rating of +/- 20v, it would be completely ok to have a 10 v signal driving it! ok thank you so much for the clarification!