I like AnalogKid's suggestion of simply buying something based on the 2596HV. That's what I would do.

However, if you still want to build your own, this is what I meant by putting the MOSFET in the ground wire. This allows you to still use an N-channel, which is cheaper and better, without the need for high-side (exceeding +48V) drive. The driver chip is probably overkill, just one that I remembered off the top of my head - you can likely get by with a much weaker driver.

Generating the +10 volts for gate drive could be as simple as using an LM317AHV with 48 volts in, since efficiency is not a concern.

 

@Vorador

If you decide to build one here is about as simple as it gets from scratch.

Thank you so much, ronv! I really appreciate it!

Looks quite complicated, though.

 

I like AnalogKid's suggestion of simply buying something based on the 2596HV. That's what I would do.

However, if you still want to build your own, this is what I meant by putting the MOSFET in the ground wire. This allows you to still use an N-channel, which is cheaper and better, without the need for high-side (exceeding +48V) drive. The driver chip is probably overkill, just one that I remembered off the top of my head - you can likely get by with a much weaker driver.

Generating the +10 volts for gate drive could be as simple as using an LM317AHV with 48 volts in, since efficiency is not a concern.

View attachment 88327

That looks simple enough! What volts should be the pulses you have shown at the input of the driver?

If I understand you right, you're first converting 48v to 10v and then using the lowered voltage to drive the MOSFET. But I don't understand the reason for putting the drain to the lower wire? What difference would it make if it's connected to the +48v wire?

Could you please tell me any place where I could learn more about MOSFET drivers? I'm totally ignorant of their necessity, function and how to use them.

Thank you soooooooo much guys!

 

That looks simple enough! What volts should be the pulses you have shown at the input of the driver?

If I understand you right, you're first converting 48v to 10v and then using the lowered voltage to drive the MOSFET. But I don't understand the reason for putting the drain to the lower wire? What difference would it make if it's connected to the +48v wire?

Could you please tell me any place where I could learn more about MOSFET drivers? I'm totally ignorant of their necessity, function and how to use them.

Thank you soooooooo much guys!

The pulses into the driver are 5 volt logic level pulses, as might come from a microcontroller or CMOS logic. That also means you have to design the control circuitry to drive the logic. The control circuitry would adjust the duty cycle and/or frequency of the pulses, so that the output voltage stays around +24V, when the load varies. If your load is really constant in its current requirements, you might be able to just use an oscillator that you adjust by turning the knob on a potentiometer manually. For hobby work, maybe that's all you need.

If you use a power supply controller chip like the LM5005, the chip may include its own regulator, so you can power the whole thing off 48 volts without the need for an external regulator, and the chip will generate the right drive, too, meaning you wouldn't need a driver chip.

Why put the MOSFET in the negative wire? There aren't strong reasons, but here is the line of thought. N-channel MOSFETS are just better than P-channel. The capacitance and on-resistance are lower for a given size. It's rare to see P-channel power FETs in commercial designs. The N-channel requires a gate drive that is maybe 8-10 volts higher than the source of the transistor, for power supply purposes where you want to switch fast. That means that if you want to put an N-channel MOSFET in the +48V rail, you need +58V to drive it. It's not especially hard to generate +58V from +48V, but it's marginally harder than generating +10V from +48V. The biggest drawback of doing it this way is that there is no common ground between input and output, if that matters to your application.

All this assumes that you want to learn about switching supply design in the first place. Remember, you can just buy that module off eBay for a few dollars that does what you want.

 

The pulses into the driver are 5 volt logic level pulses, as might come from a microcontroller or CMOS logic. That also means you have to design the control circuitry to drive the logic. The control circuitry would adjust the duty cycle and/or frequency of the pulses, so that the output voltage stays around +24V, when the load varies. If your load is really constant in its current requirements, you might be able to just use an oscillator that you adjust by turning the knob on a potentiometer manually. For hobby work, maybe that's all you need.

If you use a power supply controller chip like the LM5005, the chip may include its own regulator, so you can power the whole thing off 48 volts without the need for an external regulator, and the chip will generate the right drive, too, meaning you wouldn't need a driver chip.

Why put the MOSFET in the negative wire? There aren't strong reasons, but here is the line of thought. N-channel MOSFETS are just better than P-channel. The capacitance and on-resistance are lower for a given size. It's rare to see P-channel power FETs in commercial designs. The N-channel requires a gate drive that is maybe 8-10 volts higher than the source of the transistor, for power supply purposes where you want to switch fast. That means that if you want to put an N-channel MOSFET in the +48V rail, you need +58V to drive it. It's not especially hard to generate +58V from +48V, but it's marginally harder than generating +10V from +48V. The biggest drawback of doing it this way is that there is no common ground between input and output, if that matters to your application.

All this assumes that you want to learn about switching supply design in the first place. Remember, you can just buy that module off eBay for a few dollars that does what you want.

Thank you so much, Mr. Roderick Young! Generating high voltage pulses was the one of the main issues. I never thought about putting the MOSFET to the ground wire to as a solution to the problem. I am usually not able to make circuits work, but I'll try your suggestions.

Thanks again. Don't know how to express my gratitude in words to all of you.