How can I design a simple DC-DC buck step down converter with feedback

[Fluffyboii]

crutschow Hi, I tried replicating your circuit for experimenting in ltspice. It worked fine for 0.5ohm load but it started increasing oscillations when load is decreased. I tried different mosfets and op amps but couldn't fix the problem. Also when I go and bought the components they sold me the wrong mosfet saying yours didn't exist and I thought it was similar due to its model number but it was a npn one :/

 

[crutschow]

It worked fine for 0.5ohm load but it started increasing oscillations when load is decreased.

What do you mean "increasing oscillations"?

I tried different mosfets and op amps

My circuit uses a comparator, not an op amp.

Post you .asc file.

 

[Fluffyboii]

What do you mean "increasing oscillations"?

My circuit uses a comparator, not an op amp.

Post you .asc file.

There was an increasing sinus wave in output voltage when I chose a load resistance higher than 0.5ohm. Today I run it again but the issue was gone for some reason. Still, there is the file, It is not supposed too be like this.

 

[crutschow]

Why are you using an op amp in place of a comparator?
An op amp is not appropriate for this circuit as they do not make good comparators..

 

[Fluffyboii]

Why are you using an op amp in place of a comparator?
An op amp is not appropriate for this circuit as they do not make good comparators..

I couldn't find the model. I have the real thing in hand though. If I am not wrong you used a quad comparator. Does that mean you only use on from that packade or parallel connected all of them.

 

[crutschow]

Does that mean you only use on from that packade

Yes, you use only one from the package.
The rest you can ignore.

You also didn't include the CD4050 driver.
Will that be in the built circuit?

 

[Fluffyboii]

Yes, you use only one from the package.
The rest you can ignore.

You also didn't include the CD4050 driver.
Will that be in the built circuit?

Yes I also purchased it, only missing thing is the mosfet because they gave me the wrong one. Are you sure that there is no way to convert it to work with npn mosfets.

 

[ChrisP58]

Yes I also purchased it, only missing thing is the mosfet because they gave me the wrong one. Are you sure that there is no way to convert it to work with npn mosfets.

N-channel mosfets need to be turned on by making the gate pin more positive than the source pin. They can be used, and often are, in buck converters but the gate drive circuitry is more complex.

 

[Fluffyboii]

N-channel mosfets need to be turned on by making the gate pin more positive than the source pin. They can be used, and often are, in buck converters but the gate drive circuitry is more complex.

This actually gave me an idea. Since it npn works when the gate pin is positive I can lower the pwm signal just enough to activate the gate when it gets to peak voltage. Should be able to do it with a linear converter with voltage divider if my logic is not wrong. I still don't understand how pnp works in general tho. It seems like it is not worth doing.

 

[ChrisP58]

Generally, bipolar transistors come in NPN and PNP types, and have terminals called Base, Emitter and Collector.
MOSFETS come in N-channel and P-channel types and have terminals called Gate, Source and Drain.

 

[crutschow]

This actually gave me an idea. Since it npn works when the gate pin is positive I can lower the pwm signal just enough to activate the gate when it gets to peak voltage. Should be able to do it with a linear converter with voltage divider if my logic is not wrong.

Do you mean an N-MOSFET? NPN refers to a bipolar transistor, not a MOSFET.
As we've tried to explain that won't work.
You can try it if you like, but you will just be wasting your time.

To use an N-MOSFET you need a bootstrap driver that can generate a voltage higher then your power supply voltage.

still don't understand how pnp works in general tho.

A P-MOSFET requires that the gate be more negative than the source to turn on.
So if there is the positive supply voltage on the source, then if you ground the gate, the MOSFET will turn on.
If you raise the gate voltage to the same as the source voltage it will turn off.

 

[Fluffyboii]

Do you mean an N-MOSFET? NPN refers to a bipolar transistor, not a MOSFET.
As we've tried to explain that won't work.
You can try it if you like, but you will just be wasting your time.

To use an N-MOSFET you need a bootstrap driver that can generate a voltage higher then your power supply voltage.

An P-MOSFET requires that the gate be more negative than the source to turn on.
So if there is the positive supply voltage on the source, then if you ground the gate, the MOSFET will turn on.
If you raise the gate voltage to the same as the source voltage it will turn off.

All right, what I meant from npn mosfet was N-mosfet So in this graph green is the comparator output, red is P-mosfet output. So is it turning on when there is a negative voltage difference between the source and gate? I guess I will order the mosfet online.

Edit: Couldn't find the FDR840P for sale. They sell irf840 which is an N-Mosfet they sold me instead of the fdr840 p-mosfet. Since it is not sold online I assume finding it in one of the electronic stores would be even more difficult which I failed to do so before. Can you recommend me a P-mosfet replacement, as you can understand I don't know about mosfets I always thinked them as beefy transistors but that seems to be wrong. What about the IRF9640 for example, is it a suitable replacement. Seems to work fine in the simulation and has enough current rating.

 

[crutschow]

What about the IRF9640 for example

That has an On-resistance of 0.5Ω which will dissipate near 30W at 10A output, requiring a large heatsink and giving very poor efficiency.

Any P-MOSFET with at least a 20V rating and an On-resistance no more than about 10mΩ should work.

 

[crutschow]

I don't know about mosfets I always thinked them as beefy transistors

They are transistors and many are beefy, but they operate quite differently from MOSFETs.
Bipolar Junction Transistors (BJT) have a low impedance input (the Base) and have a significant input current through the base when ON (as determined by the transistor Beta current gain)
Metal Oxide Semiconductor Field-effect Transistor (MOSFET) have a very high resistance input (the Gate) and draw no current when ON (but the gate has a high capacitance which must be charged and discharged when turning on and off, thus requiring a low-impedance driver for fast switching).

Suggest you read some tutorials on both BJTs and MOSFETs to better understand how they work, if you are going to design with them.

 

[Fluffyboii]

They are transistors and many are beefy, but they operate quite differently from MOSFETs.
Bipolar Junction Transistors (BJT) have a low impedance input (the Base) and have a significant input current through the base when ON (as determined by the transistor Beta current gain)
Metal Oxide Semiconductor Field-effect Transistor (MOSFET) have a very high resistance input (the Gate) and draw no current when ON (but the gate has a high capacitance which must be charged and discharged when turning on and off, thus requiring a low-impedance driver for fast switching).

Suggest you read some tutorials on both BJTs and MOSFETs to better understand how they work, if you are going to design with them.

IRF4905PbF seems to have R on of 0.02 ohms and says it is a fast switching mosfet. It is also quite cheap compared to the average price of a mosfet. I don't really have many options for buying P-Mosfets online and I need a working prototype soon so I guess I am going with this one for now. I will fist use small 220mH inductors to create a working low power circuit than wind my own inductor for more power on a ring ferrite core.

 

[crutschow]

IRF4905PbF seems to have R on of 0.02 ohms and says it is a fast switching mosfet.

That should work.
The simulation shows it will dissipate near 1W of power so may need a small heatsink if you want to keep it below the boiling point.

 

[Fluffyboii]

That should work.
The simulation shows it will dissipated near 1W of power so may need a small heatsink if you want to keep it below the boiling point.

I have purchased a heatsink beforehand, will use that one.

 

[Fluffyboii]

That should work.
The simulation shows it will dissipate near 1W of power so may need a small heatsink if you want to keep it below the boiling point.

I have a problem with the circuit. The comparator leaks voltage from outside pin to output of the circuit and consumes a lot of current. My diodes ended up burning every time I tried the circuit and I have followed the problem till the comparator. I just don't know why comparator outputs around 10 volts when the inverting input is higher then the non inverting input.

 

[crutschow]

I just don't know why comparator outputs around 10 volts when the inverting input is higher then the non inverting input.

Have you tried a different comparator?

Post the schematic of the exact circuit you are testing.

 

[Fluffyboii]

Have you tried a different comparator?

Post the schematic of the exact circuit you are testing.

Its your circuit. with the LM339 and IRF4905. Just the comparator alone pulls enough current to cause a spark when there is nothing connected to it except power. I will get LM393 and test with that one. I designed a simple PCB and it should be made soon.