Buck Converter Design

[Dmitriy20]

Hello,

I’m designing a step-down dc-dc converter.
The output voltage must be adjustable from 2 to 10 volts and capable of sourcing 120 ma at
any of the voltages in the specified range with no more than 50 millivolts ripple. The
efficiency of the regulation must be better than 50% over the entire voltage range at full
load. The unregulated supply voltage is +15 volts. Resistive load from 10Ω to 1 kΩ.

The control is done through a 555 timer to generate triangular train . The switching frequency is 12.6 kHz.
LM339 - comparator, and LM741 - error amplifier.

I have tried to make it work, but it doesn’t. I'm don't know how to close the feedback loop properly. Can anyone help me to close feedback loop. All suggestions are acceptable. I have attached the circuit design.

Thank you very much for your help.

 

[mneary]

1) Microsoft Word "schematics" are inconvenient for those of us who don't have Micro$$$oft Office on every computer.
2) The input to U4 is outside its common mode range. It should not be given 0V or 15V.
3) U4 cannot get close enough to the positive rail to give you proper turn on of Q3 and Q4.
4) Q3 and Q4 are 2N3904 which are rated at only 200mA but are virtually useless above 100mA.
5) The 560 ohm load on U2A is about 25mA but the LM339 is rated tor 6mA.

Please write the equation for the output of U3. U3 = f (V1, Vout)

 

[Dmitriy20]

I cannot use any other transistors because professor doesn't allow to do that. That's why I put 2 of them in parallel, trying to split current equally.
The U4 is just voltage follower, I tried without it, but results are the same, and the pulses from LM339 are distorted. (in the beginning of each pulse spike happens)

 

[Dmitriy20]

Buck Switching Regulator

I attached pdf file!

 

[Electronworks]

why not use a dedicated chip - 33063 is low cost with all the compensation built in.

 

[mneary]

Please write the equation for the output of U3. U3 = f (V1, Vout)

 

[Dmitriy20]

Buck Converter

I just fix problem, now pulses from comparator properly turn on transistors.

I atached new schematic

The Vo of U3 will be:

Vo = (( 1 + ( R17 / R14 )) * Vpositive) - ((R17 / R14) * Vnegative)

 

[Dmitriy20]

why not use a dedicated chip - 33063 is low cost with all the compensation built in.

My professor doesn't allow me to use different IC's. So I'm stuck with a 555-timer, LM741, and LM339

 

[mneary]

I think you're saying that it works?

You could consider putting the values for R14, R17, etc. (and V1, R11, R12, and Vout) into the equation for Vo(U3). If you have spreadsheet software this is an ideal use for it.

 

[mneary]

By the way, you have R14 and R17 reversed, and you need to include the Thevenin equivalent of R11 and R12 into the value of R17.

 

[Dmitriy20]

I think you're saying that it works?

You could consider putting the values for R14, R17, etc. (and V1, R11, R12, and Vout) into the equation for Vo(U3). If you have spreadsheet software this is an ideal use for it.

I just know the basic model of feedback, so that equation is my rough idea!

 

[mneary]

Your circuit has feedback, but with R14 only 10 ohms you have almost no gain. That's why I'm leading you to write the gain equation. You need enough gain in U3 so that its output changes when Vout (the top of RL) changes.

 

[Dmitriy20]

By the way, you have R14 and R17 reversed, and you need to include the Thevenin equivalent of R11 and R12 into the value of R17.

I just derived equation, can you check it?

Vo (U3) = ( 1 + (( R14 / R17) * ( Rf2 / ( Rf1 + Rf2 )) * V1 )) - (( R14 / R17) * ( Rf2 / ( Rf1 + Rf2 )) * Vout )

 

[mneary]

( Rf2 / ( Rf1 + Rf2 )) * Vout // good. This term is the Thevenin voltage at the junction of R11 and R12.

-(R14/R17) // needs correction. You probably mean this to be the inverting gain of the amplifier, but R17 also needs to include the Thevenin resistance at R11/R12.

// This would be ((R14/(R17+(R11*R12)/(R11+R12))
// Now you can put the Vout term back together.

--- Take the V1 term apart in the same way and put it back together.

 

[Dmitriy20]

Thank you very much for help.

 

[Dmitriy20]

I'm planing to test it tomorrow in the lab!

 

[mneary]

Don't forget, I only helped you with half of the feedback equation. You have to finish it and fix the numbers before it will work!

 

[Dmitriy20]

Don't forget, I only helped you with half of the feedback equation. You have to finish it and fix the numbers before it will work!

What are you mean when you said it's only half of the feedback equation?

Another half covers what?

 

[mneary]

The right side haws two terms. One is contributed by Vout. The other is contributed by V1. The Vout term is finished, but I have left the V1 term for you to finish.

I am hoping to see that you understand what was done to create the Vout term as you repeat the process on the V1 term and put them together.

 

[Dmitriy20]

Like this:

Vo (U3) = ( 1 + (( R14 / ( R17 + (( Rf1 * Rf2 ) / ( Rf1 + Rf2 )))) * ( Rf2 / ( Rf1 + Rf2 )) * V1 )) - (( R14 / ( R17 + (( Rf1 * Rf2 ) / ( Rf1 + Rf2 )))) * ( Rf2 / ( Rf1 + Rf2 )) * Vout )