There are 2-Flavors of Regulators,
Linear, which is what you've been working with, and
Switch-Mode-Regulators,
which turn either fully On, or, fully Off, usually at very high Frequencies.

Both use Voltage-Feedback,
and depending on the design,
they may also use Current-Feedback.

Switch-Mode-Regulators require a properly designed Filter-Network on the Output,
but generally speaking,
most Linear-Regulators can be run with no Filtering of the Output.

If a Switching-Regulator is run without a proper Filter-Network,
all kinds of crazy problems are created,
and you will basically have really BAD Voltage-Regulation, and a ton of Electrical "Noise".
You might get away with this if your Load is strictly Resistive,
like a Heating-Element, or an Incandescent-Light-Bulb,
but if there is the slightest amount of Inductance, or Capacitance, in your Load,
you will have serious problems.

Switching-Regulators are usually much more efficient,
and therefore, they Dissipate much less Heat than a Linear-Regulator.

To make a Switch-Mode-Regulator work,
there is usually some additional complexity and Cost -vs- a Linear-Regulator's simplicity.

Also, many Op-Amps can NOT deal with being used as a Comparitor,
especially at High-Frequencies.

The LM358's that You were trying to use are ~$0.10 Cents each,
they are for very un-demanding jobs were CHEAP is the Top-Priority.

The LT1630CN8 Op-Amps that I specified are
the closest thing to a perfect Op-Amp that I've found,
and they ought to be, because the damn things cost ~$8.oo each,
but they have excellent manners in all situations,
so they're very easy to work with.
They will operate as a Comparitor at over 100khz without issues,
and have "Rail-to-Rail" Inputs and Outputs,
which also makes them work really well for DC-Circuits.
And, they're still available in a Through-Hole, 8-Pin-DIP-Package,
which makes them Hobbyist-Friendly.

If you are willing to spend another ~$40 to ~$50 Dollars,
I'll show You how to DIY your own Switch-Mode-Regulator.
.
.
.

thank you alot for this information. sorry if understand it wrong. so you are saying that using my opamp as comparator requaries these opamps to have very large bandwith of frequency. and they are not. so you are using op amp with negative feedback configuration with caps so that frequency is limited and doesnt alow curcuit be unstable . did i understand it correctly

 

Close, but not exactly.
The Capacitors in the Feedback-Loop are there to prevent any oscillations that may
occur up in the high Mhz Frequency region, on a very fast Op-Amp like the LT1630CN8,
they are not to "prevent" the Op-Amps from operating as a Comparitor.

The LM358 is NOT well-suited to being used as a Comparitor.

A proper, "real" Comparitor will usually have a Gain of over 1-Million, sometimes much higher.

If you will do the Math calculations,
the Current-Limiting-Section has extremely high-gain, ( gain of 1000 ),
which makes that Op-Amp act "almost" like a Comparitor.
This is enough Gain to make the Current-Limiter act almost like an On/Off-Switch,
but not so much Gain that it would create oscillation problems.
The High-Gain makes it respond extremely fast to Short-Circuiting of the Output,
and it will never allow more than ~15-Amps of Output-Current,
but it will do nothing with 14.99-Amps of Output-Current, ( if the Control is turned all the way up ).
The High-Gain allows it to work very well, even when set to less than 0.5-Amps.
Also, the High-Gain allows a very small Current-Sense-Resistor to be used.

The Voltage-Regulating Op-Amp has a Gain of ~50,
this lets the Voltage "sag" a little bit under heavy Loads.
If the Gain was much higher,
Output-Filter-Capacitors might be required for stability.
( Big-Output-Capacitors are always a good idea anyway )
.
.
.

 

Close, but not exactly.
The Capacitors in the Feedback-Loop are there to prevent any oscillations that may
occur up in the high Mhz Frequency region, on a very fast Op-Amp like the LT1630CN8,
they are not to "prevent" the Op-Amps from operating as a Comparitor.

The LM358 is NOT well-suited to being used as a Comparitor.

A proper, "real" Comparitor will usually have a Gain of over 1-Million, sometimes much higher.

If you will do the Math calculations,
the Current-Limiting-Section has extremely high-gain, ( gain of 1000 ),
which makes that Op-Amp act "almost" like a Comparitor.
This is enough Gain to make the Current-Limiter act almost like an On/Off-Switch,
but not so much Gain that it would create oscillation problems.
The High-Gain makes it respond extremely fast to Short-Circuiting of the Output,
and it will never allow more than ~15-Amps of Output-Current,
but it will do nothing with 14.99-Amps of Output-Current, ( if the Control is turned all the way up ).
The High-Gain allows it to work very well, even when set to less than 0.5-Amps.
Also, the High-Gain allows a very small Current-Sense-Resistor to be used.

The Voltage-Regulating Op-Amp has a Gain of ~50,
this lets the Voltage "sag" a little bit under heavy Loads.
If the Gain was much higher,
Output-Filter-Capacitors might be required for stability.
( Big-Output-Capacitors are always a good idea anyway )
.
.
.

after all of this very informative writing of yours i needed to gain some knowledge about what is oscillation is and why it is bad. now i need to learn more about something else i did not understand. i thought this curcuit does kind of pwm like thing. like when voltage drops opamp shorting the base to ground to get the set voltage. and when its higher then the set voltage then its puts the gate hıgh and its doing this very fast on off on off like pwm is this what is not going on here ? if it is then how is it linear curcuit. and should not get hot. which i know it does so i know it is linear but what is going on . i think i dont know what is linear means im just lost can you give more information to me please

 

Linear means that the FET is always "partially-On" / "partially-Off".
Switching means "fully-On" or "fully-Off".
Switching causes a "choppy" Output which must be Filtered.
Linear has a very smooth and quiet Output that requires very little Filtering, sometimes no Filtering.

A Switching-Regulator IS PWM, along with some other "tricks".
Your design,
if it would have worked like You imagined,
would have been a very crude, noisy, hard to control, pain in the butt.

A Switching-Regulator REQUIRES Output-Filtering,
usually a Choke-Coil, (Inductor), and Capacitors are required to make it work properly.
There are very good reasons why there are so many
different, dedicated, specialized Chips, just for Switching-Regulators

Oscillation ..............
Any Amplifier can be made to Oscillate all by its self.
All you have to do is delay the Output-Signal a little bit,
and then let some of that delayed Output-Signal get back into the Input,
now you have an instant Oscillator, super easy.

But you may not want an Oscillator,
so you have to do certain things to make sure that the Amplifier will NOT Oscillate.
You have to make sure that the Output of the Amplifier ONLY goes where You want it to go,
and that no un-wanted Electrical-Noise gets into the Input of the Amplifier.

Inductance, or Capacitance,
in the wrong-place, or in the wrong-amount, or in the wrong-combination,
can make any Amplifier Oscillate.
Oscillation is basically the Amplifier going completely crazy in a way that is not wanted.

Oscillation can be created and controlled on purpose,
and it must always be carefully controlled to achieve the desired result.

All Switch-Mode-Power-Supplies have at least one Oscillator built inside to
provide exact control of the PWM Frequency.
.
.
.

 

thank you so much i appreciate you