# Voltage regulation problem

#### psoke0

Joined Mar 31, 2017
174
hello guys . im trying to control the voltage of my power supply with this circuit.

when i remove the cap the average voltage comes to about 1.7 volts and it doesnt change with load resistance. but when i add a cap output voltage changes with load resistance and also cap value why is that

#### Attachments

• 123.7 KB Views: 3

#### ElectricSpidey

Joined Dec 2, 2017
1,814
Because the cap is charging to the peak values, instead of the average...which I assume is what you want as an output.

When you place the load, the cap discharges and the output changes.

#### psoke0

Joined Mar 31, 2017
174
Because the cap is charging to the peak values, instead of the average...which I assume is what you want as an output.

When you place the load, the cap discharges and the output changes.
no i never put out the load. i dont run the curcuit with no load .
well the peak value is 12 v not that changing values with differant loads.

#### ElectricSpidey

Joined Dec 2, 2017
1,814
Ok I see, so when there is no cap the average voltage is 1.7 when you place the cap it stores the peak voltage and releases it during the times where the output is 0...raising the average voltage.

The rate in which it dumps current into the load varies with the load resistance, therefore affecting the average voltage.

Also the bigger the cap, the more charge it has to dump, also affecting the average voltage.

Sorry, my teaching skills are a little weak.

#### psoke0

Joined Mar 31, 2017
174
Ok I see, so when there is no cap the average voltage is 1.7 when you place the cap it stores the peak voltage and releases it during the times where the output is 0...raising the average voltage.

The rate in which it dumps current into the load varies with the load resistance, therefore affecting the average voltage.

Also the bigger the cap, the more charge it has to dump, also affecting the average voltage.

Sorry, my teaching skills are a little weak.
actually you are great teacher now i understand so in order to make stable voltage with cap in curcuit i have to put a feedback for the Pulse signal right
i just want to make the curcuit give me the variable voltage i want by controling PWM signal Duty cycle

#### crutschow

Joined Mar 14, 2008
27,750
just want to make the curcuit give me the variable voltage i want by controling PWM signal Duty cycle
If you are using PWM for better efficiency, that circuit won't do it.
For high efficiency you need an inductor in series with the MOSFET and before the output capacitor.
With no inductor, the efficiency will be no better than a linear regulator, since the power will be dissipated in the on-resistance of the MOSFET when it is on. rather than being stored in the inductance to continue charging the capacitor when the MOSFET is off.

Last edited:

#### BobTPH

Joined Jun 5, 2013
3,680
That is not a voltage regulator, with or without the capacitor.

Bob

#### Papabravo

Joined Feb 24, 2006
16,828
You can't regulate something you're not even looking at. You have no feedback "mon".

#### dl324

Joined Mar 30, 2015
13,141
im trying to control the voltage of my power supply with this circuit.
You have no "control" element. How is this circuit supposed to control output voltage?

#### crutschow

Joined Mar 14, 2008
27,750
How is this circuit supposed to control output voltage?
It's a PWM circuit.
It will "control" the voltage but it won't "regulate" it.

#### psoke0

Joined Mar 31, 2017
174
If you are using PWM for better efficiency, that circuit won't do it.
For high efficiency you need an inductor in series with the MOSFET and before the output capacitor.
With no inductor, the efficiency will be no better than a linear regulator, since the power will be dissipated in the on-resistance of the MOSFET when it is on. rather than being stored in the inductance to continue charging the capacitor when the MOSFET is off.
i see so i have to make buck regulator with pwm control right ? but isnt inductor value will mess with it since wh cant change inductor value because i want from 0 volts to 12 volt adjustment in voltage

It's a PWM circuit.
It will "control" the voltage but it won't "regulate" it.
yep thats right. i typed topic wrong. i want to control the voltage

#### MrAl

Joined Jun 17, 2014
8,479
hello guys . im trying to control the voltage of my power supply with this circuit.
View attachment 231566

when i remove the cap the average voltage comes to about 1.7 volts and it doesnt change with load resistance. but when i add a cap output voltage changes with load resistance and also cap value why is that
Sorry to say this but you can neither control nor regulate the DC output voltage using that circuit, at least not in the way we normally think of controlling or regulating.

First, you know already it is not regulated because there is not yet any feedback, so i wont mention too much more about that.

However, even controlling the DC output voltage will not work very well. That is because there will always be some up and down of the output because the filtering is not adequate for a normal clean DC output. It also puts a lot of stress on the input and the MOSFET. You are seeing a lot of up and down on the output because that circuit is not the right way to do this even if you dont want feedback.

The reason you are seeing that up and down on the output is because the cap charges up very quickly and so it receives a lot of energy all in a very short time. When the MOSFET turns off, the output load then discharges the cap somewhat slowly and so you see a time when the output is very high (nearly the same as the input voltage) and a time when the output is falling slowly back toward zero.
It is unfortunate but there is no way around this without adding something like an inductor which would create a buck circuit as you know, and that is not a bad idea at all and you shouldnt have to swap out the inductor once you get the design right.
The inductor will smooth out the bumps enormously and will get the output looking much much better.

The kind of circuit you are using currently without modification is mainly used for output loads that are mostly resistive and do not have very tight requirements on the voltage specification. They can take an up and down votlage with wide variations without too much of a problem even if it still isnt the best way to do it. For a DC power supply though that's got to be one of the worst ways of doing it because the output ripple is the highest it could be which is usually the thing you really dont want.
A typical load for the kind without an inductor (like the one you show) is a regular incandescent light bulb where the pulsing does not bother the bulb too much. If you tried to drive an LED though it would probably blow it out unless it could take significant over voltage which most regular LEDs can not take because the current goes up too high.

Of course the input source has a lot to do with this too. Because the circuit during the 'on' time is so stiff the input source characteristics plays a big part in what happens at the output too (especially during the 'on' time of the MOSFET). If the source is also very stiff and very little inductance, the output will bang up high and then fade low as you are seeing. If the input source is soft the output may not bang up as high because the input source will limit the charge current to the cap.

Just as a point of design theory, if the source is an ideal source and the MOSFET is an ideal switch and the cap is an ideal capacitor, when the MSOFET turns on the current out of the source and through the mosfet and through the cap is infinite. That kind of logical idealistic analysis usually tells us something is wrong right away.

So in the end if you dont use an inductor you wont get a very nice output DC voltage even if you dont really want to regulate it.
In some cases you can use a series resistor to take the place of the inductor, but the efficiency still suffers and there is still going to be significant up and down on the output unless you get the resistance just right and the frequency high enough, and because of that series resistance you will never be able to get near the full output.

The buck circuit is the simplest type of switching power circuit to design. Dont be afraid to go there

Last edited:

#### crutschow

Joined Mar 14, 2008
27,750
cant change inductor value because i want from 0 volts to 12 volt adjustment in voltage
The inductor value won't need to change.
One inductor will work over the complete voltage range.

#### psoke0

Joined Mar 31, 2017
174
Sorry to say this but you can neither control nor regulate the DC output voltage using that circuit, at least not in the way we normally think of controlling or regulating.

First, you know already it is not regulated because there is not yet any feedback, so i wont mention too much more about that.

However, even controlling the DC output voltage will not work very well. That is because there will always be some up and down of the output because the filtering is not adequate for a normal clean DC output. It also puts a lot of stress on the input and the MOSFET. You are seeing a lot of up and down on the output because that circuit is not the right way to do this even if you dont want feedback.

The reason you are seeing that up and down on the output is because the cap charges up very quickly and so it receives a lot of energy all in a very short time. When the MOSFET turns off, the output load then discharges the cap somewhat slowly and so you see a time when the output is very high (nearly the same as the input voltage) and a time when the output is falling slowly back toward zero.
It is unfortunate but there is no way around this without adding something like an inductor which would create a buck circuit as you know, and that is not a bad idea at all and you shouldnt have to swap out the inductor once you get the design right.
The inductor will smooth out the bumps enormously and will get the output looking much much better.

The kind of circuit you are using currently without modification is mainly used for output loads that are mostly resistive and do not have very tight requirements on the voltage specification. They can take an up and down votlage with wide variations without too much of a problem even if it still isnt the best way to do it. For a DC power supply though that's got to be one of the worst ways of doing it because the output ripple is the highest it could be which is usually the thing you really dont want.
A typical load for the kind without an inductor (like the one you show) is a regular incandescent light bulb where the pulsing does not bother the bulb too much. If you tried to drive an LED though it would probably blow it out unless it could take significant over voltage which most regular LEDs can not take because the current goes up too high.

Of course the input source has a lot to do with this too. Because the circuit during the 'on' time is so stiff the input source characteristics plays a big part in what happens at the output too (especially during the 'on' time of the MOSFET). If the source is also very stiff and very little inductance, the output will bang up high and then fade low as you are seeing. If the input source is soft the output may not bang up as high because the input source will limit the charge current to the cap.

Just as a point of design theory, if the source is an ideal source and the MOSFET is an ideal switch and the cap is an ideal capacitor, when the MSOFET turns on the current out of the source and through the mosfet and through the cap is infinite. That kind of logical idealistic analysis usually tells us something is wrong right away.

So in the end if you dont use an inductor you wont get a very nice output DC voltage even if you dont really want to regulate it.
In some cases you can use a series resistor to take the place of the inductor, but the efficiency still suffers and there is still going to be significant up and down on the output unless you get the resistance just right and the frequency high enough, and because of that series resistance you will never be able to get near the full output.

The buck circuit is the simplest type of switching power circuit to design. Dont be afraid to go there
wow thank you this curcuit is not near complete ofcource i just want to learn from you guys.
The inductor value won't need to change.
One inductor will work over the complete voltage range.
okay i will put it in there and test it.
by the way this curcuit will not be like this i will put feedback curcuit to it and change duty according to the voltage i want as an output
you guys are very helpfull over the years to me. you might not remember but i asked alot of questions before to you guys
i have another question when i increase the frequency BJT transistor looks like cant handle more then 1 kHz is it true in real life too ?

#### psoke0

Joined Mar 31, 2017
174
If you are using PWM for better efficiency, that circuit won't do it.
For high efficiency you need an inductor in series with the MOSFET and before the output capacitor.
With no inductor, the efficiency will be no better than a linear regulator, since the power will be dissipated in the on-resistance of the MOSFET when it is on. rather than being stored in the inductance to continue charging the capacitor when the MOSFET is off.
hey crutschow do i need to convert this curcuit with buck converter with N channel mosfet instead of P channel i currently have in order to do this you said

#### psoke0

Joined Mar 31, 2017
174
guys is it how should look like without feedback yet ?

about 200 mV ripple how much more cap i have to add ? wow

Last edited:

#### psoke0

Joined Mar 31, 2017
174
okay guys i made some changes. when i test the first curcuit its efficiency is about 87% here it is

And with buck converter efficiency is about 82% ?? how is that possible

here it is

what am i doing wrong here guys

#### psoke0

Joined Mar 31, 2017
174

#### LowQCab

Joined Nov 6, 2012
1,110
Simply put a reference voltage into the FB (Feed-Back) Pin from your controller,
super easy, very efficient.

And I have to ask .......
Why do you want to control a Power Supply Voltage with a Micro-Controller ????
.
.
.