Hi ,
I have a generic query to ask.... i have some idea of full wave rectifiers....output of the filter cap is connected to a filter cap...... at the filter cap there would be ripple voltage depending on the value of the capacitor....but there is a limit to increase the value of capacitor.....How will we eliminate this ripple noise??....Will a voltage regulator do that?? is it because the ripple voltage is of low frequency??... How is this noise characteristics designed to voltage regulator??....
Please help.....

 

A voltage regulator eliminates essentially all of the ripple, and you use one when the circuit cannot tolerate the ripple that remains after a simple filter cap. Light bulbs and many motors are insensitive to ripple (work fine with or without it) but many other devices require a cleaner source voltage.

I don't understand your other questions.

 

I don't understand your other questions.

If ripple noise is eliminated by regulator how will we design the frequency characteristics of regulator...... Say i have a ripple frequency of 100hz......This could be eliminated by the feedback loop..... not lets say the ripple freq is 1Mhz.... possibly this 1Mhz noise would be present at the output regulator.....

 

the ripple for a full wave rectifyer is twice the line frequency. 50 hz gives 100 hz ripple, to get ripple in the mhz range, you would have to put a real high frequency into the bridge. the main purpose of the regulator is to regulate. if you try to pull too much current, you will probably get ripple first, beore it melts.

 

The spec sheet for a linear regulator will usually show the ripple rejection versus frequency from input to output.

 

Current is only delivered to the reservoir capacitor in short pulses of high current.
The current waveform is nothing like sinusoidal.
The bigger your ripple capacitor the larger in magnitude and narrower or shorter (in time) these pulses are.

Eventually the current pulses can be so large they exceed the ratings of the other components.

 

I'm wondering if everyone is talking past each other. The TS is discussing ripple frequencies greatly above any AC line I've ever heard of.

rakeshm: just what are you rectifying? From an AC power outlet or some switching supply? Some of the math is the same, some goes different directions depending on what circuit is being used.

 

If ripple noise is eliminated by regulator how will we design the frequency characteristics of regulator...... Say i have a ripple frequency of 100hz......This could be eliminated by the feedback loop..... not lets say the ripple freq is 1Mhz.... possibly this 1Mhz noise would be present at the output regulator.....

When talking about linear power supplies, ripple is at one or two times the power line frequency, such as the 100 Hz you mention. Switching power supplies have that on the primary side, plus ripple at the switching frequency on the secondary side. This is a much higher frequency (although usually not as high as 1 MHz), and is dealt with in a different manner. Which kind of power supply are you asking about?

ak

 

Hi ,
I have a generic query to ask.... i have some idea of full wave rectifiers....output of the filter cap is connected to a filter cap...... at the filter cap there would be ripple voltage depending on the value of the capacitor....but there is a limit to increase the value of capacitor.....How will we eliminate this ripple noise??....Will a voltage regulator do that?? is it because the ripple voltage is of low frequency??... How is this noise characteristics designed to voltage regulator??....
Please help.....

If you search for terms like "capacitor multiplier" and "ripple eliminator", you should find examples of a simple emitter-follower transistor circuit - the base is fed by a bias resistor and smoothed by an electrolytic, the emitter follows the base, which draws less current and only needs a small capacitor to smooth the ripple.

 

Thank you guys,
"ripple is at one or two times the power line frequency, such as the 100 Hz you mention"" ------- I just wanted to know the design considerations for rejecting ripple in a regulator..... My query was purely theoretical and has no practical significance..... Question just boils down to how did a regulator design its ripple rejection characteristics??..... How do they design ripple rejection graph?? I presume ian field points to this

If you search for terms like "capacitor multiplier" and "ripple eliminator", you should find examples of a simple emitter-follower transistor circuit - the base is fed by a bias resistor and smoothed by an electrolytic, the emitter follows the base, which draws less current and only needs a small capacitor to smooth the ripple.

 

Thank you guys,
"ripple is at one or two times the power line frequency, such as the 100 Hz you mention"" ------- I just wanted to know the design considerations for rejecting ripple in a regulator..... My query was purely theoretical and has no practical significance..... Question just boils down to how did a regulator design its ripple rejection characteristics??..... How do they design ripple rejection graph?? ......................

They design the regulator to have as high a ripple rejection as possible. It's part of the design that also determines the regulator output regulation due to any input voltage variation. The limit is likely determined by capacitive and other parasitic leakage from input to output, and also the finite gain and frequency response of the closed loop circuit gain.

The ripple rejection graph is not "designed", it is plotted from measurements on a real device by applying an input ripple of varying frequency and measuring the corresponding output ripple.

 

A linear regulator can use any of several techniques to remove incoming ripple from thee output. One is passive attenuation, with an R-C or L-C filter. Another is active attenuation with the capacitance multiplier mentioned above. But the main way a semiconductor linear regulator deals with ripple is by clipping it off. For example, let's say that the bulk capacitor has 18 VDC and 6Vp-p ripple on it. The positive peaks are 21 V and the negative peaks are 15 V. If this goes into a circuit whose output can never be above 12 V no matter what, then the ripple is effectively clipped off. This is like clipping in an audio amplifier, only on purpose. Almost all linear regulators work this way no matter what their internal design. The input can vary because of ripple, brownouts, adjustments, whatever; they hold the output at the design voltage no matter what. OK, no matter what, within reason. For every regulator IC and output voltage there is a minimum input voltage required. That difference is called the regulator's headroom. Also, these chips can respond to input changes only so fast, so they clip low frequencies much better than high frequencies.

ak

 

A linear regulator can use any of several techniques to remove incoming ripple from thee output. One is passive attenuation, with an R-C or L-C filter. Another is active attenuation with the capacitance multiplier mentioned above. But the main way a semiconductor linear regulator deals with ripple is by clipping it off. For example, let's say that the bulk capacitor has 18 VDC and 6Vp-p ripple on it. The positive peaks are 21 V and the negative peaks are 15 V. If this goes into a circuit whose output can never be above 12 V no matter what, then the ripple is effectively clipped off. This is like clipping in an audio amplifier, only on purpose. Almost all linear regulators work this way no matter what their internal design. The input can vary because of ripple, brownouts, adjustments, whatever; they hold the output at the design voltage no matter what. OK, no matter what, within reason. For every regulator IC and output voltage there is a minimum input voltage required. That difference is called the regulator's headroom. Also, these chips can respond to input changes only so fast, so they clip low frequencies much better than high frequencies.

ak

thank you

 

thank you

Its not just the integrated regulators that clip off the ripple - a simple zener diode shunt regulator does it too.

Buy they're only good for very low current draw, as soon as you draw much current the voltage drags down to where the ripple peaks are.