Hi autir,
The cap between input and ground is supposed to protect from self-oscillation in the regulator. Self-oscillation is very destructive and heat generating and should be avoided at any cost. And the output cap will protect against load variations. I do recommend using both of them and always try to have the shortest leads and distance possible between the regulator and the caps.
Yes, I suppose you could use them upside down if you like! ;D
The cap between input and ground is supposed to protect from self-oscillation in the regulator. Self-oscillation is very destructive and heat generating and should be avoided at any cost. And the output cap will protect against load variations. I do recommend using both of them and always try to have the shortest leads and distance possible between the regulator and the caps.
Yes, I suppose you could use them upside down if you like! ;D
audioguru2
- Apr 6, 2004
- 12,026
- Joined
- Apr 6, 2004
- Messages
- 12,026
audioguru2
- Apr 6, 2004
- 12,026
- Joined
- Apr 6, 2004
- Messages
- 12,026
Hi Autir,
1) Self-oscillation occurs with some circuits like a regulator that have a very high gain and a lot of negative feedback because of phase-shifts in the circuit and load.
2) The output cap improves transient response because it absorbs sharp transients that the fairly slow regulator can't deal with. If a regulator's load is suddenly disconnected, its output voltage will have a increasing-voltage transient until the slow regulator's circuit "catches-up" with it. Likewise, if a regulator's load is suddenly applied, its output voltage will have a decreasing-voltage transient.
An output transient also occurs if the input voltage suddenly changes.
I think that the LM317 regulator's datasheet shows these transients and the smoothing effect of an output cap.
3) An output cap minimises the effects of very high frequency load variation because it has a low impedance (good filtering) at high frequencies. The very high gain and a lot of negative feedback in the regulator minimises low frequency load variations and also slow input voltage fluctuations (ripple).
1) Self-oscillation occurs with some circuits like a regulator that have a very high gain and a lot of negative feedback because of phase-shifts in the circuit and load.
2) The output cap improves transient response because it absorbs sharp transients that the fairly slow regulator can't deal with. If a regulator's load is suddenly disconnected, its output voltage will have a increasing-voltage transient until the slow regulator's circuit "catches-up" with it. Likewise, if a regulator's load is suddenly applied, its output voltage will have a decreasing-voltage transient.
An output transient also occurs if the input voltage suddenly changes.
I think that the LM317 regulator's datasheet shows these transients and the smoothing effect of an output cap.
3) An output cap minimises the effects of very high frequency load variation because it has a low impedance (good filtering) at high frequencies. The very high gain and a lot of negative feedback in the regulator minimises low frequency load variations and also slow input voltage fluctuations (ripple).
Are the values of the capacitors of significance?
The absorbing of sharp transients in the output line is improved with the capacitance; a bigger capacitor would perform much better. Why just 1 pF?
Similarly for the input capacitor (is it an RF-lowpass filter to filter out AC?).
The absorbing of sharp transients in the output line is improved with the capacitance; a bigger capacitor would perform much better. Why just 1 pF?
Similarly for the input capacitor (is it an RF-lowpass filter to filter out AC?).
audioguru2
- Apr 6, 2004
- 12,026
- Joined
- Apr 6, 2004
- Messages
- 12,026
Hi Autir,
You should use the capacitors recommended on the datasheet by the IC's expert designers. They know that a ceramic disk cap is much better at very high frequencies than an electrolytic cap, and that larger ceramic disks are also poorer.
They recommend a 0.1uF ceramic disk output cap, not only 1pF. 0.1uF is one hundred-thousand pF.
The input cap is needed when the main filter cap is at a distance because the wiring has inductance which causes a high impedance at high frequencies and the regulator needs a low impedance at its input to remain stable. Big ceramic disk caps have inductance and electrolytic ones have much more.
You should use the capacitors recommended on the datasheet by the IC's expert designers. They know that a ceramic disk cap is much better at very high frequencies than an electrolytic cap, and that larger ceramic disks are also poorer.
They recommend a 0.1uF ceramic disk output cap, not only 1pF. 0.1uF is one hundred-thousand pF.
The input cap is needed when the main filter cap is at a distance because the wiring has inductance which causes a high impedance at high frequencies and the regulator needs a low impedance at its input to remain stable. Big ceramic disk caps have inductance and electrolytic ones have much more.
audioguru2
- Apr 6, 2004
- 12,026
- Joined
- Apr 6, 2004
- Messages
- 12,026
Trigger, a 0.1uF cap has a value that is much too low to smooth mains ripple. The main filter cap will have a value of thousands of microfarads to do that.
As the datasheet says, the 0.1uF cap at the regulator's output improves its transient response. Transients are abrupt, fast and high frequency changes in the output voltage. A 0.1uF cap filters those high frequencies very well.
As the datasheet says, the 0.1uF cap at the regulator's output improves its transient response. Transients are abrupt, fast and high frequency changes in the output voltage. A 0.1uF cap filters those high frequencies very well.
What I mean is both Ceramic capacitor and Electrolytic capacitors are worked together as a DC smoothing.
You are right audioguru, 0.1uF Ceramic capacitors are as a low pass filter to shunt high frequency to ground while electrolyic capacitors are more like to handle the charge and discharge time of the DC voltages during the rippling.
You are right audioguru, 0.1uF Ceramic capacitors are as a low pass filter to shunt high frequency to ground while electrolyic capacitors are more like to handle the charge and discharge time of the DC voltages during the rippling.
A
Alun
- Jan 1, 1970
- 0
Also ceramic capacitors have a lower effective series resistance than electrolytics.
audioguru2
- Apr 6, 2004
- 12,026
- Joined
- Apr 6, 2004
- Messages
- 12,026
Hi Autir,
I have used BC Components (used to be made by Philips and Siemens) 370 series polyester caps for audio coupling for many years. "They work better" than ceramic or electrolytic for audio because they have very low distortion. For a power supply, distortion doesn't matter and since the polyester cap is small it performs as well as a ceramic cap. Either cap will be fine.
You should use the 0.33uF input cap anyway since the regulator will need it when the electrolytic cap dries out.
I have used BC Components (used to be made by Philips and Siemens) 370 series polyester caps for audio coupling for many years. "They work better" than ceramic or electrolytic for audio because they have very low distortion. For a power supply, distortion doesn't matter and since the polyester cap is small it performs as well as a ceramic cap. Either cap will be fine.
You should use the 0.33uF input cap anyway since the regulator will need it when the electrolytic cap dries out.
???
audioguru2
- Apr 6, 2004
- 12,026
- Joined
- Apr 6, 2004
- Messages
- 12,026
Hi Autir,
He, he. ;D
That's right, the electrolytic will leak or just "dry out" in a few years or many years depending on its quality. Then the regulator will go crazy without an input cap and maybe destroy your load.
He, he. ;D
That's right, the electrolytic will leak or just "dry out" in a few years or many years depending on its quality. Then the regulator will go crazy without an input cap and maybe destroy your load.
I always thought that electronics are indestructible by time, and the only way to damage them is improper voltage.
So, for example, a computer motherboard will become unusable when its electrolytic capacitors dry out?
What happens when the cap dries out? Is its capacitance reduced, or its maximum voltage ?
So, for example, a computer motherboard will become unusable when its electrolytic capacitors dry out?
What happens when the cap dries out? Is its capacitance reduced, or its maximum voltage ?
Similar threads
