I'm not quite sure which one of these filters is better or .if they have the same performance. Both have the same total resistance and capacity. (View in Fixed font)
+5V in -----10 ohms----+-----------+------ out | | 100uF 100uF | | ground ground+5V in -----20 ohms----+------ out | 200uF | ground
Mac
(Oops! My fixed font was not... fixed. This one should be better )
I'm not quite sure which one of these filters is better or .if they have the same performance. Both have the same total resistance and capacity. (View in Fixed font)
+5V in -----10 ohms----+-----------+------ out | | 100uF 100uF | | ground ground+5V in -----20 ohms----+------ out | 200uF | ground
Mac
Do you mean to have another resistor in there" 10 ohms total in example 1,
20 in 2. Try writing the Vo/Vi equation when you show all the resistors. Do you know how to do the math with complex impedances?
No they don't. I see two different resistors.
Assuming you want to know which has a lower cutoff frequency, its the 20 ohm version. But whether that is better for your application can't be answered without more information. Source and load impedances, load current, frequency characteristics of the noise, ripple voltage limits and a bunch of other stuff is going to affect the final performance of either filter.
If you meant to compare two filters with the same resistor value, then the simple answer is they are equivalent as long as frequencies are low enough that the components can be treated as ideal lumped elements. If that's not true (and its not true above tens of kilohertz) then the pair of parallel capacitor would usually have lower total ESR and will filter low frequency ripple and noise better.
If you intend to use this filter in a power system, I suggest some smaller ceramics in parallel to shunt the higher frequencies which the larger electrolytic caps won't help with. Tantalum electrolytics will become increasingly useless above a hundred kilohertz or so, aluminum will give out at even lower frequencies. Its not uncommon to use several different smaller ceramic values in parallel to cover a wider bandwidth in sensitive applications.
Steve
Dear God. I should take some sleep, there is indeed a second 10 ohm resistor has corrected below:
+5V in -----10 ohms----+----10 ohms-------+------ out | | 100uF 100uF | | ground ground +5V in -----20 ohms----+------ out | 200uF | groundI can deal with impedances but with complex impedances, I don't think so :(
I forgot the second 10 ohm resistor in the circuit (as shown below). I want to add a filter between a USB 5V supply from my laptop to a USB sound card. The 5V is contaminated with switching noises. The USB sound card has a microphone input that is quite sensitive and the noise "leaks" through in the sound. Using battery cures the problem but I want to use the USB power. So the noise is more high frequency but still in audio range, few Khz at most!
+5V in -----10 ohms----+----10 ohms-------+------ out | | 100uF 100uF | | ground ground +5V in -----20 ohms----+------ out | 200uF | ground
You're going to lose your +5V supply regulation for the sound card as the card draws varying current, and drops will occur across the resistors.
Just how bad is the "contamination", and what's its frequency makeup? You might be able to get away with just using a single capacitor close by the sound card.
The first version is a two-pole filter, and at high frequencies, it rolls off at 40 db/decade, while the second one rolls off at 20 db per decade. The second one has more attenuation at lower frequencies, around 100 Hz. They have equal attenuation at about 500 Hz. If you are concerned about noise at a few kHz while keeping the series resistance low, the first version is better.
-- john
The 1st filter is a 2 pole filter. Its ultimate attenuation is 12 dB/octave, and its ultimate phase shift is -180 degrees (assuming a resistive load). There will be a frequency above which this filter will have more attenuation than the 2nd filter. This frequency depends on the load. . The 2nd filter is a single pole filter. Its ultimate attenuation is 6 dB/octave, and its ultimate phase shift is -90 degrees (assuming a resistive load). . The equations for the frequency responses of the filters are rather lengthy. Re-post if you want them.
"Jon" a écrit dans le message de news: snipped-for-privacy@y43g2000cwc.googlegroups.com...
Uhhh???
-- Thanks, Fred.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.