hello,
This is my first speaker building and during the process of designing my ported enclosure I saw that my speaker will have an issue with the cone excursion of the woofer (Dayton audio TCP 115-4). I figured that I will solve the problem with an high pass filter to attenuate and remove the frequency bellow the tuning of my box.
From there (and with the help of windISD) I saw that I need a filter with cutoff frequency of 65Hz with -12dB/octave (so a second order filter).
So I was thinking about applying this filter between the preamp and the amp as this part of the spectrum don't not need to amplify.
I saw that a combination of a capacitor and an inductor can do this but when I try some calculation and as the input impedance of my amp is 10KOhms I came with some crazy values (something like 50H for inductor and 1.2F for capacitor).
I try to save money on this build and I was thinking before the amp was cheaper than after but with component of this value it's a failure :-(
My calculation :
C=1/(4*Z*Pi*Fc)
L=Z/(Pi*Fc)
C = capacitor value in Farad
Z= impedance
Fc = cutoff frequency
L = inductor value in Henry
so :
C=1/(4*10000*Pi*65)
C=1.22F
L=10000/(Pi*65)
L=48.97H
I don't know if my calculation are totally false or if there is a better solution to do this ? maybe active filter ?😕
thanks you for your help !
This is my first speaker building and during the process of designing my ported enclosure I saw that my speaker will have an issue with the cone excursion of the woofer (Dayton audio TCP 115-4). I figured that I will solve the problem with an high pass filter to attenuate and remove the frequency bellow the tuning of my box.
From there (and with the help of windISD) I saw that I need a filter with cutoff frequency of 65Hz with -12dB/octave (so a second order filter).
So I was thinking about applying this filter between the preamp and the amp as this part of the spectrum don't not need to amplify.
I saw that a combination of a capacitor and an inductor can do this but when I try some calculation and as the input impedance of my amp is 10KOhms I came with some crazy values (something like 50H for inductor and 1.2F for capacitor).
I try to save money on this build and I was thinking before the amp was cheaper than after but with component of this value it's a failure :-(
My calculation :
C=1/(4*Z*Pi*Fc)
L=Z/(Pi*Fc)
C = capacitor value in Farad
Z= impedance
Fc = cutoff frequency
L = inductor value in Henry
so :
C=1/(4*10000*Pi*65)
C=1.22F
L=10000/(Pi*65)
L=48.97H
I don't know if my calculation are totally false or if there is a better solution to do this ? maybe active filter ?😕
thanks you for your help !
For a RC filter, with R = 10000 ohms, the capacitor value you calculated is almost correct but it should be divided by 10 000 000.
If you want to do a LC filter, then the impedances required for calculation are those of the capacitor and inductor at 65 Hz. The two form a voltage divider which varies with frequency to reduce the high frequencies. You will need to choose a value for the capacitor or inductor and then calculate the value of the other using your cutoff frequency. There are online calculators that can be used.
If you want to do a LC filter, then the impedances required for calculation are those of the capacitor and inductor at 65 Hz. The two form a voltage divider which varies with frequency to reduce the high frequencies. You will need to choose a value for the capacitor or inductor and then calculate the value of the other using your cutoff frequency. There are online calculators that can be used.
I find 173.1378 nF and 34.6276 H when the filter is driven from an ideal voltage source, only resistively terminated on the output side and you want a Butterworth response. The inductance is impractical because of the low frequency and high impedance.
An active filter is indeed more practical. You could make a second-order Sallen and Key stage or a second-order multiple-feedback stage.
An active filter is indeed more practical. You could make a second-order Sallen and Key stage or a second-order multiple-feedback stage.
Yes, micro-Farad, not Farad.
Low frequency makes big coil.
Low frequency makes big cap.
Low impedance makes smaller coil and bigger cap.
Coils get to be a worse problem than caps. We want low impedance.
What impedance can your *source* drive?
Many solid-state preamps and other sources can drive 1K, have internal impedance like 100 Ohms. Because the internal impedance is uncertain, and we want a well defined impedance to the filter, we pad-out the source with a little less than 1K and call the total 1K. We also have a shunt load, the power amp, and we pencil this as 10K.
A moment's poking of the Idiot Assistant got the attached plan, which is roughly correct, especially since we can not afford a "2 Henry" choke of high precision. It will look like a tube amp output transformer, but the gap has to be carefully adjusted (with a hammer) to get near 2H, and this will still vary with signal level (which means bass distortion).
This is why we "always" do such tasks with opamp filters. We get an exact response with nearly ideal parts of low cost and small size. At the bottom of the audio band, active filters have been THE technique even back in WWII when an "active" meant real money.
Low frequency makes big coil.
Low frequency makes big cap.
Low impedance makes smaller coil and bigger cap.
Coils get to be a worse problem than caps. We want low impedance.
What impedance can your *source* drive?
Many solid-state preamps and other sources can drive 1K, have internal impedance like 100 Ohms. Because the internal impedance is uncertain, and we want a well defined impedance to the filter, we pad-out the source with a little less than 1K and call the total 1K. We also have a shunt load, the power amp, and we pencil this as 10K.
A moment's poking of the Idiot Assistant got the attached plan, which is roughly correct, especially since we can not afford a "2 Henry" choke of high precision. It will look like a tube amp output transformer, but the gap has to be carefully adjusted (with a hammer) to get near 2H, and this will still vary with signal level (which means bass distortion).
This is why we "always" do such tasks with opamp filters. We get an exact response with nearly ideal parts of low cost and small size. At the bottom of the audio band, active filters have been THE technique even back in WWII when an "active" meant real money.
Attachments
Thank you all for replying and for the informations !
Do you have any advices or step by step process to build an active filter like this ? or should I buy it somewhere ?
I'm a totally new in this field and I don't know how to choose the right components.
thank you one more time for your answers !
Do you have any advices or step by step process to build an active filter like this ? or should I buy it somewhere ?
I'm a totally new in this field and I don't know how to choose the right components.
thank you one more time for your answers !
For a Sallen and Key high-pass, you can calculate the component values with an online calculator like Sallen-Key Low-pass Filter Design Tool. The attachment is example output of this tool.
For the filter capacitors, low-tolerance polypropylene or polystyrene foil capacitors or low-tolerance C0G/NP0 multilayer ceramic capacitors are the best you can get. Plain old MKT capacitors are not quite as good, but still work quite well. Ceramic class II capacitors (such as X7R, X5R, Z5U, Y5G) are to be avoided as filter capacitors: they are quite inaccurate and they distort.
For the filter resistors, metal film through-hole resistors or thin-film SMD resistors are preferred, although carbon film through-hole or thick-film SMD also works well.
The op-amp can be any old unity-gain stable audio op-amp that doesn't have too much input bias current. An OPA134 (single) or OPA2134 (dual) should work fine, or a TL071 or TL072 if you want to keep it cheap.
If there is a long cable between the filter and the main amplifier, you may have to add a resistor of 100 ohm or so between filter output and cable. This helps to keep the op-amp stable despite the cable capacitance.
At the input of the filter, you can add a small passive low-pass filter to get rid of any RF rubbish and also to help keep the op-amp stable. For example something like 100 ohm and 100 pF.
The supplies of the op-amp(s) will need to be decoupled. 10 uF aluminium electrolytic capacitors placed close to the op-amp(s) will do, or 100 nF class-II ceramic when the supply is reasonably nearby.
Finally you need a supply. Usually regulated +/- 15 V supplies are used, +/- 12 V is equally suitable. I'm sure a search engine can find some examples for you.
For the filter capacitors, low-tolerance polypropylene or polystyrene foil capacitors or low-tolerance C0G/NP0 multilayer ceramic capacitors are the best you can get. Plain old MKT capacitors are not quite as good, but still work quite well. Ceramic class II capacitors (such as X7R, X5R, Z5U, Y5G) are to be avoided as filter capacitors: they are quite inaccurate and they distort.
For the filter resistors, metal film through-hole resistors or thin-film SMD resistors are preferred, although carbon film through-hole or thick-film SMD also works well.
The op-amp can be any old unity-gain stable audio op-amp that doesn't have too much input bias current. An OPA134 (single) or OPA2134 (dual) should work fine, or a TL071 or TL072 if you want to keep it cheap.
If there is a long cable between the filter and the main amplifier, you may have to add a resistor of 100 ohm or so between filter output and cable. This helps to keep the op-amp stable despite the cable capacitance.
At the input of the filter, you can add a small passive low-pass filter to get rid of any RF rubbish and also to help keep the op-amp stable. For example something like 100 ohm and 100 pF.
The supplies of the op-amp(s) will need to be decoupled. 10 uF aluminium electrolytic capacitors placed close to the op-amp(s) will do, or 100 nF class-II ceramic when the supply is reasonably nearby.
Finally you need a supply. Usually regulated +/- 15 V supplies are used, +/- 12 V is equally suitable. I'm sure a search engine can find some examples for you.
Attachments
- Status
- Not open for further replies.