…and not resistors?
That is, why don’t passive x-overs use RC and CR filters, instead of LC and CL?
(And yes I know that speaker impedance is normally used in a CR configuration for high-pass filters, but my question is on a more theoretical level, just out of curiousity.)
That is, why don’t passive x-overs use RC and CR filters, instead of LC and CL?
(And yes I know that speaker impedance is normally used in a CR configuration for high-pass filters, but my question is on a more theoretical level, just out of curiousity.)
At the high currents associated with loudspeakers, using resistors with woofers would wastefully convert lots of energy to heat, leaving less to operate the woofer.
RC filters are used in low current applications like inside the electronic circuits as found in radio tuners etc.
Perhaps someone can back this up with theory?
RC filters are used in low current applications like inside the electronic circuits as found in radio tuners etc.
Perhaps someone can back this up with theory?
Last edited:
Here's a previous thread on the topic which I think you may have seen since you mention the RC high pass tweeter filter: https://www.diyaudio.com/community/threads/rc-filters-for-audio.342709/#:~:text=Using a passive RC crossover filter between the,damping of the woofer's response is seriously degraded.
Some seem to be happy to waste power if it gets them the result they want. However, an RC low pass filter that sets a low source impedance for the speaker, will be too difficult for the amplifier to drive.
Inductors are not as common in line level circuits since they are difficult to build for a high enough self resonance, and require care in implementation.. otherwise they might be more popular. Speaker crossover inductors are much easier to produce.
Inductors are not as common in line level circuits since they are difficult to build for a high enough self resonance, and require care in implementation.. otherwise they might be more popular. Speaker crossover inductors are much easier to produce.
It is indeed a power loss / impedance issue between the amplifier and speaker. To roll off highs, you would use a series resistance and capacitor to ground on the driver side. Yes, power loss is extreme.
In signal circuits, they do sometimes use inductors. Low frequency units are large and need lots of shielding. An inductor is a coil, and coils pick up stray magnetic fields really well! Additionally, large with expensive shielding makes them a very expensive circuit element.
Want to see inductors in audio circuits? Look at some early graphic equalizers.
Another consideration is that two reactive elements (L and C) roll off at -12dB/oct. A resistor - capacitor circuit rolls off at -6 dB/oct. One reactive component.
In signal circuits, they do sometimes use inductors. Low frequency units are large and need lots of shielding. An inductor is a coil, and coils pick up stray magnetic fields really well! Additionally, large with expensive shielding makes them a very expensive circuit element.
Want to see inductors in audio circuits? Look at some early graphic equalizers.
Another consideration is that two reactive elements (L and C) roll off at -12dB/oct. A resistor - capacitor circuit rolls off at -6 dB/oct. One reactive component.
Another point particularly significant for woofers / midbass etc.: when you insert series R in the low pass, you're also in effect artifically increasing the system Q, so as well as wasted power & needing to account for heat dissipation, you'll need to design the box / baffle / whatever load with that revised Q in mind -which in the case of cabinets will also mandate increased volume, since of the dominant Fs, Vas & [effective] Q factors in box design, Q tends to have a disproportionately large effect relative to the others.
You also have zero control anymore over the Q-factor of the filter.
Technically, cascading RC filters, can give you an higher order filer.
But the Q-factor wil be close to nothing at a certain point.
Exactly Scott!
That's one reason active systems sound so much better than using a passive crossover.
That's one reason active systems sound so much better than using a passive crossover.
In general terms I'd agree, since with a few exceptions as far as I'm concerned the less series / loop resistance there is in circuit the better. Although in fairness implementation (and facilities) will always play a big part too.
A resistor will use energy and get hot. It will also need to be large for higher wattage amps.
The inductors losses are very little and for higher power it just needs thicker wire.
The inductors losses are very little and for higher power it just needs thicker wire.
Simply because resistance is not a frequency dependent property and inductive reactance and capacitive reactance are.
'Just'. 😉
Agreed though -RC low pass filters are almost unknown at speaker level for this and all the other reasons mentioned.
Last edited:
It took a long time to get to this. An inductor is reactive and will only pass low frequencies. A resistor (other than the Q affects mentioned) only adjusts level.
dave
Last edited:
Hi Dave,
Exactly. At audio frequencies or a pure resistance you are 100% right.
Inductors:
Being able to use two complimentary reactive components saves on parts and effectiveness in passive crossovers. Not losing a ton of energy and messing up damping factors nearly as much as crossovers do already is another huge reason to do this. We do use resistors in zobel networks to dissipate power, thereby correcting impedance in driver voice coils.
Exactly. At audio frequencies or a pure resistance you are 100% right.
Inductors:
Being able to use two complimentary reactive components saves on parts and effectiveness in passive crossovers. Not losing a ton of energy and messing up damping factors nearly as much as crossovers do already is another huge reason to do this. We do use resistors in zobel networks to dissipate power, thereby correcting impedance in driver voice coils.
Inductors are impossible to avoid really.
Your drivers voice coil is a inductor to begin with😳
As for why not using RC filters for low pass in speakers, it was already explained quite well.
I can feel my physics professor cringe. There is no such thing as a slow frequency. Speed = Wavelength x Frequency. An inductor will pass signals with long wavelength (relatively speaking), speed being constant for the frequency range in question.
Last edited by a moderator:
To put this in mathematical terms:
With only R and C all the poles and zeros are real, ditto for just L and R, with L and C (and no R) all the poles and zeros are imaginary. In reality you have L and C and R in a speaker system as the load is partly resistive and components are not ideal, so the poles and zeros can be arbitrary complex frequencies(*). This is what allows controllable Q-factors and thus allow a usable response curve to be chosen (Butterworth, LR, etc etc).
(*) with a passive filter poles can never have positive real parts though.
Also active filters are different and you can do anything with just R and C in an active filter - its all to do with the differential equations that pop-out of the circuit analysis and whether they are first order or second order...
Last edited: