Assuming the 2 drivers are the same (both woofers) then adding an inductor in series (doesn't matter where) would create a 6dB octave low pass filter due to the inductors increasing impedance with frequency.
A cap would produce a 6dB/octave high pass filter due to falling impedance with frequency.
If the speakers are different (woofer+tweeter) then an inductor will prevent the tweeter from working and a cap would prevent the woofer from working.
A cap would produce a 6dB/octave high pass filter due to falling impedance with frequency.
If the speakers are different (woofer+tweeter) then an inductor will prevent the tweeter from working and a cap would prevent the woofer from working.
There are two types of First Order Crossover Networks - Series and Parallel. The type most often seen is the Parallel type, but there is no reason why it can't be done in Series.
In a Series circuit, the speakers are in series with each other and the crossover components are in parallel with the speakers.
In a Parallel circuit, the speakers are in parallel with each other and the crossover components are in series with the speakers.
See the attached diagram.
In a Series crossover, at high frequencies, the capacitor is a dead short and the coil is an open. Meaning that only the (HF) tweeter is in the circuit. At low frequencies, the coil is a dead short and the capacitor is open. Meaning that only the woofer (LF) is in the circuit.
At frequencies in-between, the capacitor and coil have varying resistances (impedances) according to frequency and the appropriate signal are still delivered to the appropriate speaker and the appropriate degree.
Think of it as a variable frequency controlled voltage divider. The coil and cap change impedance with frequency and create a voltage divider that balances the signal between woofer and tweeter to the degree appropriate at a given frequency.
Notice that it is wire opposite that of the Parallel circuit. In a Series, the cap is in parallel with the Woofer. In a Parallel circuit, the cap is in series with the Tweeter.
Even though a Series circuit is valid, to a logical mind, a Parallel just looks right, and is therefore more frequently used.
steve/bluewizard
In a Series circuit, the speakers are in series with each other and the crossover components are in parallel with the speakers.
In a Parallel circuit, the speakers are in parallel with each other and the crossover components are in series with the speakers.
See the attached diagram.
In a Series crossover, at high frequencies, the capacitor is a dead short and the coil is an open. Meaning that only the (HF) tweeter is in the circuit. At low frequencies, the coil is a dead short and the capacitor is open. Meaning that only the woofer (LF) is in the circuit.
At frequencies in-between, the capacitor and coil have varying resistances (impedances) according to frequency and the appropriate signal are still delivered to the appropriate speaker and the appropriate degree.
Think of it as a variable frequency controlled voltage divider. The coil and cap change impedance with frequency and create a voltage divider that balances the signal between woofer and tweeter to the degree appropriate at a given frequency.
Notice that it is wire opposite that of the Parallel circuit. In a Series, the cap is in parallel with the Woofer. In a Parallel circuit, the cap is in series with the Tweeter.
Even though a Series circuit is valid, to a logical mind, a Parallel just looks right, and is therefore more frequently used.
steve/bluewizard
Attachments
Quite true Steve.
An interesting point is that the series configuration, as you detail it, is that it's impossible to not sum to unity as the speakers are in series and the L/C just directs the frequencies. Whatever doesn't go to the woofer must go to the tweeter and vice versa.
On the other hand, it is possible to mess up a parallel configuration as the two filters are now independent and crossover frequency is set by L and Rwoofer for one and C and Rtweeter for the other. The tolerances of L and C and the flatness of Rwoofer and Rtweeter are critical.
An interesting point is that the series configuration, as you detail it, is that it's impossible to not sum to unity as the speakers are in series and the L/C just directs the frequencies. Whatever doesn't go to the woofer must go to the tweeter and vice versa.
On the other hand, it is possible to mess up a parallel configuration as the two filters are now independent and crossover frequency is set by L and Rwoofer for one and C and Rtweeter for the other. The tolerances of L and C and the flatness of Rwoofer and Rtweeter are critical.
Well, as long as the subject has come up, I found myself pondering the nature of this Series circuit wondering how you would design it.
One one hand, if we look at it as a pure voltage divider, we probably would want to use high impedance for the cap and coil. For example, at the crossover point, and assuming 8 ohm speakers, would we want both the coil and cap to be, say, 80 ohms each? This would keep the coil and capacitor out of the circuit. By that I mean, the coil and cap would have no (or extremely) little effect on the circuit impedance. 80 ohms in parallel with 8 ohms is essentially not there, but the combination of an 80 ohms coil and 80 ohm capacitor would still divide the voltage into equal parts for the woofer and the tweeter.
Now, if we use our Parallel crossover as a model, then at the crossover point, we want the capacitor and the coil to be 8 ohms. This makes sense when the cap and coil are in series with the speaker, but is it necessary when the cap and coil are in parallel with their respective speakers?
One thought that occurs to me, assuming we make the coil and cap 8 ohms at the crossover, and ignoring phase because it gives me a headache, is that we have an 8 ohm cap in parallel with an 8 ohm speaker, which is in series with another 8ohm/8ohm combination. Which in turn means 4 ohms in series with 4 ohms for a total circuit impedance of 8 ohms (again, ignoring phase angle).
Likely you want the crossover component impedance to match the speaker impedance, but is that really necessary?
Again, this has nothing to do with the original question, it was just an additional thought that occurred to me.
Steve/bluewizard
One one hand, if we look at it as a pure voltage divider, we probably would want to use high impedance for the cap and coil. For example, at the crossover point, and assuming 8 ohm speakers, would we want both the coil and cap to be, say, 80 ohms each? This would keep the coil and capacitor out of the circuit. By that I mean, the coil and cap would have no (or extremely) little effect on the circuit impedance. 80 ohms in parallel with 8 ohms is essentially not there, but the combination of an 80 ohms coil and 80 ohm capacitor would still divide the voltage into equal parts for the woofer and the tweeter.
Now, if we use our Parallel crossover as a model, then at the crossover point, we want the capacitor and the coil to be 8 ohms. This makes sense when the cap and coil are in series with the speaker, but is it necessary when the cap and coil are in parallel with their respective speakers?
One thought that occurs to me, assuming we make the coil and cap 8 ohms at the crossover, and ignoring phase because it gives me a headache, is that we have an 8 ohm cap in parallel with an 8 ohm speaker, which is in series with another 8ohm/8ohm combination. Which in turn means 4 ohms in series with 4 ohms for a total circuit impedance of 8 ohms (again, ignoring phase angle).
Likely you want the crossover component impedance to match the speaker impedance, but is that really necessary?
Again, this has nothing to do with the original question, it was just an additional thought that occurred to me.
Steve/bluewizard
Hi Thawach,
Didn't exactly understand your question so if my answer doesn't mean anything to you, apologies.
Richard Small was the first person I know of to use the series crossover described here in his seminal AES paper "Constant Voltage Crossovers" Vol 19, No.1, Jan 1971.
Steve,
You're on the right track. In both series and parallel circuits you're relying on the impedance vs. frequency characteristics of the L/C to send the frequencies to the right driver. In the series design, you want the cap across the woofer to shunt all the high frequencies to the tweeter and vice versa for the tweeter and L. So the C must be much greater than 8ohms at low frequency so the power goes to the 8ohm woofer and the inductor must be much less than 8ohms at low frequency to not voltage divide the signal too much. Hence the need for very low ESR components. At low frequency the whole speaker looks like an 8ohm woofer in series with a low impedance cap; at high frequency it looks like an 8 ohm tweeter in series with a low impedance inductor. The 8 ohm impedance is therefore maintained over the whole spectrum.
Mtn Bob: sorry to threadjack but it's all good stuff!
Didn't exactly understand your question so if my answer doesn't mean anything to you, apologies.
Richard Small was the first person I know of to use the series crossover described here in his seminal AES paper "Constant Voltage Crossovers" Vol 19, No.1, Jan 1971.
Steve,
You're on the right track. In both series and parallel circuits you're relying on the impedance vs. frequency characteristics of the L/C to send the frequencies to the right driver. In the series design, you want the cap across the woofer to shunt all the high frequencies to the tweeter and vice versa for the tweeter and L. So the C must be much greater than 8ohms at low frequency so the power goes to the 8ohm woofer and the inductor must be much less than 8ohms at low frequency to not voltage divide the signal too much. Hence the need for very low ESR components. At low frequency the whole speaker looks like an 8ohm woofer in series with a low impedance cap; at high frequency it looks like an 8 ohm tweeter in series with a low impedance inductor. The 8 ohm impedance is therefore maintained over the whole spectrum.
Mtn Bob: sorry to threadjack but it's all good stuff!
Series vs. parallel crossover topologies have very little to do with the original poster's question. Richie00boy answered it admirably.
Series crossovers have at least one serious limitation, they (at least 1st order) require very low DCR on the inductors or you will have excessive tweeter excursion (and resulting distortion) issues.
They don't "automatically sum to unity" unless the drivers are ideal and identical in all ways...and if they were you wouldn't need a crossover
There is an extent to which they compensate for non ideal conditions, but most of what is said about them by aficionados is voodoo nonsense.
Series crossovers have at least one serious limitation, they (at least 1st order) require very low DCR on the inductors or you will have excessive tweeter excursion (and resulting distortion) issues.
They don't "automatically sum to unity" unless the drivers are ideal and identical in all ways...and if they were you wouldn't need a crossover
There is an extent to which they compensate for non ideal conditions, but most of what is said about them by aficionados is voodoo nonsense.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.