I thought the answer would be easier to find...
A first order crossover has a capacitor in series with the tweeter and an inductor in series with the woofer.
A second order has the above, plus an inductor in parallel with the tweeter and a capacitor in parallel with the woofer.
There are ultra simple crossovers that are just a series capacitor as a high pass for the tweeter.
Why aren't there any first order crossovers just with capacitors - one in series with the tweeter and one in parallel with the woofer?
A first order crossover has a capacitor in series with the tweeter and an inductor in series with the woofer.
A second order has the above, plus an inductor in parallel with the tweeter and a capacitor in parallel with the woofer.
There are ultra simple crossovers that are just a series capacitor as a high pass for the tweeter.
Why aren't there any first order crossovers just with capacitors - one in series with the tweeter and one in parallel with the woofer?
Looking at the simple first order crossover arrangement:
A capacitor in series with a tweeter forms a voltage divider with the lower frequencies appearing across the capacitor and the higher frequencies appearing across the tweeter.
A capacitor in parallel with a woofer would not provide the necessary frequency dependent voltage division function.
Instead we use an inductor in series with the woofer to form a voltage divider with the higher frequencies appearing across the inductor and the lower frequencies appearing across the woofer.
A capacitor in series with a tweeter forms a voltage divider with the lower frequencies appearing across the capacitor and the higher frequencies appearing across the tweeter.
A capacitor in parallel with a woofer would not provide the necessary frequency dependent voltage division function.
Instead we use an inductor in series with the woofer to form a voltage divider with the higher frequencies appearing across the inductor and the lower frequencies appearing across the woofer.
With nothing else in the woofer leg, the capacitor in parallel there will also drag the impedance down at high frequencies, which you don't typically want.
Here's a quick XSim model of what I think you're asking about. This used impedance files I had handy, but I left the frequency responses flat so the filter action is more obvious. A simulator is handy to get a feel for what things do in-circuit, and there are free and easy-to-use options out there now.
Here's a quick XSim model of what I think you're asking about. This used impedance files I had handy, but I left the frequency responses flat so the filter action is more obvious. A simulator is handy to get a feel for what things do in-circuit, and there are free and easy-to-use options out there now.
Last edited:
Capacitors are cheaper than big inductors. The Peavey SP4-XT crossover has a 33 uf cap parallel the 2 1508-8HESF woofers. I don't know if the woofers are in series or parallel, but the -8 typically means 8 ohm impedance. The RX22 tweeter has a 2.5 uf series the tweeter and a 350 uH parallel the tweeter behind the cap. 350 uH is a cheap inductor. I believe SP4-XT is rated at 4 ohms. Another SP4 is rated 4 ohms with 3.2 ohm minimum.
Last edited:
That's a short at high frequencies...you'll break many amplifiers with that circuit.
The amplifier is a voltage source, so the capacitor across the amp output terminals does nothing for the crossover.
Some amplifiers could have stability problems as well.
Some amplifiers could have stability problems as well.
Bands that buy a 113 lb 1000 w continuous rated Peavey SP4-XT do not IMHO produce much high frequency output.
Update, there are a couple of flag terminals series the woofers before the 33 uF cap that says "1.6 mH"
Theoretically, if audio amplifiers were current sources, a capacitor in parallel with the woofer would actually act as a low-pass filter. But then a capacitor in series with the tweeter as hp would make no sense.
Thanks all, I’ve learned something today.
The reason I asked, I took a gamble on these speakers for not much money, labelled as scandinavian but clearly not. I’ll probably just use them for testing with faulty amps but as they currently only have a single 2.2uF cap in series with the tweeter I thought I’d do some experimenting with a crossover to see how much difference it makes.
I have quite a few audio grade caps (hence the question), but no inductors and no local supplier so I’ll order one online.
I’ll try a simple second order and can work out the values of the components, but interested in suggestions as the what frequency to target for the cutover.
The current 2.2uF equates to somewhere up around 13kHz depending what I assume for the R value, but typically a two way is between 3-6kHz isn’t it? I was thinking of aiming around 5kHz for this one.
Of course it’s all complete guesswork, I know nothing about the drivers, suitability of the enclosure etc but I’m not expecting too much from the end product. Just want an educated guess.
The reason I asked, I took a gamble on these speakers for not much money, labelled as scandinavian but clearly not. I’ll probably just use them for testing with faulty amps but as they currently only have a single 2.2uF cap in series with the tweeter I thought I’d do some experimenting with a crossover to see how much difference it makes.
I have quite a few audio grade caps (hence the question), but no inductors and no local supplier so I’ll order one online.
I’ll try a simple second order and can work out the values of the components, but interested in suggestions as the what frequency to target for the cutover.
The current 2.2uF equates to somewhere up around 13kHz depending what I assume for the R value, but typically a two way is between 3-6kHz isn’t it? I was thinking of aiming around 5kHz for this one.
Of course it’s all complete guesswork, I know nothing about the drivers, suitability of the enclosure etc but I’m not expecting too much from the end product. Just want an educated guess.
A 2.2 uF capacitor in series with your nominal 8 ohm tweeter results in a crossover frequency of around 9,000 Hz.
Add a 0.28 mH inductor in parallel with the tweeter terminals and you can reduce the crossover frequency to around 6,000 Hz.
A 3.3 uF and 0.40 mH will get you down to around 4,000 Hz.
Last edited:
The nominal 8 ohm tweeter has a vc measurement of 5.5 but again I have no graphs to go by.
I’ve got audio grade caps at 2.2, 3.3, 4.7, 6.8 and 10uF and can use the tools to work out the right inductance.
The question is more about what frequency to target. The tweeter is labelled a DT3 and externally looks like a copy of a Goodmans DT3, so for fun I’ll base it on that spec and xover at 4kHz
I’ve got audio grade caps at 2.2, 3.3, 4.7, 6.8 and 10uF and can use the tools to work out the right inductance.
The question is more about what frequency to target. The tweeter is labelled a DT3 and externally looks like a copy of a Goodmans DT3, so for fun I’ll base it on that spec and xover at 4kHz
I would base the calculations on the nominal impedance rather than the DC resistance of the voice coil.
The similar Monacor DT75/8 dome tweeter has an impedance that varies little from its nominal 8 ohm value over its operating frequency range of 3,000 Hz to 20,000 Hz
Its crossover has to be no lower than 3,000 Hz with a second order (12 dB/octave) crossover.
4,000 Hz is a good compromise in your case.
I decided to try a proper 12db crossover from another set of speakers (2 inductors, 2x caps) and it improved the sound significantly…..I then replaced the old 2.2’s with some 4.7’s and they have more clarity but still a bit harsh.
Not worth spending any more time on these ones sadly.
Not worth spending any more time on these ones sadly.
There are some designs with just 1 cap, and it's on the tweeter.
The real issue you'd have is the impedance. Grab XSim or VituixCAD and see what happens. With a second order filter, the first component raises the impedance, the second drops it. Remove the first and you are basically creating a short to ground above/below the "crossover" point, if you could call it that.
As a result, you'll often see 2nd order components with little R in series. They actually keep the impedance of the other section from dropping too low.
I guess in theory you might be able to substitute a resistor for the first component, so a Tweeter with a resistor, then coil in parallel, but lordy the wasted power. 😀
Last edited:
For the reasons above, it's extremely difficult as the impedance load drops to something unreasonable. That said, it's not completely impossible: RC high and low-pass filters are the oldest there is (along with RL high and low pass filters). Ignoring other issues of sensitivities, on-baffle frequency / phase & impedance loads temporarily, you could go with an RC low pass, but (and it's a big 'but') you'll a/ be reducing system sensitivity & burning up power in the series resistor (how much depending on how large a resistor you need to get the system impedance to some acceptable level) and b/ be artificially raising the effective Q of the system, so you'll need to design your enclosure load with this altered Q in mind.