Don't know why you post links to Art Ludwig's pages but here is a link to my old site. (click Impatient)
Please take note of the section on TP crossover (Transient Perfect). The TP 2-way is the crossover Art refers to on his page,
Also here is a link to the FRC where you can look at a number of Excell spreadsheets I developed based on extension of th work of Thiele, Vanderkooy and Lipshitz and the old B&O filler driver approach by Bækgaard.
Since you seem to be a student of the past work I assume you would be familiar with all these.
Quite true, and you could insert this comment into many of the threads around here.
I'm still getting my head around the design of the higher order series crossovers. What made this hard was the lack of good examples, but I have this good Troels Vifa XT25 design to study:
Vifa PL18WO-
And Tony Gee at Homemade Humble HiFi has done endless designs. The Classic here is a good one to look at:
Humble Homemade Hifi
What you seem to do is give the input LC circuit ca. 5 ohm resistive damping to counter the fact it is getting less direct damping from the LC nearer the drive units. Otherwise it's gonna ring horribly. Still haven't done the exact transfer function, because I've yet to finalise a good higher order example circuit.
Oh, worth mentioning that the optional shunt LCR circuit at the input of Tony Gee's Classic design is just to tame the impedance peak for valve amps. Voltage amp users won't need it.
Vifa PL18WO-
And Tony Gee at Homemade Humble HiFi has done endless designs. The Classic here is a good one to look at:
Humble Homemade Hifi
What you seem to do is give the input LC circuit ca. 5 ohm resistive damping to counter the fact it is getting less direct damping from the LC nearer the drive units. Otherwise it's gonna ring horribly. Still haven't done the exact transfer function, because I've yet to finalise a good higher order example circuit.
Oh, worth mentioning that the optional shunt LCR circuit at the input of Tony Gee's Classic design is just to tame the impedance peak for valve amps. Voltage amp users won't need it.
Last edited:
I'm not sure WHAT it's gonna do, speakerdoctor...😀
An externally hosted image should be here but it was not working when we last tested it.
A quick estimate on the 0.27mh and 15uF near the 8 ohm drivers says that part of the circuit (Z=4.23R) is damped with a Q of 0.5 and resonance around 2.5kHz. So I'd reckon the 0.47mH, 8uF, 4.5R input (Z=7.6R plus resistance) might just be right too around 0.5 Q. But I won't be risking my amp till I understand it all. 😉
Attachments
Last edited:
It's become evident that there are at least two accomplished DIY speaker builders (Gee and Graveson) out there in the WWW who, on occasion, opt for a series xo.
It may be worth asking them why since they do build a LOT of speakers and listen a LOT to how they sound.
At least I would value their subjective opinions. Who knows, one or both could surprise us with their rationale.
It may be worth asking them why since they do build a LOT of speakers and listen a LOT to how they sound.
At least I would value their subjective opinions. Who knows, one or both could surprise us with their rationale.
Troels always amazes me by getting the capacitor values wrong where he uses a 6 ohm input attenuating resistor on tweeter 3rd order butterworths. They are meant to be the same, aren't they? But here's his series Xover for our pleasure:
An externally hosted image should be here but it was not working when we last tested it.
He's going the opposite way to Tony Gee on inductor values too. But then no pure resistance on the input. 😕
HI,
R1's main job appears to be to attenuate the tweeter, though
being placed where it is suggests it might damp something.
I can't really see how L2 and C3 interact as described.
Given the voltage split for L1 And C1+R1, after than
you got what looks like a second order parallel filter
+ zobels on each driver, but joined together at the
split point, which in parallel would be common return.
Its certainly not intuitive if your used to parallel x/o's.
rgds, sreten.
R1's main job appears to be to attenuate the tweeter, though
being placed where it is suggests it might damp something.
I can't really see how L2 and C3 interact as described.
Given the voltage split for L1 And C1+R1, after than
you got what looks like a second order parallel filter
+ zobels on each driver, but joined together at the
split point, which in parallel would be common return.
Its certainly not intuitive if your used to parallel x/o's.
rgds, sreten.
Last edited:
This is INTERESTING! 😀
People here will know my admiration for the musical qualities of the deceptively ordinary looking paper-coned Acoustic Research loudspeakers. Lo and behold in 1981, AR were using a series crossover in their AR-48s model betwen bass and midrange with a quite extraordinarily big inductor on the bass or midrange, depending how you like to look at it...😎
People here will know my admiration for the musical qualities of the deceptively ordinary looking paper-coned Acoustic Research loudspeakers. Lo and behold in 1981, AR were using a series crossover in their AR-48s model betwen bass and midrange with a quite extraordinarily big inductor on the bass or midrange, depending how you like to look at it...😎
Attachments
You can build a LOT of speakers even without actually hearing them 😀 I mean, you can build a new design in a day, and the quality is in accordance.
The PL18-XT25 is no different with other series design I have heard. Not my cup of tea.
No, I was talking about the other one. Lately I read a thread where someone with much less experience than him was discarding his W15-XT25 design for a ARSXO design which was designed only by ears. I laughed out loud. That's an irony. I think mediocre designs are acceptable when there is time constraints in the design (per speaker).
I must admit that Jay's comments there didn't warm my heart. Troels Gravesen is the ultimate speaker enthusiast. He just likes trying stuff. If a speaker doesn't work for you, well, maybe it's just not optimised for your amp. Because amps and output impedance make a huge difference to the whole design process. 🙂
I was struck by this splendid WLM Stella design which jacozz kindly brought to my attention.
This is a very pricey and beautifully optimised design employing a very basic crossover, tweeter attenuator and rather basic Visaton drivers, a £20 5" paper bass and a and a £10 3" paper tweeter.
It's actually quite AR (Acoustic Research) in concept, bar the reflex. What appeals to me is adding a £10 4" midrange with a series crossover and even improving it. The trick with series crossovers is the drive units should be rather similar. We're just filling the horrible gap between the woofer and the tweeter here in two way designs. 😎
I was struck by this splendid WLM Stella design which jacozz kindly brought to my attention.
An externally hosted image should be here but it was not working when we last tested it.
This is a very pricey and beautifully optimised design employing a very basic crossover, tweeter attenuator and rather basic Visaton drivers, a £20 5" paper bass and a and a £10 3" paper tweeter.
It's actually quite AR (Acoustic Research) in concept, bar the reflex. What appeals to me is adding a £10 4" midrange with a series crossover and even improving it. The trick with series crossovers is the drive units should be rather similar. We're just filling the horrible gap between the woofer and the tweeter here in two way designs. 😎
A fellow audio society colleague of mine writes reviews for 6moons high end web site. He has a pair of WLM La Scala's. They are also 2-way with simple 6 db/oct xover. Looking at pics of the xover, it appears it's st'd parallel network.
He did not write the review at the link below.
6moons audio reviews: WLM La Scala
He did not write the review at the link below.
6moons audio reviews: WLM La Scala
I had a look at that crossover, and I don't know WHAT to make of it! 😕
Looks a bit like a series to me the way the coil and cap are connected across the terminals with a (2.2R resistance?). But it makes no sense at many levels, especially since we can't see how the drivers are wired up. I think WLM just lashed something up there to create a bit of FUD.
Can't blame them. They gotta keep their secret sauce, er, secret...😀
Looks a bit like a series to me the way the coil and cap are connected across the terminals with a (2.2R resistance?). But it makes no sense at many levels, especially since we can't see how the drivers are wired up. I think WLM just lashed something up there to create a bit of FUD.
An externally hosted image should be here but it was not working when we last tested it.
Can't blame them. They gotta keep their secret sauce, er, secret...😀
Download the imagehifi review in German at this link
WLM Hauptseite
It's got a better xo pic with driver wires attached. Tell me what you think.
WLM Hauptseite
It's got a better xo pic with driver wires attached. Tell me what you think.
Those photos are posted by the MD of WLM. I expect he is aware of secrecy issues. 🙂
For sure if that was really a PUNY 1W 2R2 resistor between the speaker terminals, it would fry in about 5 seconds of hard driving whether with a coil and cap or not. I'd expect the 5-10W wirewound is really the key player and the tweeter attenuator is just doing the normal AR job. 😀
I find the PAC tweeter idea more interesting in WLM's more high-end designs. It seems to borrow from the AR-2a midrange offsets:
Oh God, WLM must hate me! 😉
For sure if that was really a PUNY 1W 2R2 resistor between the speaker terminals, it would fry in about 5 seconds of hard driving whether with a coil and cap or not. I'd expect the 5-10W wirewound is really the key player and the tweeter attenuator is just doing the normal AR job. 😀
I find the PAC tweeter idea more interesting in WLM's more high-end designs. It seems to borrow from the AR-2a midrange offsets:
An externally hosted image should be here but it was not working when we last tested it.
Oh God, WLM must hate me! 😉
Oh, this one! 😎
That's a bit different, isn't it. Will mull that over...
An externally hosted image should be here but it was not working when we last tested it.
That's a bit different, isn't it. Will mull that over...
- Status
- Not open for further replies.