I don’t know if this is helpful, but it’s possible to become a bit rigid in our design approach methodologies, and this can sometimes impede the creativity that DIYing otherwise grants the builder.
You seem like a very energetic and passionate person, with ample enthusiasm for this build. I believe what you have here is a fantastic opportunity to build what you really dream of owning, but it’s also a chance to explore some improvements that can be tailored to your listening space and desires.
I know firsthand that, historically, the drivers which Wilson Audio used were custom made to have T&S electromechanical properties best suited to the designer’s goals. The drivers available to DIY markets are different, and the crossover schematic for the Alexandria XLF may be difficult to find. The speaker you build may look the same yet be functionally different, and this type of outcome has been documented before.
It might be worthwhile to entertain the hypotheticals for some physical changes, like closer tweeter-to-midrange spacing for better near-field response and lobing behaviors, changing the driver layout from WWMTM to just WWTM and using asymmetric slopes for a broader and tilted vertical pattern (this can maintain a flatter response if you stand up from the seated listening position). Other ideas may include impulse aligning the tweeter and midrange, radiused edges for reduced edge diffraction, and the removal of some structural features that are suspect for inviting resonance in the original design.
This isn’t to sow doubt in the original design, but instead highlights just a few of the many avenues your build can take.
Sharp edges, so no gradual transition from 4π to 2π radiation, various gaps in the baffle without proper damping, causing diffraction artefacts certainly and resonances possibly. A big reflecting area just underneath the lower mid doesn’t help either. I won’t comment on the MTM configuration, but it’s hard to get that right, certainly with this center-to-center spacing. So vertical lobing will be pretty present. There are pros and cons to that.
I would second this suggestion. If you like the style of the Wilson Alexandras, have a look at something like his ATS4, Illuminator 4 or Illuminator 5 projects. Troels was inspired by the Wilson speakers in these designs and, having built two of his projects, I can attest that he knows what he is doing.
Alex
But Montana member would like a clone of the XLF, not something of a lesser appearance Mr.Gravesen
is offering. 🙂
is offering. 🙂
Sharp edges cause diffraction and have nothing to do with the spread pattern when going from 2pi to 4pi radiation. Diffraction occurs at approx 1500-2000 Hz in this design and 2pi to 4pi transition occurs between 150 to approx 700 Hz.
Diffraction mainly effect the tweeters and in this design the felt surrounding the dome intend to reduce diffraction.
There are a little bit more to do regarding diffraction but Wilson has addressed the main diffraction in this design.
The whole principle with adjustable depth for each driver to the listening position is to prevent timing issues, and when timing is properly adjusted you have very little lobing that you need to worry about.
I don't know if "clone" is appropriate for the present discussion, since the OP is starting off with different drivers. The Alexandria is a 5-way design and almost impossible for a hobbyist to clone successfully
I suggested Troels' projects (as did Erik) on the basis that the "style" of the designs is based more or less on the bigger Wilson speakers, but he offers complete projects with dedicated crossovers and good measurements.
As well you know, aesthetics and quality of finish are up to the builder. Some of the builds on Troels' page are quite stunning - for instance here.
Alex
Your suggestion was reasonable, but OP desires mtm section, and that's what I meant under lesser appearance.
It's obvious that clone won't be possible, not to the last bit, outer cabinet measures perhaps may suffice to OP.
It's obvious that clone won't be possible, not to the last bit, outer cabinet measures perhaps may suffice to OP.
... a relative statement.... 😉
... to maintain a healthy ad income relation ....
I'm sure they are fine.
//
You might want to reconsider this statement. I have no intention to start a discussion about this here. But do you have access to polars and BD plots of this speaker?
I heard Wilsons in Montréal and Boston; both times they left me absolutely cold.
In contrast, at an early Burning Amp I was chatting tech with another participant when I couldn't help putting the convo on hold while I wandered to the next room in search of what was making that wonderful sound. It turned out to be a modestly sized fullrange driver powered by a not-overly-large amplifier, and this humble marriage did such a fantastic job on a singer with small combo that I had to shut up and listen.
It didn't fare so well on large-scale orchestral works, but if someone's taste ran to jazz singers or the like I wouldn't hesitate to name this as their target system (if I could remember it!).
So horses for courses, and sometimes small British monitors can be just what the doctor ordered.
As for the Wilsons... well, my opinion of them like NASA quizzing interns on how many mistakes they can find in Armageddon. Brace yourself & bring a lunch.
I do not need to reconsider anything.
I just commented on an incorrect statement from you.
Diffraction because of sharp edges is a separate fenomenon than 2pi to 4pi radiation pattern.
Diffraction is caused between driver and a nearby edge.
2pi to 4pi radiation involves the full baffle width and starts at 1/2 wavelength of the baffle width.
To keep a straight frequency response you need to compensate for the 4-6dB loss with 4pi rad instead of 2pi, and that is called baffle step compensation.
The transition from 2pi to 4pi is very little influenced if the edges are sharp or not.
And above all it occurs in a complete different frequency range compared to diffraction.
Think about WHY baffle step happens. If you can’t figure that out, read acoustics handbooks. Even Olson gives you a clue.
I do not need to think about anything you suggests.
You are wrong.
Accept it.
Here you have a typical result of the reduction in amplitude because of a 30 cm wide baffle.
340/2x0,3m = 566 Hz. There you have the point where the your 2pi to 4pi starts and then it stops at around 80-150Hz when room boundaries like rear and side walls make your 4pi radiation pattern to 2pi again.
Sharp edges on the baffle have nothing to do with radiation pattern as you stated.
If you have the driver placed center of the baffle you have your diffraction point for the edges to the driver at half the baffle width = 566x2 or 1132 Hz where you have the approx. first diffraction reflection.
You can reduce diffraction by rounding of the edges, but it have no impact on the 2pi to 4 pi transition that occurs between 80-566 Hz in this example.
There is a bit more to it than this, but there you have the basic principle of 2pi to 4pi radiation compensation or baffle step compensation, and it has very little to do with diffraction which is what you talk about in your statement.
For the record. Olson was very clever, but he didn't get everything right.
You are wrong.
Accept it.
Here you have a typical result of the reduction in amplitude because of a 30 cm wide baffle.
340/2x0,3m = 566 Hz. There you have the point where the your 2pi to 4pi starts and then it stops at around 80-150Hz when room boundaries like rear and side walls make your 4pi radiation pattern to 2pi again.
Sharp edges on the baffle have nothing to do with radiation pattern as you stated.
If you have the driver placed center of the baffle you have your diffraction point for the edges to the driver at half the baffle width = 566x2 or 1132 Hz where you have the approx. first diffraction reflection.
You can reduce diffraction by rounding of the edges, but it have no impact on the 2pi to 4 pi transition that occurs between 80-566 Hz in this example.
There is a bit more to it than this, but there you have the basic principle of 2pi to 4pi radiation compensation or baffle step compensation, and it has very little to do with diffraction which is what you talk about in your statement.
For the record. Olson was very clever, but he didn't get everything right.
Last edited:
To make it clearer...
Region marked in red is region effected by change from 2pi to 4 pi radiation pattern, and region marked in green is region is effected by diffraction.
This for a 30cm wide baffle with driver placed in centre between edges.
Region marked in red is region effected by change from 2pi to 4 pi radiation pattern, and region marked in green is region is effected by diffraction.
This for a 30cm wide baffle with driver placed in centre between edges.
Sorry, I've no intention to come over rude but: you don't present polars nor BD plots of the bespoke Wilson. Instead you want to discuss acoustics. Then please: why does baffle step occur? And: I didn't start about diffraction at sharp edges of enclosures. You did.
We are not dicussing XLF here but general principles, and you try to lecture me as I know less than you when the opposite is true.
You have not got the full picture clear for you when you relate sharp edges to 2pi to 4pi transition in spred pattern.
They are not closely related.
Eigther you take notice and learn or you don't.
Thats up to you.
Thanks for the discussion...
You have not got the full picture clear for you when you relate sharp edges to 2pi to 4pi transition in spred pattern.
They are not closely related.
Eigther you take notice and learn or you don't.
Thats up to you.
Thanks for the discussion...
OK. I'll try. Baffle step occurs because on a baffle, at any point, the sound radiated from the transducer, is reflected and thus sums with the incoming sound. In an ideal situation that can lead to +6dB on a far point perpendicular on that baffle (easy form of the wave equation). The summing of course only adds up to 6dB if both sound sources are in phase (that is why the ripples occur above the baffle step). And for larger wavelengths the size of the baffle isn't big enough to contribute (the reflected sound of course is weaker and you 'need' a certain area).
Now if a baffle has sharp edges, the character of this summing phenomenon changes very fast, leading in an abrupt change of the mentioned summing at the said far point, with a sudden drop of the combined response (SPL). And thus creates an abrupt baffle step, hard to compensate and leading to a challenge in reaching a gradual sound power response.
I think this is short enough. Of course you could add comb filtering due to the 'diffraction' at the sharp edge into the equation (in fact I addressed it), but I leave it here for now.
Now if a baffle has sharp edges, the character of this summing phenomenon changes very fast, leading in an abrupt change of the mentioned summing at the said far point, with a sudden drop of the combined response (SPL). And thus creates an abrupt baffle step, hard to compensate and leading to a challenge in reaching a gradual sound power response.
I think this is short enough. Of course you could add comb filtering due to the 'diffraction' at the sharp edge into the equation (in fact I addressed it), but I leave it here for now.
You can not make a clone of Alexandria, because you have different woofers. Also, you can't buy the same midrange and tweeter as in Alexandria because they are specially made for Wilson Audio. So, acoustical clone of Alexandria is not possible.
But you can make look-alike (visually) clone of Alexandria. Making crossover for your "clone" is job for an expert (or a very good DIY enthusiast with measurement equipment and crossover simulation software) - hire one.