• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Linkwitz LX521.4 Review (and measurements!)

Do you have any links to the work done by Laub, Boers and Finke?
Hi Burning Sounds

Of course I have. Here you have them.

Gerrit Boers' absolutely full constant directivity 20-20 KHz design and implementation, including the assembling of Hypex power amps, Hypex DLCP crossover and custom designed planar tweeter. Astounding work in my opinion. The resulting polar measurement is out of this planet. Nude drivers is the way to go, I also think. Use Google Chrome for a very exact translation from Netherlands to English. The real show begins around page 6 or 7, if my memory serves well.


Charlie Laub has also constructed a true CD transducer 20-20 KHz. In my opinion, he implements a great and quite innovative way to use two inexpensive 1" nude dome tweeters with waveguides to achieve constant directivity performance up to the highest frequencies. This is Achilles heel in LX521, because I suppose SL was forced to set up a very high crosspoint to the Seas tweeters (above dipolar peak and above diffraction zone) to get some control on the directivity of the 27TFFNCs. This was due to the fact that the Seas tweeters showed a bad pattern in that zone from 2 to 6 KHz, so a dead end. As a consequence of this fact, I think SL was forced again to add much more passband to the MU10RB high mid driver than what he was willing. It should have been crossed up at a significative lower frequency, at about 3 to 4 KHz, but his selected tweeters didn't allowed him. Could it be a better transducer without these impositions? Maybe yes, maybe not. The thread itself is a fantastic work by Charlie Laub, with quite interesting opinions by other informed writers.


And finally, here is the thread by Rudolf Finke. He has also married two opposed and nude smaller dome tweeters but without waveguides. Despite getting a quite good horizontal response off-axis, the vertical directivity resulted a disaster and he abandoned the idea.


Best regards
 
Last edited:
Do you have any links to the work done by Laub, Boers and Finke?
Laub here... If you (or any forum member) wants a copy of my manuscript from 2022 about "minimally baffled" speakers just drop me a message and I will send it to you. It's an evolution of the 20-20k dipole whitepaper that is linked above by @JeyB.
 
Yes, I do pull them out for listening. I know Sigfried recommended at least 3 feet from the front wall, but realize many 521 owners bring them much further into the room, which doesn't have a very appealing WAF lol, which is very much at play in my situation. I did pull them into the room for the measurements that I made. That being said, I don't think one should have such wild frequency/SPL response fluctuations.

I think each speaker is ~ 70 lbs. How do you move them to & fro, on dolly’s?
 
Hi Burning Sounds

Of course I have. Here you have them.

Gerrit Boers' absolutely full constant directivity 20-20 KHz design and implementation, including the assembling of Hypex power amps, Hypex DLCP crossover and custom designed planar tweeter. Astounding work in my opinion. The resulting polar measurement is out of this planet. Nude drivers is the way to go, I also think. Use Google Chrome for a very exact translation from Netherlands to English. The real show begins around page 6 or 7, if my memory serves well.


Charlie Laub has also constructed a true CD transducer 20-20 KHz. In my opinion, he implements a great and quite innovative way to use two inexpensive 1" nude dome tweeters with waveguides to achieve constant directivity performance up to the highest frequencies. This is Achilles heel in LX521, because I suppose SL was forced to set up a very high crosspoint to the Seas tweeters (above dipolar peak and above diffraction zone) to get some control on the directivity of the 27TFFNCs. This was due to the fact that the Seas tweeters showed a bad pattern in that zone from 2 to 6 KHz, so a dead end. As a consequence of this fact, I think SL was forced again to add much more passband to the MU10RB high mid driver than what he was willing. It should have been crossed up at a significative lower frequency, at about 3 to 4 KHz, but his selected tweeters didn't allowed him. Could it be a better transducer without these impositions? Maybe yes, maybe not. The thread itself is a fantastic work by Charlie Laub, with quite interesting opinions by other informed writers.


And finally, here is the thread by Rudolf Finke. He has also married two opposed and nude smaller dome tweeters but without waveguides. Despite getting a quite good horizontal response off-axis, the vertical directivity resulted a disaster and he abandoned the idea.


Best regards

Hi Jey
I just saw your description of Gerrit Boers' speakers with “constant” directivity - apparently better than Linkwitz final design
awaiting confirmation but assuming you’re not making it up lol - Much appreciated
Went to the thread in the link …11 pages - in Dutch … now hoping someone able to search the thread than me will direct me to specs & especially the measurements

I just saw your equipment list in your footer. And an octagonal room. Very impressive

Anyhow … what is a monkey box?

Cheers
 
Laub here... If you (or any forum member) wants a copy of my manuscript from 2022 about "minimally baffled" speakers just drop me a message and I will send it to you. It's an evolution of the 20-20k dipole whitepaper that is linked above by @JeyB.

Hi Charlie

Thank you for offering. Though I’m much more interested in Jey’s
“constructed a true CD transducer 20-20 KHz” which I initially interpreted as a single driver. (emoticon Cannot be)

Is the transducer described in its final form, preferably with measurements.. somewhere in the 28 page thread Jey linked to, or elsewhere?

Cheers
Richard
 
I see the source of your confusion... Using "transducer" in the sentence above is not correct and it should read: "Loudspeaker"
 
Is the transducer described in its final form, preferably with measurements.. somewhere in the 28 page thread Jey linked to, or elsewhere?

So to rephrase the question correctly …

Is there somewhere within the 28 page thread where there are measurements of the main attributes of the completed project, ie directivity, distortion, frequency response?
 
Hi Jey
I just saw your description of Gerrit Boers' speakers with “constant” directivity - apparently better than Linkwitz final design
awaiting confirmation but assuming you’re not making it up lol - Much appreciated
Went to the thread in the link …11 pages - in Dutch … now hoping someone able to search the thread than me will direct me to specs & especially the measurements

I just saw your equipment list in your footer. And an octagonal room. Very impressive

Anyhow … what is a monkey box?

Cheers
Hi Richard

As you know, a transducer is a device that converts variations in a physical quantity, such as pressure or brightness, into an electrical signal, or vice versa. I meant transducer as a set of various electroacoustic drivers, meaning the whole set, ie the loudspeaker. Loudspeaker is a more adequate word and easier to understand, like Charlie said. Sorry

On the other hand, my room is not octagonal. It's a orthogonal rectangular prism, a rectangular parallelepiped.

A monkey box, monkey coffin is a quick description of a boxed loudspeaker. I've first time heard this term employed by Peter Aczel from the Audio Critic fanzine. It doesn't mean in any way there is a dead monkey inside ready to be buried :)

I'm not constructing Gerrit Boers' loudspeakers. I'm trying to design, with my limited knowledges, another true constant directivity prototype, mainly with nude drivers from 120 Hz up, just to discover what difference is perceived psychoacoustically compared to my own LX521.

I feel the need to make it. But maybe a full range constant directivity system (with near perfect figure of eight radiation pattern) doesn't implies a better subjective sound experience. It's possible that it's enough with some clever directivity control. Who knows...

But my proto won't be a single driver loudspeaker. It will be 4 way, 7 drivers. So I think I've induced some kind of confussion relative to my use of the term transducer.

I've found really great info and measurements in Charlie's 20-20 KHz thread from 2018 who is helping me to take decissions in my project. In 2022 he has made another more advanced louspeaker.

Best regards
 
Last edited:
Well, we can all constabtly learn things, I do every day and it is one of the joys of living!

My experience with my Lx521 is they upgraded the wrong driver. I have not found any issue with the lower midrange driver which now has an expensive replacement. And while the upper midrange driver measures perfectly and I can tune any which way with my pc based crossover/dsp - i just find it sounds "harsh" compared to other speakers I have - Genelec, Neuman etc. I just dont think its a nice sounding driver no matter what I do. Considering how little the tweetets actually do, I dont think they need replacing, but do need shelving down. The bass is phenomenal!
While I don't find the UM driver to sound harsh it can sound slightly edgy on some recordings, but is perfectly fine on others, so I'm not completely convinced it is the driver that is at issue here.

I too, use a PC based x-over/eq (JRiver), but also have the Linkwitz ASPv2. I prefer the PC based x-over as I find it subjectively more realistic (whatever that means - it's hard to describe the difference without getting into audiophile speak :D). From version 29 JRiver could do all pass filters so I have been able to copy the transfer function of the ASP. The transfer functions are not identical, but the difference is very small and I've been able to match exactly with a bit of judicious filtering.

Do you use SL's original configuration to shelve down the upper range or his later one where he removed this?
 
The newer ASPv.2 is essentially a reversion back to the original crossover design. It uses a state-variable topology vice the Sallen-Key configuration, plus a few notches/peaks to fine-tune the drivers.
The LX521 system has not changed in general concept since day one.

Also, the need for all-pass filters is removed if using a DSP setup since you can program pure delays to achieve the phase-coherency.
 
The newer ASPv.2 is essentially a reversion back to the original crossover design. It uses a state-variable topology vice the Sallen-Key configuration, plus a few notches/peaks to fine-tune the drivers.
The LX521 system has not changed in general concept since day one.

Also, the need for all-pass filters is removed if using a DSP setup since you can program pure delays to achieve the phase-coherency.
Yes, I know what ASPv2 does, that's why I asked you in an earlier post if that was what you were referring to.

I agree that the LX521 system has not changed in general concept, but there have been what appear on the surface to be a few minor changes that do affect how it sounds IMO.
 
another true constant directivity prototype, mainly with nude drivers from 120 Hz up, just to discover what difference is perceived psychoacoustically compared to my own LX521.

I totally understand your motivation, and your goal

re nude drivers
You probably know that the 521’s baffle shape was arrived at by experimentation. angularity aside, relative to his previous speaker the Orion in terms of baffle size I’d say, ball-park it’s “70% nude”. It’s a shame he didn’t share the shapes he tried and their results

“won't be a single driver loudspeaker”
Of course. Unless you’re aware of a driver that is from an alternative reality lol … mission impossible

“ It will be 4 way, 7 drivers”’

Sounds about right. At diyaudio, poster andyr made a nude baffle dipole. a 3-way
From memory
the frame is metal
7 drivers: 1 tweeter, four midwoofers and two bass drivers
the four midwoofers are SB Acoustics, arranged in a square pattern abutting each other.
I don’t know why he didn’t put the tweeter in the centre, but it’s in the usual position above them

Probably worth a look. Let me know if you have trouble finding it
It got very high praise from a very good amplifier designer

As of now I know nothing about Charlie's. It could be even better?

Cheers
Richard
 
“Charlie's 20-20 KHz …
In 2022 he has made another more advanced loudspeaker”

Is that somewhere, later in the 2019 thread?
Or a different thread?

the 2019 thread has nearly 600 posts - I’d be grateful if someone could point me to measurements of the final version

Cheers
Richard
 
For those interested in this challenge, please visit also the websites
www.dipolplus.de by Rudolf Finke
https://musicanddesign.speakerdesign.net/products.html by John Kreskovsky

I have been happy with CD dipole from 200Hz up. Monopole bass is so easy and crossed to dipole it has cardioid response around xo. I use planar transducers like Gerrit. Vertical pattern is always compromised in non-coaxial multiways.


ainogneo v11 in 20ms 16 norm Directivity(hor) lines.png ainogneo83 2x4HD v35 in vertical 0-30 4ms 13 lines.png ainogneo83 2x4 conf36 MMM 500ms 112.jpg
 
Last edited:
I totally understand your motivation, and your goal

re nude drivers
You probably know that the 521’s baffle shape was arrived at by experimentation. angularity aside, relative to his previous speaker the Orion in terms of baffle size I’d say, ball-park it’s “70% nude”. It’s a shame he didn’t share the shapes he tried and their results

“won't be a single driver loudspeaker”
Of course. Unless you’re aware of a driver that is from an alternative reality lol … mission impossible

“ It will be 4 way, 7 drivers”’

Sounds about right. At diyaudio, poster andyr made a nude baffle dipole. a 3-way
From memory
the frame is metal
7 drivers: 1 tweeter, four midwoofers and two bass drivers
the four midwoofers are SB Acoustics, arranged in a square pattern abutting each other.
I don’t know why he didn’t put the tweeter in the centre, but it’s in the usual position above them

Probably worth a look. Let me know if you have trouble finding it
It got very high praise from a very good amplifier designer

As of now I know nothing about Charlie's. It could be even better?

Cheers
Richard
Hi Richard

You are right that SL tried many different baffles for the LX521, until he found the one that best suited his choice of drivers and his purpose of controlling directivity in order to evenly illuminate the room. Just a guess, but I think that if SL could create a new loudspeaker, it would probably lack the baffle, except for an open, folded structure for the bass. From the Phoenix all the way to the LX521, he progressively reduced the deflector with each new project that saw the light. Finally in the LXmini, his latest design, he finally eliminated it completely.

I don't know of any single full range controller that serves a CD application either. Physics plays against that purpose.

In post 43, Charlie said he could send, to those who are interested, his 2022 manuscript on the dipole with minimal baffles. Send him a PM please. That dipole is not in the 2019 thread or any other, I believe. On whether it's superior to his 2018 prototype, I can't know since I haven't heard either of them. But what I am sure of is that it makes very interesting reading and deals very well with various technical aspects of importance.

My idea is that if one system allows me to enjoy music more than another, then I consider it superior. But this is very subjective territory. In the end, it depends on the personal preferences of each of us.

If you can indicate the link to Andyr's design, I would appreciate it. It looks interesting that prototype.

Best regards
 
Yes, I know what ASPv2 does, that's why I asked you in an earlier post if that was what you were referring to.

I agree that the LX521 system has not changed in general concept, but there have been what appear on the surface to be a few minor changes that do affect how it sounds IMO.
Yes, all of the changes/revisions have affected the sound.

The change to DSP in the .4 version was a significant one since it altered the out-of-band roll-off rates for the drivers. But, there were other changes made as well, so it was difficult to associate this upper-midrange issue to the roll-off changes.
In hindsight, it would have been preferable to incorporate the "cascaded" configuration into the DSP setup as well.
 
SL created the LX521 baffle shape by repeated trial and error. He started with some shape and then cut away here and there while remeasuring. What you have today is basically as little baffle as possible, as he found that to give the best pattern while still being able to have enough left into which he could mount the drivers. The baffle shape and response pattern was far from the only aspect of the loudspeaker that was important to him. He valued low energy storage and low distortion, especially IMD, and he did extensive testing to find the drivers that were used in the original LX521 speaker. This is why, for example, he used two dome tweeters - they have very good performance even if not a great dipole pattern. For my own projects I choose pattern over distortion performance, so I use AMT tweeters that are dipole or close to it throughout their range. I outline the design choices and design procedure in my 2002 paper and provide some technical background about why a "nude" driver is best for midrange and tweeter.
 
Back
Top Bottom