This thread is a follow up to http://www.diyaudio.com/forums/multi-way/263359-ugly-lx-mini-clone.html
In this thread I'll take some of the ideas from the other thread, but improve upon them sonically and cosmetically.
In this thread I'll take some of the ideas from the other thread, but improve upon them sonically and cosmetically.
Here's a pic of the Linkwitz LX Minis.
Here's a pic of the Gradient Helsinkis
Here's my idea:
The LX Mini has a tweeter that's cardioid, and a monopole midrange.
The Gradient Helsinki has a constant directivity tweeter, a cardioid midrange, and a dipole bass unit.
My idea is to take some of the goodness of the LX Mini, but try to improve upon it cosmetically and sonically. To improve it cosmetically I'm going to use a commercial loudspeaker for the midrange and the treble. By using a commercial unit I'm assured that the fit and finish will be superior to what I can do. To improve on the LX sonically I am going to smooth the polar response with the use of a waveguide, and extend the frequency response by going with a 3 way. In many ways this is similar to the Helsinki, which also uses waveguides and is a three way.
I've heard the Helsinki and they're fantastic.
For my midrange and treble, I'm using the JBL Control Now. Here's some cool features of the speaker:
1) it has a 3/4 dome loaded in a constant directivity waveguide
2) it has dual midranges with underhung neodymium motors for low distortion
3) The MTM configuration narrows vertical directivity, minimizing the 'floor bounce' while maintaining wide horizontal dispersion
These retailed for close to $300, but I bought a pile of 'em when Costco was selling them for $50 each.
On paper, the Control Now has the potential to outshine the drivers in the LX Mini. The horizontal polars of the LX Mini are nothing to write home about, due to diffraction off of the surfaces of the two drivers, and the peculiar mounting scheme:
^^ LXmini polars ^^
The Control Now has a waveguide loaded tweeter that offers controlled directivity, extension to 20khz, and high efficiency. The Control Now has more low frequency displacement than the LX Mini.
There's a lot to like.
Unfortunately, I've always found that my Unity horns are more articulate than my Control Now, and my Summas image better.
So I wanted to do some investigation of the crossover, because I thought it might be the culprit. The JBL drivers are good, especially at this price. I don't think the lack of soundstage and articulation can be blamed on the drivers.
Here's a polar measurement of the Control Now.
There's a lot of issues with this plot:
1) The midbass is about five dB hotter than everything else
2) The midrange is recessed
3) The polars are a bit bizarre; at high frequency the speaker is loudest on-axis, but in the two octaves from 500hz to 2khz the speaker is louder OFF axis than ON axis.
Quite a strange set of measurements. It helps explain why the speaker "looks good on paper" but doesn't measure particularly well.
I've had these speakers for ages, I've measured them before, and for the longest time I thought JBL intentionally voiced them so that the treble and the bass is "hot." The ol' "JBL sound" that they were infamous for in the 70s and 80s.
But then I sat down and thought about it for a minute, and it occurred to me that the problem may be that the midrange drivers are not equidistant. IE, when you set up the Control Now in the "obvious" way, the midrange at the top is about 15cm behind the midrange at the bottom. That gap introduces a delay of about half a millisecond. And a delay like that can wreak havoc with imaging cues, it can make it impossible to get the drivers in-phase at the crossover point, it can dramatically alter the frequency response. Particularly in the midrange, where it's so important.
To combat those issues, I ran another set of measurements, this time with the two midrange drivers perfectly aligned on the vertical axis.
Lo and behold, that basically fixes EVERYTHING. It improves the polar response, it flattens the frequency response, it reduces the comb filtering that's cause by using multiple midranges. It's a dramatic change. Once you tilt the speaker, it measures +/- 4dB from 90hz to 18khz.
Not too shabby for a $50 speaker from Costco!
^^ LXmini polars ^^
The Control Now has a waveguide loaded tweeter that offers controlled directivity, extension to 20khz, and high efficiency. The Control Now has more low frequency displacement than the LX Mini.
There's a lot to like.
Unfortunately, I've always found that my Unity horns are more articulate than my Control Now, and my Summas image better.
So I wanted to do some investigation of the crossover, because I thought it might be the culprit. The JBL drivers are good, especially at this price. I don't think the lack of soundstage and articulation can be blamed on the drivers.
Here's a polar measurement of the Control Now.
There's a lot of issues with this plot:
1) The midbass is about five dB hotter than everything else
2) The midrange is recessed
3) The polars are a bit bizarre; at high frequency the speaker is loudest on-axis, but in the two octaves from 500hz to 2khz the speaker is louder OFF axis than ON axis.
Quite a strange set of measurements. It helps explain why the speaker "looks good on paper" but doesn't measure particularly well.
I've had these speakers for ages, I've measured them before, and for the longest time I thought JBL intentionally voiced them so that the treble and the bass is "hot." The ol' "JBL sound" that they were infamous for in the 70s and 80s.
But then I sat down and thought about it for a minute, and it occurred to me that the problem may be that the midrange drivers are not equidistant. IE, when you set up the Control Now in the "obvious" way, the midrange at the top is about 15cm behind the midrange at the bottom. That gap introduces a delay of about half a millisecond. And a delay like that can wreak havoc with imaging cues, it can make it impossible to get the drivers in-phase at the crossover point, it can dramatically alter the frequency response. Particularly in the midrange, where it's so important.
To combat those issues, I ran another set of measurements, this time with the two midrange drivers perfectly aligned on the vertical axis.
Lo and behold, that basically fixes EVERYTHING. It improves the polar response, it flattens the frequency response, it reduces the comb filtering that's cause by using multiple midranges. It's a dramatic change. Once you tilt the speaker, it measures +/- 4dB from 90hz to 18khz.
Not too shabby for a $50 speaker from Costco!
Last edited:
If anyone wants to play along at home, the JBL Control Now is easily found on eBay, and Dayton Audio just cloned it in the guise of their "Quadrant" line.
The same as (most) all things PB does.
Keep on, keepin on John.
Last year did a similar and noticed that there wasn't enough direct to the reflecting soundfield in a pluto-like speaker. Like you, goofing around just to see what there is to see / possible improvements to be made. Insight into that has given me more than a few ideas for a future dsp based active project.
It's going to be the same setup:
A up-firing mono pole woofer with a cardioid satellite perched on top.
The difference is that the satellite will be a two-way, not a full range.
The same as (most) all things PB does.
Keep on, keepin on John.
Last year did a similar and noticed that there wasn't enough direct to the reflecting soundfield in a pluto-like speaker. Like you, goofing around just to see what there is to see / possible improvements to be made. Insight into that has given me more than a few ideas for a future dsp based active project.
"Possible improvements" based on a fundamental misunderstanding of the goals of the original design.
Nothing wrong with having his own experiments but the title is in poor taste, this is not a homage; it is merely expressing his design prejudices. An exercise in proactive trolling.
SL wants the scattered rear radiation. LX mini is meant to be part open baffle. Using a forward radiating wave guide makes this a fundamentally different concept.
Not using one OB drive unit precisely positioned and suspended over a monopole makes this a completely different concept.
Titles like this should go in another popular thread on this site: "what really grinds your gears".
The enclosure is about to become swiss cheese.
This is going to look a lot like an LX mini soon.
If the mods want to change the title, that's fine
This is going to look a lot like an LX mini soon.
If the mods want to change the title, that's fine
Possible improvements have todo with psychoacoustical changes in our hearing. No misunderstanding of any concept SL is marketing. Having built similar back in the '80's what SL's design goals are is nothing more than the basics of such a design and it's built in limitations. Sigfried does this for K.I.S.S. factor and cost.
I'm not arguing they are poor examples of what can be achieved, quite the contrary, but is not the be all / end all. Here we in DIY can design such that exceed what another has done. SL wasn't the first with such a design, nor was I 30+ years ago. 🙂
I'm not arguing they are poor examples of what can be achieved, quite the contrary, but is not the be all / end all. Here we in DIY can design such that exceed what another has done. SL wasn't the first with such a design, nor was I 30+ years ago. 🙂
Here's a pic of my JBL Control Now after being turned into swiss cheese. I cut twenty 1" holes into the enclosure. I would have added more but the top and bottom panels on the Control Now are made of aluminum. Due to that, my hole bit balked.
Besides adding the holes, I also added fiberglass insulation to attenuate the rear wave.
An externally hosted image should be here but it was not working when we last tested it.
Kimmo Saunisto was definitely an inspiration. That's one of his speakers from his site, pictured above.
Here's some pics of the speakers in action :
The speaker in the center is a Kef coax. The speakers on the side are a diy design I'm working on, documented in a thread called "LX Mini Homage". I have some unfinished Synergy horns and some Gedlee Summas in the garage. They're too big for this room. (I live in a tiny home.)
The speaker in the center is a Kef coax. The speakers on the side are a diy design I'm working on, documented in a thread called "LX Mini Homage". I have some unfinished Synergy horns and some Gedlee Summas in the garage. They're too big for this room. (I live in a tiny home.)
I have full polars, both before and after I gutted the enclosure.
I need to post them.
Before the enclosure was perforated, it had an F3 of below 100hz.
Once I turned it into swiss cheese, the F3 went up to around 400hz. It starts to look like a cardioid around 1khz.
So it basically looks like this:
1) In the 3.5 octaves from 2khz to 20khz it's basically constant directivity, due to the waveguide
2) In the octave from 1-2khz it's omnipolar. This is because the radiators are much smaller than the wavelength, and because the radiation from the back of the cone is attenuated by the fiberglass inside the enclosure
3) In the octave from 500hz to 1khz it's cardioid, because the radiation from the back of the cone is escaping the enclosure via the holes
I gotta admit the beamwidth looks a little funky. The waveguide does a much better job of controlling directivity than the cardioid does. But it sounds EXCEPTIONAL. Literally one of the most successful projects I've done in a long time.
My hunch is that the good sound is due to a couple of things:
1) Narrow enclosures image really nice. My Summas image better than any horn I've ever heard, but these Control Nows image better than the Summas now that they're a cardioid. I think the narrow baffle width has something to do with that.
2) I think that early reflections really cause a LOT of issues. For instance, I heard some Wilson Audio Sophias a week ago, and the Wilsons had real depth and width in the soundstage. But the owner of the Wilsons had pulled them way, WAY far away from the walls. The demo room was about the size of my entire house. I just don't have the space to arrange my speakers like that, so cardioid is an excellent alternative. (And I believe that's why Linkwitz opted for cardioid also.)
Here's the overall polar response of the three-way system. Unfortunately I made the mistake of using single-octave smoothing, accidentally. So the resolution isn't as good as it should be.
Here's the polar response of the midrange and tweeter, with 6th octave smoothing
Here's the unequalized response of the midbass and the mains. Note that the midbass is much more efficient, due to horn loading.
Here's the equalized response of the midbass and mains
Some random observations:
1) The waveguide does it's job well, but it only works down to 2700hz
2) The cardioid response occurs in the 1.5 octaves from 800hz to 2700hz
3) Despite the stuffing, the dipole peak is quite apparent at 400hz
In dipole designs, the size dependant phenomenons such as peak and roll off govern the crossover point. At any frequency absorbing works only if it has significant width relative to the wavelenght. At 400 Hz, you need roughly 8 inches of absorbing material in order to make the wave admit there is any such measure present.
May I add something to your system description. The midrange of the LX Mini is not monopole. It is 1 phi omnidirectional ratiator in the horizontal pane down to the frequency with wavelenght long enough to consider the floor as a near boundary, lower than that it is 1/2 phi radiator, but sustains omnidirectional action gradually widening the radiation angle to more than 180 degrees towards the bass.
SL made the speakers tall for a reason.
Thank you for the thread, I learn from it as it is interesting and the information is served in an easy to comprehend manner!
May I add something to your system description. The midrange of the LX Mini is not monopole. It is 1 phi omnidirectional ratiator in the horizontal pane down to the frequency with wavelenght long enough to consider the floor as a near boundary, lower than that it is 1/2 phi radiator, but sustains omnidirectional action gradually widening the radiation angle to more than 180 degrees towards the bass.
SL made the speakers tall for a reason.
Thank you for the thread, I learn from it as it is interesting and the information is served in an easy to comprehend manner!
Last edited:
Interesting!
Here's the overall polar response of the three-way system. Unfortunately I made the mistake of using single-octave smoothing, accidentally. So the resolution isn't as good as it should be.
Here's the polar response of the midrange and tweeter, with 6th octave smoothing
Here's the unequalized response of the midbass and the mains. Note that the midbass is much more efficient, due to horn loading.
Here's the equalized response of the midbass and mains
Some random observations:
1) The waveguide does it's job well, but it only works down to 2700hz
2) The cardioid response occurs in the 1.5 octaves from 800hz to 2700hz
3) Despite the stuffing, the dipole peak is quite apparent at 400hz
Do you have corresponding figures for the unedited speaker?
The figures doesn't seem to agree with you on what smoothing you are using... First says 1/3 octave and second 1/12 octave.
/Anton
You don't need to totally absorb the back wave for this to work. That's the point. It needs to be attenuated and delayed. I use about 5" thick recycled denim and it works quite well to 200hz or so.
Hi Patrick,
LOVE IT!
I have always enjoyed my JBL Control25s and 28s, even bought a pair of control 1s to take away on holiday with me and driven with a Tripath amp + battery.
have LX521s as well, so can see what you are getting at here. The tweeter on the JBL Control is great for the cost - and its cheap if you blow it! although you got to be pretty unlucky with the protection circuit.
quick question.
what did you do for the upward firing woofer with this?
Brad
LOVE IT!
I have always enjoyed my JBL Control25s and 28s, even bought a pair of control 1s to take away on holiday with me and driven with a Tripath amp + battery.
have LX521s as well, so can see what you are getting at here. The tweeter on the JBL Control is great for the cost - and its cheap if you blow it! although you got to be pretty unlucky with the protection circuit.
quick question.
what did you do for the upward firing woofer with this?
Brad
- Status
- Not open for further replies.