Measurements of Kef LS50

[Zvu]

On axis and roughly 30 deg off axis. Calibrated Behringer ECM8000, M-audio FastTrack USB, Holmimpulse. Gate is good to about 85-90Hz.

Measurement conditions



Midwoofer and tweeter





Simulation of Kef's factory crossover> frequency, impedance and phase response









Other LS50 measurements for comparison:

• https://www.stereophile.com/content/...r-measurements
• https://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=941:n rc-measurements-kef-ls50-loudspeakers&catid=77:loudspeaker-measurements&Itemid=153

Playing with it for an hour: on axis and 30 deg off axis

[ergo]

Very interesting and thanks for sharing. I've found that speaker very interesting, especially after hearing KEF-s presentation about the R&D behind it in AES event some years ago.

The LS50 whitepaper attached below shows it a bit smoother, but in all likelihood they have used wider averaging and the scale difference plays a role here too.


But it's very much likely that with tuning and maybe adding a bit of complexity to the xover you can get even better sound out from that speaker. Commercial design will need to compromise some things especially in that price range.

[Zvu]

Yup, scaling is definitely different and it is possible that they've measured prototype for this whitepaper but in mass production there are tolerances up to 10%. Even with that i see the bump between 500-1000Hz dip at 1.5kHz and peak at 4kHz.

Listening to the factory crossover for some time i have to give them credit for it. It doesn't sound half as bad as it measures. You can hear that it is not linear but to what extent would be beyond me if i didn't measure it.

[Zvu]

As usual, i'm posting .txt files that you can convert to .frd or .zma if your simulator needs it. I derived frequency files with about 10ms gate time. I didn't merge it to get low frequency response of woofer (under 120Hz) because i'm not interested in it. In the .zip file that is available for download from my DB account there is another .zip file that can be uploaded to HolmImpulse if you are not satisfied how i extracted sim files so you can do it as you see fit. All measurements are done from about 108cm on axis and approximately 30 degrees off axis. It is not good as 0-180 in 10 degrees increments but if you cross it over between 2000-2800Hz you won't get into trouble. Any fine tuning i recommend to be done by measuring and listening of your drivers.

If there's anything that you want me to do (other than measuring TS for woofers because i'm not doin' it ) feel free to ask.

https://www.dropbox.com/s/zni55ujjo8yzgh9/Kef%20LS50%20measurements%20and%20sim%20files.zip?dl=0

Few comments:

In Kef LS50 they've seem to done around 8dB baffle step compensation for midwoofer. That sounds like a lot but surprisingly it doesn't sound that bad at all. I guess that it doesn't go much low so it kinda evens out. Tweeter level is set, as can be seen, a 1.5-2dB higher than midwoofer. Again, i'm surprised that it isn't perceived like overly bright. You can hear it being louder but it doesn't make you want to turn the volume down - quite strange actually, unlike any metal dome i've listened so far. Also, good thing i've prepared and measured both cabinets because there are few measured differences between drivers. You'll see.

So, that being said, feel free to share what you've achieved with this or with your LS50's. I know i will.

[Zvu]

I've been listening for a few days this crossover:






Loudspeaker No.2 has few slight changes in crossover and one significant:






The surround on that midwoofer didn't get enough coating so its resonance is quite visible. I decided to solve it with LC filter. 2x4,7uF in series and 7.5mH/Rdc=12.5ohm. 2.2uF will also work but i ran out of 2.2uF capacitors and i was eager to listen to it. The coil is Jantzen, catalog number 000-1541 and can be bought at HiFiCollective for about 3 euros a piece. Its outer diameter is 26mm and it is 15mm high but without it the response looks like this and that was unacceptable:





So, last night i've made some measurements of modified Kef LS50. Bare in mind that this is 1dB grid.

0, 10 and 20 degrees



30, 40 and 50 degrees



60, 70 and 80 degrees



Here I had to make the grid 2dB to be able to see the level of 90 degree against 0 degrees

So 0, 80 and 90 degrees




Tangerine phase plug obviously works like a charm. There will be some tweaking by ear but all in all, I'd say that's it.

And now, traditionally, a little livelier display

Image not available

[ergo]

A very nice result indeed. You've taken a good speaker and tuned it to max and at least on screen it looks very good measurement wise and probably sounds good and well balanced as well.
Thanks for sharing!

[Zvu]

Yes, it surely is nice. Still lacks that authority, down low, that larger cone loudspeakers have. I'll be adding two 6.5" woofers per one LS50 and i'll cross it over at 150Hz or so. 150uF in series and 2mH in parallel with LS50 midwoofer turns it in to high performance and high power midrange. Its 50mm aluminium wire wound voice coil can take loads of power if deprived of frequencies under 150Hz. I'll make woofer cabinets separate to try WtmW or tmWW to see what works best for me. I want to keep the LS50 cabinet and all of the great engineering that went in to it. Only the fancy port will be violated and end up with a sock in it

I always have this picture in my mind only my stands would serve more than one purpose:




So, phase one - finished.
Phase two, loading...

[Juhazi]

Yes, LS50 yells for a real woofer crossed around 150Hz, not just a sub.

https://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=941:n rc-measurements-kef-ls50-loudspeakers&catid=77:loudspeaker-measurements&Itemid=153

[Zvu]

Here is horizontal radiated power at angle of 180 degrees (+/-90 degrees in regard to on axis):




And here are individual curves:

[sfdoddsy]

I enjoyed the LS50 more than most box speakers when they came out, but you've identified their biggest problem. They simply don't go low enough to cross to a sub. Unless you have stereo subs which are happy playing up to the 200Hz mark. In other words woofers.

I used the SLS 10 with good results, and then proper sub below that.

But now you're talking a 3 box system.

[Zvu]

We'll see. Woofers i ordered earlier are here for some time now. I've cut boards for cabinets. I need to assemble it and do the measurements. I've simulated it though to see if it would work.

Been busy these days with work, kids and making plans and preparations for summer holiday. I'm reluctant to take more responsibilities if i see i can't do it all properly at the same time. I'm letting cabinet building rest a little while i finish what i started and in the meantime modified PSM XO for LS50 works quite well so music keeps flowing.

[AJINFLA]

I enjoyed the LS50 more than most box speakers when they came out, but you've identified their biggest problem. They simply don't go low enough to cross to a sub. Unless you have stereo subs which are happy playing up to the 200Hz mark. In other words woofers.

Plenty of "proper" sub drivers that can handle up to 200hz (and beyond). Dayton RS series come to mind. Integration that high can be tricky, unless the subs are close, ideally, stands for the monitors.
Or just stick the whole thing in one box.

cheers

[dwk]

Ha - of course AJ chimed in just before me. I have a pair of Q150's that I'm planning to try to integrate with a RSS210HF in a single box similar to what he did (although externally amped). That may ultimately prove too big for even my rather large desk. An Anarchy would probably be the fallback. And, to tie it in with that other thread Jon and Steve have going, I'm hoping to use my CNC to eventually make some cool cabs for them. If you have the room, the RSS210's can easily go up to 500+ without problems - really nice drivers for the $$$.

I really need to break out my measurement gear and see whether pursuing xover mods similar to what Zvu has done to the LS50 would be as beneficial with the Q150. I'm planning a DSP xover solution between the coax and the woofer, but am planning to stay passive on the mid/tweet. If the power response is ok, then I can handle FR in the DSP, but the results Zvu is showing are definitely impressive .

[Dave Bullet]

Very impressive work as always. You should apply for a job @ KEF.

I'm surprised the original xo 1.5 - 4KHz hump doesn't make you want to climb the walls. I'm very sensitive to that area being prominent.

Why only a 6.5" woofer to extend the bottom end? Are you backing this up with a subwoofer anyway? Budget or enclosure size constrained?

If you want to keep a narrow front, do you think a larger side mounted woofer could work given your rolloff is in the low 100's?

[Zvu]

Very impressive work as always. You should apply for a job @ KEF...

Hey Dave, glad to see you.

Regarding Kef, i wouldn't let something that measures worst than the drivers can deliver go into production, and they tailor the crossover per price tag. They'd fire me in a hart beat.

........I'm surprised the original xo 1.5 - 4KHz hump doesn't make you want to climb the walls. I'm very sensitive to that area being prominent...

I was equally surprised. I could clearly hear that it is louder than the rest but it wasn't insulting. How's that possible, i really don't know. I did listen them about 15 degrees off axis.

Knowing what problems you had with your loudspeakers, i threw in notch filter at tweeter that can be used to fine tune that part for the ones that don't like it flat. Although, you still haven't measured off axis response with your new crossover. Summing those curves, you might end up with flat listening window response. I'm a patient guy, i'll wait

....

Why only a 6.5" woofer to extend the bottom end? Are you backing this up with a subwoofer anyway? Budget or enclosure size constrained?

If you want to keep a narrow front, do you think a larger side mounted woofer could work given your rolloff is in the low 100's?

1. Budget always plays very important role - at least for me. Performance plays equally large role too. So you see, i'm in a problem. I want it very good, but i want it at affordable price. That isn't always possible.

2. Size isn't an issue for me but through the years i came to the conclusion what i do and don't like. I really like how good bookshelf deliver lows. This might sound strange but that's how it is. Floor standing ported speakers do go lower but the bass becomes less precise to my ears. I'm always for ported bookshelf and few sealed subs. That is only matter of personal preference, not saying one is better than other. Smaller woofers, that i choose and prefer, usually have lower moving mass (Mms). I'm always for 2x6" rather than 1x8". Cone surface might be the same and Mms, but there are two points of sound origin (it will interact with the room differently) and two motors where there would be one (less compression and hopefully distortion). If i wanted even more dynamics (since i don't tune speakers bellow 40Hz) i'd just use more 6" drivers. Largest i'd do are 8" for home drivers. With PRO drivers it's another story. They can give me what i want and are sensitive but they are quite hard to move. On low volume they sound a bit lifeless and when i turn it up to get it sounding good (in my room), it is too loud.

3.Thing is that i live in apartment - rooms i listen music in are 4x4m and 4x5m. Tuning low in rooms of that size that are made of brick creates mayhem with bass and sub bass. All of my loudspeakers are tuned F3 to about 40-45Hz so they work with the room. One could see this as a flaw since i don't know the room someone will be using that loudspeaker, but the logic is next - if you have smaller room, you don't want your lows to go wild and become mushy. If you have large room - what the hell are you trying to achieve with 5-6" bookshelf there ? So in larger rooms i'd definitely recommend few sealed subs crossed at 80Hz or even 60Hz. Then the bookshelf can reach the loudness providing it doesn't distort much that low.

4. I don't like side mounted drivers. Again, thing of personal preference, but i always have the feeling that i lose something when not listening direct sound. I could be testing that soon enough with LS50 and bass cabinets - i can always turn them 90degress off axis and give it a listen.

[Zvu]

Ha - of course AJ chimed in just before me. I have a pair of Q150's that I'm planning to try to integrate with a RSS210HF in a single box similar to what he did (although externally amped). That may ultimately prove too big for even my rather large desk. An Anarchy would probably be the fallback. And, to tie it in with that other thread Jon and Steve have going, I'm hoping to use my CNC to eventually make some cool cabs for them. If you have the room, the RSS210's can easily go up to 500+ without problems - really nice drivers for the $$$.

I really need to break out my measurement gear and see whether pursuing xover mods similar to what Zvu has done to the LS50 would be as beneficial with the Q150. I'm planning a DSP xover solution between the coax and the woofer, but am planning to stay passive on the mid/tweet. If the power response is ok, then I can handle FR in the DSP, but the results Zvu is showing are definitely impressive .

RSS210-HF is one of the top 8" woofers in my book. Their price in Europe is ridiculous though. The only woofer i know of that can work in sealed enclosure of normal volume and reach down to 40's with no apparent problem. I like everything about it. I don't think Q150's need crossover as high as 500Hz. It just might be that it can work 150-200Hz. But only measurements and listening can tell. Mms of RSS210-HF is highish for me so i probably wouldn't let it past 150Hz.

Agreeing about using DSP for bass but i'd make high pass passive filter for Q150 mid-woofer as well. I don't want to poop on other people's lawns but before you decide if you will be using DSP for mid-hi, try just to listen Q150 with passive crossover with and without DSP. No interventions, no filters, DSP set to flat and letting the sound go through it on one speaker, while other plays without DSP in chain. If there isn't any noticeable difference, then i'd use DSP. MiniDSP 2x4 that i tested this way muddied the sound of mids so i gave up on it. I used Visaton B200 wide range for this test with no passive crossover.

I have really high hopes for Fusion plate amps though.

[dwk]

RSS210-HF is one of the top 8" woofers in my book. Their price in Europe is ridiculous though. The only woofer i know of that can work in sealed enclosure of normal volume and reach down to 40's with no apparent problem. I like everything about it. I don't think Q150's need crossover as high as 500Hz. It just might be that it can work 150-200Hz. But only measurements and listening can tell. Mms is highish for me so i'd probably wouldn't let it past 150Hz.

Yes - performance in acceptably-sized sealed enclosures is exactly why I settled on it. Not many other drivers that work well in this application.

And I agree - with the Q150 I'd target 150-200 xover - I was just pointing out that the 210HF can be thought of as a 'woofer' rather than a 'subwoofer', and play up high enough to mate with true midrange drivers.

Agreeing about using DSP for bass but i'd make high pass passive filter for Q150 mid-woofer as well. I don't want to poop on other people's lawns but before you decide if you will be using DSP for mid-hi, try just to listen Q150 with passive crossover with and without DSP. No interventions, no filters, DSP set to flat and letting the sound go through it on one speaker, while other plays without DSP in chain. If there isn't any noticeable difference, then i'd use DSP. MiniDSP 2x4 that i tested this way muddied the sound of mids so i gave it up. I used Visaton B200 wide range for this test with no passive crossover.

I have really high hopes for Fusion plate amps though.

I'm not surprised at your results with the 2x4 - the earlier MiniDSP units were definitely marginal performers. The newer HD units are better, but it seems they really might have finally gotten it right with the new SHD units. Pricey at $1100 though for only a 2-way output (although it includes Dirac Live, which is probably where some of the money goes)

I'm engaged in exactly this type of listening/evaluation process at the moment though. My original plan was to use a MiniDSP Nano-Digi to feed external DACs. Listening to it in flat/bypass though, I really think that even just the sample-rate-conversion process is audible, and you can't bypass that. I probably need to do a bit more testing, though. My current plan is a PC based solution centered around BruteFir (an oldie, but still a goodie). Output will be one of three choices:

• MiniDSP UDIO-8 out to external DACs
• Okto Research DAC-8 Pro if I decide to stretch the budget a bit
• Motu 8A, 8D or Ultralite-Mk4 if I can firmly establish that these actually work under Linux

I'm currently leaning towards the MiniDSP UDIO-8. Output over USB is direct (i.e. no SRC), and Kal Rubinson is using it in his multichannel rig and claims it's as good as any other spdif source he's tried. SRC is applied to the spdif inputs though, which works well for secondary sources given the constraints of fixed-rate FIR filtering.

Part of my leaning towards the UDIO is that I'm currently listening with the amazingly cheap combo of a Khadas Tone Board into a JDS Atom used as a preamp, with a Neurochrome Mod-86 providing power to the Q150s. The Khadas + JDS Atom is a whopping $200 total, and it sounds amazing - good enough that if I can just filter the spdif out and integrate the woofers I'll probably be a happy camper.

[Zvu]

Yup. I too am listening my Khadas TB for a few months now. I have compared it with some big guns.

For the ones that don't know what are we talking about:

https://www.audiosciencereview.com/forum/index.php?threads/review-and-measurements-of-wesiontek-khadas-tone-board-dac.4823/

[ttan98]

Yup. I too am listening my Khadas TB for a few months now. I have compared it with some big guns.

For the ones that don't know what are we talking about:

https://www.audiosciencereview.com/f...oard-dac.4823/

Can you give your impressions of the outcome of this comparison between Khadas and the big guns?

[Zvu]

The cheapest one was 5 times price of Khadas. Few of them were 10 times Khadas price. Most of the time we couldn't tell which dac is playing. Certain music did bring out the differences but they were subtle in my opinion. My personal impression is that little Khadas works well no matter the price. For 10 times the money i don't expect 10 times better sound but i do expect night and day difference.

What i consider subtle might be a lot for someone and vice versa so i can't help with that.

I used my HP thinclient with Daphile on its ssd. Usb to Khadas, step attenuator, my diy amp and modified Kef R300.

Since diy scene here is fairly strong, mods are already being done by others. Changing to Oscon caps, diy power supplies etc.

I'm waiting for guys from Khadas to fix what is now known as ess distortion hump and to throw in balanced out so i can order it and compare with this one. It will be interesting to see if there will be any perceivable difference.

[JonMarsh]

Plenty of "proper" sub drivers that can handle up to 200hz (and beyond). Dayton RS series come to mind. Integration that high can be tricky, unless the subs are close, ideally, stands for the monitors.
Or just stick the whole thing in one box.

cheers

I'm with AJ on this one. . Heck, most of my favorite "woofer" drivers are actually "short throw" (12-14mm) sub drivers, or advertised as that by the original manufacturer. Wavecor SW223BD02, for example. Or the Aurasound 12" underhung VC drivers I used in the Isiris.

But I'm in violent agreement that you can't treat them like subs and put them over in the corner; they must be setup with the "mini-speaker" aligned, basically like a multi-piece three way (Wilson Audio, anyone?)

[Juhazi]

A mid-size coaxial with an integrated subwoofer is a very nice combination for hifi! With some dsp'ery it can have smooth room-response 20-20kHz and gives the benefit of distributed subs. Extremely tight and controlled, even punchy bass and prescise imaging.

I cloned the Gradient 1.4 and made it active with Hypex FA123 board. Woofer is SB 29NRX75-6, coaxial SEAS MR18. The project is here https://www.diyaudio.com/forums/mult...ml#post5591356 Measuring and setting W-M xo was painful indoors during the winter. Now they are serving my daughter, I guess I must call them finished!

[AJINFLA]

Very cool Juha. I will have to slog the 9 page DIY thread later. Is that the Magnesium tweeter? I once asked Madisound if a umm, "replacement" was available for say a damaged C18, but no luck. If Aluminum, good too (thanks), just that I think the Magnesium has the best very top end dispersion of the family, the bane of many a coax. Actually my only critique of the R series KEF is what I thought (my perception of a factory R series tower) was a slightly closed in top. The Q had this too though...

Mms of RSS210-HF is highish for me so i probably wouldn't let it past 150Hz.

Hi Zvu, this seems intuitively correct, that I believed it myself many moons ago, given so many practical examples of big heavy "slow" bass subs. But it's actually the inductance that plays a key roll here and despite the highish Mms, the Daytons low inductance and stiff aluminum cone (thanks mass!) push the resonances out of the usable audio band, so they can be used quite high. I actually used a HF in a 2way with the DXT (@ 1.8k) for a contracted project! . DSP for sure.
I agree with you, the Minidsp stuff is really hit or miss. I usually stay clear except for prototyping. The Fusions on the other hand, are stellar and absolute bargains. We'll see with the QC issues, had some firmware drama with a couple units, since resolved.

cheers

[Juhazi]

The tweeter is a spare part for SEAS T18REX/XACG which is not listed at SEAS at all, they are custom models used by several Finnish speaker brands and diy people.

https://www.hifitalo.fi/seas_t18rexxacg

I destroyed one of the MR18 textile tweeters and borrowed the alu from my other speakers using that T18. The alu has less distortion and different response above 10kHz, because it sits differently in relation to the cone. I can't hear the difference... Hifitalo has those spares, email them in English!

My 10" high Le woofer is sealed box downfire and crossed 300Hz LR2 (acoustic). Bass sounds amazing, I can't hear the transition between drivers at all in any circumstances! LR4 might be a different story.

[Zvu]

....
Hi Zvu, this seems intuitively correct, that I believed it myself many moons ago, given so many practical examples of big heavy "slow" bass subs. But it's actually the inductance that plays a key roll here and despite the highish Mms, the Daytons low inductance and stiff aluminum cone (thanks mass!) push the resonances out of the usable audio band, so they can be used quite high. I actually used a HF in a 2way with the DXT (@ 1.8k) for a contracted project! . DSP for sure.
I agree with you, the Minidsp stuff is really hit or miss. I usually stay clear except for prototyping. The Fusions on the other hand, are stellar and absolute bargains. We'll see with the QC issues, had some firmware drama with a couple units, since resolved.

cheers

Hi AJINFLA,

I'm not worried that it will not play up to 500Hz and measure good. I'm wondering about how will it sound at lower volume because I've found that lower mass woofers are more responsive and are better with micro dynamics at low level listening. I generally get best results when i let midrange do all the midrange it can. At 150Hz cone movement of Kef concentric is minimal so it will not affect the highs, everything from 150Hz and up comes from a single point, lobing between woofer and midrange is practically non existent and 9gr cone does all of midrange which means a lot when listening at low volume.

How did you deal with 3rd and 5th harmonic with your 1800Hz cross over ? It looks sinister:

[AJINFLA]

How did you deal with 3rd and 5th harmonic with your 1800Hz cross over ?

I applied a notch and low pass filters .
What you show is an unfiltered driver. Of course as you say, if the Dayton is prohibitively expensive, there may be better choices there. I have some Wavecore SW215s (49g mms) arriving tomorrow, which will be my first experience with brand. On your side of pond, the Reckhorn D165s seem to be a bargain, so 2x might be an option for you at reasonable cost vs a single 8.

cheers,

AJ

[AJINFLA]

Hifitalo has those spares, email them in English!

That didn't work so well for me...until you intervened.
I suspected the Alu, but your picture almost made tweeter look white like the Magnesium. Isn't that what Jorma (RIP) used in the 1.4?
One thing you can see in the KEF/Tangerine vs the Seas (old school coax) is the "pinching" around 8k. I usually apply a narrow-ish notch to help mitigate the excess sound power in this region (the KEF seems to have this inherently, though a bit broad). Helps take some edge of sibilant tracks. Luckily, even drywall starts to be absorptive at these frequencies, so the room/position/taste matters here also.

cheers,

AJ

[Juhazi]

Yep, the Gradient 1.4 has MR18 cone and C18 Mg tweeter, perhaps some custom job with surround and spider etc. Gradient has been using mainly SEAS drivers for many decades.

[Zvu]

... On your side of pond, the Reckhorn D165s seem to be a bargain, so 2x might be an option for you at reasonable cost vs a single 8.

cheers,

AJ

8 pieces of Reckhorn D165 (4 per side) cost 240€ and are still 60€ cheaper than one pair of RSS210HF.

[AJINFLA]

8 pieces of Reckhorn D165 (4 per side) cost 240€ and are still 60€ cheaper than one pair of RSS210HF.

4 always better than 2 .
Though of course I'd be tempted to do your sub tower stands as cardioids if using 4/side. A simple way (for simple folks like me) is to combine low Q drivers like the Reckhorn enclosed, with some high Q drivers dipole open baffle. Can get as much as -20db rear reduction. Did something like that to shore up a coax long ago at a DIY gathering. Used a Peerless XLS and a High Q TCSounds open baffle sub.



Makes them relatively easy to position with regards to the wall behind, especially in smaller rooms, etc.

cheers,

AJ

[Zvu]

I am on hold for projects like that. First i'd like to make or buy 8-10 amp channels - preferably AB class. When i do that, then i can consider cardioid and few other things i want to try. Doing it passive is plausible but i make enough work for myself out of simple classic 3 way passive crosover.