• 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!

Making a KEF LS50 Clone

Aijan

Member
Joined
Jul 13, 2021
Messages
21
Likes
39
KEF makes some of the best coaxial speakers in the market, and they use the same driver platform in all their speakers ranging from the entry-level Q series to all the way up to Blade. The drivers in their higher end speakers usually have stronger motors and better crossovers, but the higher prices don't always justify the level of performance improvements.

kef_q150_and_ls50_meta_side_by_side.jpg


KEF's acclaimed LS50 (and later LS50 Meta) speakers use a single coaxial driver which looks identical to the entry level Q150 from the outside, but the speakers sound quite different. If you equalize both speakers to have about the same frequency response to level the playing field, the LS50 will still have a better defined, less diffuse sound.

kef_ls50_meta_inside.jpg


The LS50 has a superior cabinet that features thicker walls, substantial cross braces, a flexible port, and finally KEF's version of Constrained Layer Damping (CLD) to minimize internal resonances. Since you can buy two to three pairs of the Q150 for the price of one pair of LS50 Meta depending on whether there's a sale or not, I've decided to build a LS50 clone using a Q150 driver to see if they would sound similar after equalization. The clone wouldn't have the beautiful curved front baffle and better crossover of LS50 Meta, but it would otherwise be very similar in construction.

The KEF Q150 is a narrow speaker -- only 18 cm (7") wide while the LS50 is slightly wider at 20 cm. My version would be even wider at 22 cm as I decided to reuse the original port, which had such a wide flare that it wouldn't fit if the cabinet was narrower.

The KEF Q150's front baffle is quite thick at 36 cm (1.4"), but the remaining walls are a rather thin at 12.7 mm (1/2"). There's only one small horizontal brace and very little stuffing inside, so the cabinet is naturally prone to resonances.

kef_q150_bare_bones_cabinet.jpg


I decided to use 18 mm thick MDF just like the LS50 for all cabinet walls except for the front baffle which would stay the same at 36 mm. The cross braces would also be made out of 18 mm MDF. KEF uses something similar to 3M VHB tape to have a flexible connection between the braces and the cabinet walls as a form of CLD, so I decided to give it a try.

I started with cutting the interlocking braces and glued them to the surrounding walls by using double sided 3M VHB tape instead of wood glue. The VHB tape is supposed to act as a CLD layer to loosely couple the braces to the cabinet walls.

kef_ls50_clone_braces.jpg


The outside walls are glued to each other with wood glue.

kef_ls50_clone_side_walls.jpg


The front baffle was made out of multiple layers of MDF. I opted to use M4 threaded inserts and screws in place of the original's wood screws, as all it takes a few driver test fits to strip wood screw holes. I also rounded over all cabinet edges except the rear side with a 12.7 mm (1/2") router bit, and sanded everything smooth. The cabinet weights around 6 kg (~13 pounds).

kef_ls50_clone_front_baffle_attached.jpg


The cabinet was painted white with a spray gun with the paint I already had at hand, but I may use a different color scheme in the future.

kef_ls50_clone_painted.jpg


The KEF Q150 uses a simple first-order crossover board to divide the audio signal between the mid-bass driver and tweeter. Although the crossover board had clearly been designed to be mounted on the back side of the terminal cups, KEF chose not to do that for some reason. They mounted the board on the bottom side of the KEF Q150 cabinet instead. Since the original terminal cups didn't have the necessary standoffs to mount the crossover board, I purchased another set with standoffs, and used M3 nylon screws to mount the board after tapping the standoffs of course.

kef_q150_driver_and_crossover.jpg


Here is the finished cabinet with the driver and trim ring mounted. I reused the rubber grommets from the original to mount the trim ring. I was excited to test the speaker out, so there is no clear coat to protect the paint on the speaker yet.

kef_ls50_clone_finished_front.jpg


The center of the port is positioned one-fourth of the way from the right and the top to cancel some of the internal standing waves. The terminal cups are also mounted vertically. I think they look better that way.

kef_ls50_clone_finished_rear.jpg


So, how does it sound? Quite good actually. The original had a great sound stage thanks to the well designed coaxial driver, but was undermined by the cabinet. With the new cabinet, I'm getting better bass definition, and the upper notes have better clarity.

After listening to the speaker, I made impedance measurements of the original KEF Q150 and my LS50 clone. My measurement setup isn't super accurate, but the wiggles that indicate the resonances in the original (orange) are mostly gone in my LS50 clone (green). The difference is especially visible at around 700 Hz.

kef_ls50_clone_and_q150_impedance_measurements.png


That's all for now. Please let me know what you think.
 
Like the cabinetry, but tbh, what you have created is more like a Q150+ than an LS50 clone.

This is not bad but just a matter of expectation management. Also, get your excitement but really need more measurements to show definitive improvement. Unless you have further plans, would suggest you adjust the thread title.
 
Last edited:
It looks really good. However, the LS50 is sloped (yes I know hard to replicate). By putting the driver in the middle you maximize edge diffraction effects, it would be acoustically better to have it off-center in both directions, even though it might look worse.

What is driver difference, what is enclosure, what is crossover, who knows. It would certainly be interesting to modify the Q150 crossover and compare that with the DIY enclosure.
 
Thank you for your comments. Much appreciated.

I personally think that the differences between the KEF Q150 and LS50 are mostly down to the cabinets used as the drivers are quite similar and DSP can be used to shape both speakers' frequency responses to a similar target. Two speakers with very similar frequency responses can sound different if one of the cabinets have more resonances than the other as some notes linger more.
 
Like the cabinetry, but tbh, what you have created is more like a Q150 meta than an LS50 clone.

This is not bad but just a matter of expectation management. Also, get your excitement but really need more measurements to show definitive improvement. Unless you have further plans, would suggest you adjust the thread title.
Not sure what this has in common with the 'Meta' models. Aren't they all based on the metamaterial absorption tech for the tweeters?
 
Not sure what this has in common with the 'Meta' models. Aren't they all based on the metamaterial absorption tech for the tweeters?

yes, I should correct to a Q150+ or a Q150 CLD.

I would expect the LS50 crossover is better too but seems the OP questionably downplays its role.
 
Last edited:
Thank you for your comments. Much appreciated.

I personally think that the differences between the KEF Q150 and LS50 are mostly down to the cabinets used as the drivers are quite similar and DSP can be used to shape both speakers' frequency responses to a similar target. Two speakers with very similar frequency responses can sound different if one of the cabinets have more resonances than the other as some notes linger more.
There could be differences in Xmax or distortion (inductivity/copper in the motor) for the drivers, among other (minor) things.
Building a new cabinet I would use roundovers much more generously (see below) and not put the driver exactly in the center.

index.php
 
I don't think you can replicate the cabinet which is two pieces and includes an injection molded plastic baffle, neither the port which is also proprietary, and finally there is the meta material. So if you want to do an LS 50 Meta clone, you'd need the baffle, port, and the driver with the meta material on the back (which is an LS 50 Meta driver). At which point you're just fabricating a wooden cabinet, bracing, and building a crossover board.

If you want to do that as a project, cool. But, almost top obvious to point out, the real deal isn't all that expensive and will be stellar with the proper care and feeding.
 
I think the guys from the other posts iare nitpicking too much :) I for one really like what you're trying to do, no matter how boxy your box might be. It would be really cool to see the differences in response between this and the original speaker.
 
I think it's an admirable effort and cool outcome. I haven't seen braces attached with VHB before, but it makes intuitive sense and appears to have worked. It's an upgrade to the Q150 absolutely. Getting rid of the resonances is an achievement already.

I agree though, to say it's a LS50 clone, you want to see some on- and off-axis FR measurements at least.
 
Is your trim ring perfectly flush?

When I was using a Q100 coaxial in my centre channel it made a shocking difference even being misaligned by 1-2mm. 3D printing a new one with a perfect contour is worth considering.

The stock Q100/Q150 crossover is mediocre, the breakup modes of the woofer are not well suppressed. Would strongly recommend throwing a notch filter in there, or outright redesigning it.
 
Thank you for the reminder. It's actually about a mm in, but doesn't make much of a difference according to the quick in-room measurement I made.

Yes, the KEF Q150's crossover is as simple as it gets, but I didn't bother modifying the crossover because I use DSP to fix frequency response issues.
 
Thank you for the reminder. It's actually about a mm in, but doesn't make much of a difference according to the quick in-room measurement I made.

Yes, the KEF Q150's crossover is as simple as it gets, but I didn't bother modifying the crossover because I use DSP to fix frequency response issues.

DSP doesn't always fix things like breakup modes, there can be ringing that doesn't manifest as a sharp peak in frequency response. For example, here is what I measured from the Q100 woofer and tweeter.

Subjectively I found a small improvement by changing the crossover to 3rd order slopes at 2500Hz. Evidently KEF found some benefit themselves, because the LS50 is a 2nd order design compared to the Q100/Q150 1st order design.
q100 woofer spectral decay.jpg

q100 tweeter spectral decay.jpg
 
Subjectively I found a small improvement by changing the crossover to 3rd order slopes at 2500Hz. Evidently KEF found some benefit themselves, because the LS50 is a 2nd order design compared to the Q100/Q150 1st order design.

I think alex-z has a good point. The main reason why I built a new cabinet is to reduce internal resonances, which don't always manifest themselves in frequency response charts. Modifying the crossover or going with an active crossover can further improve the sound.
 
Great little project. Maybe you can get a used LS 50 driver, sometimes they get scratches or something and end up on the used markets.
It would be interesting to see the differences between the drivers, if there are any.

I also agree with the other guys, your next target should be the crossover, cloning that from the LS50 should make a difference.
 
Back
Top Bottom