TJLs from Kevin, Arizona, US

Hello Troels,
I recently finished building a pair of your TLJ loudspeakers and wanted to send you some comments on both the construction and their sound.  Sorry this is rather long, but I wanted to give you the best feed-back I could.
First, I would like to thank you for your extraordinary generosity in sharing your designs with others.  I also want to mention that I appreciate your extended discussions on the genesis of your designs- what works, what does not, and how you addressed these problems.  Combined, you provide valuable insight into both the variables of loudspeaker design and your own preferences regarding sound reproduction.
Building these speakers was straightforward, but I made a few changes to your design that may be worth considering.  First, I increased the thickness of the front baffle and the back to 1 ½ inches, rather than your indicated 1 and 3/4 inches respectively.  I did this because I wanted a slightly stiffer baffle, and I was concerned about the material's thickness at the base after cutting the ½ by 3/8 inch reveals at the top of the base.  I also had only a limited amount of ½ inch MDF.  Secondly, I slightly raised the base (1 ½ inch) to help accommodate the crossovers.  I compensated for these changes by increasing the depth of the cabinets, keeping the internal volume the same.  The bitumized felt, mass-loading panels proved a problem where I live ( a small town in rural Arizona, USA), and the auto-parts offerings seemed rather expensive.  I solved the problem by using a product called "Whispermat", a 1/8 inch rubberized base bonded to ½ inch acoustic foam. The Whispermat was attached to all interior panels with adhesive- not brads or staples.  According to some on the web, Whispermat provides 90% of the effectiveness of "Black-Hole 5" at a small fraction of the cost, you may want to try this product.
To accommodate the crossovers, I increased the depth at the base for the crossovers, and built a modified plinth around the cabinets, thus providing a little more space.  This plinth is composed of 1 ½ inch square pieces with a 30 degree miter on the tops, beginning 1/8 inch wide of the cabinet sides.  This miter helps reduce the chunky look of squared sides. I did not simply build a solid base and attach it to the bottoms of the speakers because again, I wanted as much depth to the crossover compartment as I could get. From an aesthetic perspective, I really like your treatment of the base, with a shadow space between the cabinet walls and the base, but implementing your idea was a problem.
As for the rest of the construction, all went as you advised.  I cut ellipses in the braces, but then rounded them over and sealed the edges with wood-workers glue diluted about 25%.  I also rounded-over the inner woofer cut-out as you suggested, and sealed the end grain.  By the way, this sealing of end-grain in internal panels is suggested by George Short of "North Creek".  In addition to the Whispermat, I also mounted polyfil to the tops, bottoms, and behind the woofer, and added some loose batting in the top chamber behind the tweeter, behind the woofer, and in the chamber below the port. For the port, I used a 3 inch tube, and adjusted for the greater diameter by shortening the length slightly.
Fitting the crossovers inside the bottom of the cabinets was a major concern.  I wanted to use foil inductors, and the 2.2 Mh inductor is nearly 3 ½ inches in diameter, while the .68 is only marginally smaller.  I basically followed your drawing, but turned the largest inductor 90 degrees to provide a little more space.  This explains my changes to your design in order to provide more depth in the bottom of the cabinet, and why I built the plinth the way I did.  Because of size constraints, I used 14 G Alpha Core for the LP and 16 G for the tweeters.  I also used Mills resistors and Solen caps.
The finished cabinets are veneered in Pau Ferro, a Bolivian hardwood, with black lacquered bases.  3/4 inch brass cones are mounted under the base, and improve image focus.  The narrow baffle, relatively modest height, and depth makes an attractive addition to our home.
I have included 2 attachments to this email with pictures of a finished speaker and the crossover.

Okay, so how do they sound?

I have spent over 100 hours breaking the drivers in and I can finally offer some comments.
First, I had a problem with the cabinet stuffing.  The speakers, while exhibiting extraordinary soundstaging, depth, and imaging, as well as a wealth of detail, seemed to be constricted.  Dynamics were not what I expected, and there was a recess in the upper bass through the mid-range.  This was noticeable on instrumental music, especially where the cello played a prominent role, and continued up through soprano voices.  The speakers were quite listenable, and in most respects excellent, but there was something wrong.  Finally, I removed the loose polyfill from the cabinets- there were only a few ounces in each, and the transformation was quite dramatic.  The speakers opened-up and the missing dynamics suddenly made an appearance, as well as a substantial increase in upper bass and lower mid-range weight.  With this simple change, I now had a better appreciation of your speaker's potential.
Tonal balance is excellent, perfect for my own personal preference.  As mentioned above, bass is solid and articulate, full without being flabby.  The orchestral foundation is well reproduced.  Mid-range, especially voice is palpably reproduced.  In a darkened room, the recreation of voices is spooky.  On several occasions, my dog thought someone was hiding inside one of the speakers!  Treble is clear and relaxed- violins sound wonderful, while triangles and cymbals have a believable shimmer.  Brass has the proper combination of rasp and weight, something difficult to describe but readily apparent.  You are quite correct in describing these speakers as suitable for jazz, rock, classical, or chamber music, but I get the sense that they become a little congested with large-scale classical recordings.  On the other hand, I have never heard a better vocal reproduction, especially female singers!
Soundstaging and imaging are first class. These speakers are replacing a sub-satellite system consisting of Vifa P13/OW1 monitors and a sealed 15 inch subwoofer.  I am pleased to find a sub unnecessary with these speakers.  The previous combo had very precise imaging and very deep bass.  With the TLJ, even though they do not reproduce the bottom octave, images appear larger- every part of the presentation is bigger and fuller.  Soundstage is very wide, with appropriate recordings extending beyond the outside edges of the speakers.  Depth is sensational, nearly as good as I have ever experienced.  At first, imaging did not seem as precise, but with familiarity I believe the images are much more three-dimensional and recreate real bodies in an acoustic space, rather than the relatively one-dimensional but pinpoint imaging of my monitors.  The drivers you selected are very detailed and I believe make a significant impact on both their imaging and musical satisfaction.  Nuances of orchestral scoring are clearly reproduced here in a totally effortless manner.  This is not hyper-detail like one finds in certain ribbons or older metal cone drivers, but rather a believable recreation of the musical event.  I used to think that the lowest octave played a key role in reproducing space and ambience, but with these speakers, I now see that detail on the micro-dynamic level is at least as important.  Even without the full bottom octave, these speakers provide a compelling sense of the recording venue.
I have only one, rather small problem with these speakers and that concerns their efficiency.  Although I have not made any measurements, I guess their efficiency is in the low 80s db. range.  I am pushing my 60 watt tube amp to its limits on orchestral passages, and although I have not audibly clipped it yet, I need to be careful.  At 60 watts, the speakers certainly fill a moderately sized room, but there is simply not enough head-room to play at concert-hall levels.  Unfortunately, I can not even consider trying these speakers in triode mode, with only 30 watts/channel.  This is a shame because these speaker's  palpability would probably be even more dramatic with the sweetness of triode.  For those running solid-state amps, they will probably want around 200 watts to play at high levels.  I should put this into context, since according to my family, these speakers play very loud already!
While on the topic of ancillary equipment, I want to share a few other observations.  First, the detail these speakers possess make them a fine analytical tool for evaluating up-stream components.  For example, I have been experimenting with different NOS tubes in my amp, and the speakers clearly respond to the different brands.  After finding a good match for my equipment, the NOS tubes substantially increase the image focus, soundstaging, ambience, and palpability of these speakers.  I also tried different AC cables on my pre-amp- a component I have lived happily with for nearly fifteen years, and the improvement was readily apparent.  Until now, I had been convinced that power cords were over-hyped tweaks- not anymore.  Although I have decent equipment, I doubt I have reached the limits of what your speakers are capable of.
Now a few questions for you.  I recently noticed on your website that you are working on a 3-way version of these speakers.  What are your goals for this project?  What are you trying to improve?  I assume you are trying to address the occasional mid-range congestion, is this correct?  Would this new configuration improve sensitivity, or does the sensitivity of the W18 limit this?  Also, I have not found any information on the mid-driver you use, not even on the SEAS website. Is this brand-new, or is there a different catalog number?  My best guess is that it is an Excel 4.5 inch. metal mid.  Is this correct?  How is the project coming?  I like building cabinets in my spare time, and would consider building a pair if the improvements justified the costs.
In closing, I want to thank you once again for your generosity in sharing your talents and efforts, I am thrilled with these speakers and you should know that I think of you whenever I play them.

Kevin Reid
Flagstaff, Arizona
USA

Comments to Kevin's TJLs:

Thanks a lot to Kevin for a thorough review of the TJL loudspeaker. And thanks for allowing me to publish your response on my website.
You are the first person to respond this way to the TJL, actually I have no idea if anyone ever build it, despite a substantial amount of mails and questions.
To your questions at the end of the mail:
Any speaker is a compromise in terms of cost and performance and no wonder the commercial two-way floorstander concept is popular - simplicity and relatively low production costs compared to the more elaborate three-way. I was/am a bit worried about releasing the magnesium driver constructions. Despite very low production tolerances on the SEAS drivers compared to e.g. Scan-Speak, the fine-tuning of the needed notch-filters to eliminate cone break-up is mandatory to get the best from the set-up. One of the problems that are shared by most 6" +1" two-ways is the need to crossover to the tweeter around 2500-3500 Hz, where really any 6" driver will start beaming when the wavelength equals the diameter of the cone. Most 6" drivers will have a membrane diameter of some 14 cm = 2460 Hz and apparent compression on more complex material will start making a noticeable impact on performance. Thus the - almost retro - idea of making a three-way with a 6" + 4" + 1", maintaining the same drivers but adding the W11 or W12 SEAS magnesium drivers for midrange at around 800 Hz. I have had a pair of W11 very cheaply and will use these for my final TJL3W construction - but at the same time implement the W12, which is destined for a coming 4-way construction. The only apparent difference between the W11 and W12 is the chassis - which may have a minor impact on sound, but shouldn't affect technical performance or require any changes to the crossover. I expect them to be readily interchangeable but sound slightly different. We'll see.
In the Seventies we wouldn't hesitate to add a 4" driver for the upper midrange and lower treble, simply because this is the optimum size for a driver handling this frequency range. The drivers may have been more tolerable and crossovers were often simple 1st orders. Making a point of crossover around 800 Hz is difficult - or was difficult - but with the availability of PC-based measuring systems and not least great simulation programmes, this has become much less time consuming. And for the diy people, we have all the time needed to do the work and adding a mid-driver doesn't add considerably to the overall cost. Whether it will be worth while will have to be seen - and heard. My first step in this direction was the SP38/13 where I used the 3806/8200 driver, a 38 mm tweeter, easily handling up to above 10 kHz. This allowed me to take the point of crossover down to 1600 Hz with improved sound compared to the SP95 or SP98. So, the purpose of this 3-way exercise will be to gain a little more speed and sparkle from the upper midrange/lower treble and enhanced dispersion characteristics.
The overall sensitivity will not rise due to this change and it's one of the TJL's compromises. I could have used the W18EX001 and get an extra dB or two but with the sacrifice of low-end extension. I have no problems with the 2 x 60 W Copland valve power amp, but agreed, these speakers need a sturdy amplifier to get the best of it. A 2½-way TJL was also considered, but then the impedance easily falls below 3 ohms and this is not valve territory either. The Acapella SE has the sensitivity needed for most valve amps around 40-60 watts but size, cost and complexity is another issue here. Compression on large orchestral works? Hmm.... I went to a HIFI show recently to hear the mega-buck B&W Diamond and yes, they have an incredible bass performance but I was disappointed to hear compressed midrange from the 6" kevlar mid-driver. I mean, how much membrane area does it take to get midrange right and reproduce a symphony orchestra? I've heard the second biggest Focal speakers and they come close having 2 x 6" for midrange.

I case you want to contact Kevin: kevinsreid@att.net