Thiel Owners

True enough, wider baffles isolate the waveform launch at the speaker, as opposed to the room. His included assumption is that the curved baffle is effectively diffusing meaningful surface anomalies and that an infinite baffle is ideal. I might agree with the later, but I am challenging the assumption of a perfect-curved baffle.

Also note that I judge models with similar side and top environments to sound more natural than those with small-dimension horizontals and large dimension verticals, both between drivers and above the tweeter. 

One of our forum members is outfitting his CS3.5s with my current surface treatment for comparison to the stock control unit.

Unsound - I think you took issue with my paradigm restatement (but I could be wrong.) I see a paradigm as a lens through which to view the territory under exploration. The values and results might not be challenged or changed. What changes is the perspective, the point of view. I came to realize that 80% of my CDs were not suitable for sharing or showing off or enjoying. I am adopting a new paradigm which takes responsibility for that situation.
Regarding Dunlavy, I had issues with John's approach (my exposure was late 90s). I didn't fall in love with his products. And soft baffles were just part of his rather inclusive approach. His cabinets had sharp corners and I could hear them more than Thiels'. Of parallel meaning to me is the early Vandersteen baffle-less cabinets, which reminded me of my own early work of hearing and minimizing diffraction effects. I liked the rounded edge solutions we developed at Thiel and those solutions were later refined with larger radius easing. But that seems to have been the end of the story, and it shouldn't have been, and wouldn't have been if I had stayed at Thiel.

My "Realizing the Artists' Dream" paradigm encouraged me to pursue why and by what methods the V and D products attracted so many avid adopters. Both brands were enormously successful compared to Thiel's limited appeal. Last summer, my explorations got serious when reading cabinet wall vibrations with a stethoscope and found significant chatter on the CS3.6s baffles. The baffles weren't moving detectably at 3" well braced thickness. The noise was on the surface. That noise sounded a lot like the "hardness" attributed to Thiel products at high volume and musical complexity; and the 3.6 was far worse than the 2.2, which was worse than the PowerPoint1.2. I gradually came to terms with accepting responsibility for whatever I might find - easier since I was personally involved with 2.2 and 3.6 cabinet development.
I am now several iterations deep into soft, layered overlays on Thiel's curved baffles. The results are different than my assumptions predicted, and very encouraging. My work is also informed via previous work with laminar and turbulent flow management. So, I am thinking and experimenting a lot with "soft" baffles.

Presently I seek out any and all CDs, regardless of pedigree, to learn what they might teach. And I am constantly surprised that nearly all of them pull me into their Artists' Dream, their musicality and charm. I cross-check every cut with Cans and stock 2.2s and 1.6s, stock and modified.

Regarding Andy's point of cost. Indeed products are limited by their budgets. Part of the appeal of this exploration is that an effective solution will cost less than premium components and wire; plus the improvements are in parallel to those electronic improvements. Not either-or, but both-and.

Unsound - in developing the CS5, I evaluated a tapered cabinet, large enough for the woofers at the bottom and as narrow as possible at the top. We never mocked up or tested it. The 3-dimensional cabinet geometry would have added considerable cost and production engineering time-to-market delays. CS5 product development was on the fast track. I hadn't heard of Jim's "wider baffle" statement.

I lobbied for a CS5.2 with that tapered 3-D geometry, plus a frontally-contoured baffle to put all the drivers in their proper physical alignment. Half of that extensive crossover (dozens of parts?) is for time delay for the upper and lower midrange drivers. That requirement would vanish and the signal path would be enormously shortened, allowing ultra-quality components. And so forth and so on.

As a company we chose to concentrate on our core lower-price market, leading to home theater products and subwoofers and a lot of scrambling. A different Thiel Audio would have resulted from pursuing a 5.2 and other upscale products.

Unsound - wave guides might have merit in the right designer's toolkit. But I am not that guy. The famous "lobing" is caused by vertical offset between the tweeter and midrange as a function of wavelength and crossover slope mating. The 3.5 drivers are very close, providing a pretty large vertical window for any seated listener more than 8' away. I consider that set of parameters as baked in to the fundamental product design. Wide horizontal energy dispersion is also quite good in the stock product; I'm keeping stock geometry.I'm working with surface treatment to absorb the energy propagating along the baffle surfaces.

I do not know Bau's or Dunlavy's later work, but your near triangle makes good sense. Before I left Thiel Audio, I experimented with tapered cabinets, which require CNC router-cutters with differing angles per each taper angle. In a production environment, quite a bit of complexity is added, plus no matter how sophisticated the cutter, the sharp veneered panel edge is somewhat damaged. We stuck with rectangular panels where the CNC-mounted saw could cut precise, sharp mitres.

We have identified a replacement midrange, with a new tweeter in the works - giving me the courage to address this classic model, which sold more than 5000 pair in a 5 year run from 1987 to 1992. That in many ways was a high-point of connection with our audience and market. At this point, I cannot address the equalizer. And today there are good subwoofers to augment the 3.5's extraordinary sealed bass.

Prof - there is some low-hanging fruit for the 2.7 upgrade. A 400uF electrolytic midrange series feed cap (although well bypassed) is certainly limiting the clarity-directness. Other XO tweaks too. The 2.7 baffle is nicely optimized, but I bet we'll gain some harmonic delicacy with treatment.

Prof - thank you for your thorough comparisons and commentary.I have read Stereophile's reviews of the Joseph Perspectives and admire the outcome. I am a big fan of magnesium as  a driver material - the SEAS Graphene material seems great to me. Thiel products never got to that level of refinement with their associated costs. But in my fantasies, I would develop a tweeter with such a diaphragm to take the breakup above 40K Hz without electronic intervention. My experience with microphones says that magic would flow.

For the record, we can improve the 3.7 performance with passive component upgrades (like Beetlemania's 2.4s) and a little baffle treatment. Don't sell them yet.

Prof - the 2.7 baffle is 3" thick and well-braced. I doubt that it is moving much. The baffle treatment I am developing is a 3-layer felt and fabric overlay to make the baffle surface virtually disappear. This territory is new to me, and I am making progress, and have 4 pair under development with 3 co-conspirators.

JA - circling back to 1.6 and 1.7 sales: Rob says approximately 2500 pairs of 1.6 and 1000 pair of 1.7 were sold. The preponderance of evidence suggests the 1.7 has 4th order Butterworth XO slopes for time coincidence, but not phase coherence.

The elements of a speaker all work together as a whole - a pot of soup. Trying to separate any single element leads to confusion.

Unsound - thank you for those links. I remember the Angelus; Bau was putting the tweeter into the (relative) infinite baffle domain (like Jim's desire for the CS5 to have a wide, curved baffle). And his woofer wanted a wider propagation environment at its high end. A rather quirky implementation of good ideas. Spica had a solid high-performance appeal. I've never heard one.

The Dunlavy interview is excellent. I would project the he and Jim would agree on everything. Dunlavy brought tons of ability and experience to his work and built a successful company which permitted his development of his good ideas. As I mentioned, I only heard his early products, which I judged as not thoroughly refined. But I would like to hear some later Dunlavy speakers. I bet they're excellent.

Request for contact information - I have Thiel's original Sony SCD-1, and it needs repair. Someone here noted a tech doing this work. Can someone please re-post that lead or any leads to someone repairing and/or upgrading the SCD-1. Thank you. Tom

Jacksky - a long answer may be of interest to some on this thread. I just lost a few paragraphs and will find some time soon. Hint: the 1.6 is the sixth generation of the model 1 (with pre-history of O4 and O4a).
Tom

Jacksky - I posted a long version answer perhaps a year ago on this thread. 3 is the model 3 and .5 is the 5th generation. The CS5 is a 5-way, and has no passive radiators. The bass is sealed and all 3 woofers extend below 10Hz.

Thanks Beetle. I'll point out an error in that old post, when I was dusting off my remembering stick. The sequence of the CS1 series (6.5"2-way) goes O4, O4a, CS1, CS1.2, CS1.5 (5th generation), CS1.6, CS1.7. The model 1 represents the most changes of any Thiel model family, due mostly to it being able to accept trickle-down upgrades from the many products above it.

Jazzman7 - Thanks for the lead to the CS1.5s at TRM. They're on their way here as my intended proof of concept workhorses. I'll apply what I've learned to these 1.5s due to their relatively simplicity, low risk and firm foundation. Are there any 1.5 fans here?

YY - I suggest you find an alternate reality where you can turn your room on its side. That 21' x 25' room with a 12' ceiling would be fantastic!

Andy - I'm interested in learning of your work and look forward to reading more on your thread.

As an observation of your opening statement, I read what seems like a false-start. I have followed John Atkinson's reviews, observations and attitudes with great interest since meeting him in the 1980s. Your semi-quote: "Since if a speaker can produce a step response correctly, therefore it is time-phase coherent, and therefore it must be "good" - is not something I hear him saying. He does say that "all else being equal, phase coherent speakers tend to produce exceptional imaging". (My semi-quote). I state and know that Jim would agree and think that John would agree that phase coherence (however measured) does not produce a "good speaker". I think that phase coherence (as an objective or success) increases the difficulty of making a good speaker by a large multiple, and that many attempts fail in many ways, including Thiel's attempts.

I look forward to your posts in your coherence thread and this one.
Tom

I also find the topic of phase fascinating, including the auditory neurology and learning that helps us hear, as I have expounded on this thread.

There are some confusing terms. Time alignment is a great term, but was trademarked by Ed Long around 1980, which is weird. Time coincident means the same thing, and I recommend it. Time coherence merges two concepts, leading to confusion. Also, Atkinson and others have hijacked "coherence" to include the smooth phase transitions in non-coincident systems with large phase shift - as long as the phase transitions are smooth, some call it phase coherent. Thiel, Vandersteen and many physicists reserve the term "coherence" to mean that the phase response remains minimum, ie. it does not depart from flat. A good measure is "excess phase" approaching zero. When terminology is scrambled, concepts are harder to define. 

George - I always wonder whether my comments are clarifying or just muddy the waters darker. My survey of the web suggests that there are strongly held but contrary opinions on the subject. I, as you might guess, have paid attention from various angles including music playback and record production. I take comfort that my opinions align with the best of the bunch, those recordings that play well on all systems.

Here's a link to a good explanation of the territory. It demonstrates the 'stealth phantom' of linear phase filters, whether they are analog or digital.

https://www.audiomasterclass.com/newsletter/the-difference-between-minimum-phase-and-linear-phase-eq...

I'm off to a dinner meeting. I agree that the terms are used pretty loosely.BTW: Thiel considered in-room power response to be the number one parameter.

T

Andy - regarding your last paragraph on the 3.7 charts. Those charts would indeed FAIL my definitions of both "time coincidence" and "phase coherence". BUT I don't buy the charts. Note they are taken at 50" and on the tweeter axis, both of which meet Stereophile's MO, but are illegitimate for the system under test. At 100" and 35" ear height, those measurements actually yield clean triangles without those false anomalies.

I am really not the guy to try to explain this stuff. It's pretty deep and complicated by real and imagined factors, and my knowledge is real, but not fluent. Here is what I can say: Jim was an engineer's engineer, and an honest mathematician and physicist. He pursued the problems to their root causes and engineered thorough solutions. I have only conversational knowledge of what he knew to his bones. That said, I was in the lab every day for 20 years and helped devise and implement the test set-ups and cross-checks. For frequencies lower than our chamber could reliably measure (100 Hz±) we had a roof system (100'x 500' flat roof) which we correlated with both straight and ground-plane techniques. It is germane that our test system included impulses for step responses at 1/3 octave intervals from 10Hz to 30kHz. Thiel speakers pass those pulses from bottom to top, including through the crossover regions. I don't know of a more stringent test for coherence, and to my knowledge, all speakers from the 03 in 1977 meet that requirement.

Unsound - right on. Dunlavy took them to task and Vandersteen chided in an interview. Thiel chose to not respond as a policy, but discussed it with John. The biggest problem with them is their (usually) 50" mic distance, which does not allow the wavefront to integrate, and then JA reading in to the ragged scan, which was caused by the technique, not the speaker. They published excess phase charts on older models, which are near zero, and even at 20kHz, they drift less than 10°. In our development measurements, all models (in my time, and probably later) fell within that 10° from minimum, except the CS5, which was <5° (plus Jim published a time delay spec which I remember being in the microseconds.)

Andy - I don't remember using "phase coincidence" or "time coherent", since I find both terms confusing.

The separate drivers are placed in 3-D space to sum properly when listening at 35"±~4" and greater than 8', optimized for 3M / 10'. That is a stated constraint, which we believed to be reasonable for real people in real listening situations. A lot of the confusion revolves around magazines/reviewers not being able to measure in those real-world situations.  

Here's the Stereophile measurements of the CS1.5: https://www.stereophile.com/content/thiel-cs15-measurements
Regarding Vandersteen, I plead ignorance of any particulars. Richard didn't go places that were unfriendly. He knew the pitfalls of those measurements and had plenty of sales to side-step that playing field, although he and JA were on the best of terms. From everything I have read, he was pursuing the same expression of minimum phase as Thiel.

Beetle - Stereophile's next publisher after Larry Archibald did not tolerate the editor producing content that put their major advertisers in dim light.
Andy - Design intent is 8' minimum distance. Farther away yields larger listening window.
Jon - power response is the energy dispersed through the whole room, which affects overall tonal balance, especially with phase coherent systems where reflected energy begins at the speaker as coherent waveforms.
Robert - I like the tenor of your observations. Keep it up.

As you all know, I am taking on the 3.5 after the 1.5. Both models are from the golden era and both sold more units to more loyal customers than any other products (the CS2 also fits that description, but its actualized design limitations and lack of parts takes it out.) My 1.5s are on their way. I'm looking for a pair of 3.5s to borrow or buy.

Rules - I like the term "time coincidence" as less confusing.Also note that the term of art 'minimum phase' includes the least possible, given all constraints. Note also that the phase lead vs lag in crossover regions does completely cancel each other to net zero. Bass tuning does not have that counteracting slope, so bass phase leads and lags with its resonant behavior. DSP can do stuff, no doubt, but with its own consequences.

Unsound - IF there were a boutique company refurbishing classic Thiel speakers, I think the CS2 would be a prime candidate. Let's spin a little: There were 5500 pairs made, built like little tanks, available for short money. One might probably insert CS2.2 drivers (as successors are developed) with necessary XO mods, or 3.5 tweeters and/or midranges. Solveable.  With today's tools, such as SpectraFoo and Klippel, a more robust anti-diffraction solution could be developed to replace the 1/2" MDF grille with perhaps a 1" equivalent like the CS1.5 AND the 04a rubber elastomer strips could be increased for very effective coupling. The CS2 has 6-9s coils and wire and styrene ultra bypasses around 1uF styrene bypasses for plenty of salvageable XO value. With my developing 'controlled wave-launch' baffle / grille technology, I am confident we could get class-leading performance.

The necessary ingredients are not in the design or technology departments, they are in the youthful vigor department - need of the right team to pull off the ongoing project. I'm working on that.

If you or anybody has a pair of CS2s, I hereby offer DIY coaching to hot-rod them. 

Unsound - the CS2 and CS2.2 are conceptually the same product. Same driver sets, same crossover points. I believe whatever drivers we develop for the 2.2 going forward can be retrofitted to the CS2, along with the upgrades being developed for the other models.  I'm looking for a pair.

jazzman - thank you for the news; sad but grounded in the facts of our times. Simon, and by extension the whole Audio Consultants organization, was a dream come true. I believe that if the marketplace were filled with retailers of his caliber, our marketplace would be far healthier and maybe even vibrant.

It may interest this group that when New Thiel assembled its first team including National Sales Manager Steve DeFuria, among his first moves was to re-enlist some core, respected retailers to re-focus Thiel in the marketplace. Audio Consultants was his first approach and acceptance, and they would have been a king-pin if New Thiel had stayed on its stated course of being true to Jim's vision. I don't have any figures, but I suspect that Audio Consultants may have been Thiel's top-selling dealer of all time. 

Hi Guys - a report from the trenches. This week I have received my CS1.5s and 3.5s with EQ plus an Adcom 5300 amp. Thank you forum for providing these units for the hotrod garage! My setup is getting fairly flush, allowing comparison between various products with various sources and amplification changes via throwing a few switches. Also, I'm accumulating measurement data with FuzzMeasure to correlate what I'm hearing with detailed data. And add the decades of reviewer and user experience and your long-term observations and opinions, plus my recollections of how and why various developments occurred . . . the picture is getting much clearer.
Today's fun was hooking up the 3.5s I picked up yesterday from Harry Lavo. What a trip; I haven't heard those in decades. I get it. You 3.5 guys saved me from dismissing them as obsolete due to replacement driver hassles. Rob and l are actively finding optimum drivers to take the 3.5 into the future. That 20 cycle sealed bass is stunning with realism and immediacy. The 3.5 was the last product designed before CNC machining and all-MDF construction. I'll do the measurements, but my stethoscope says the walls are much quieter than the newer products. Their 1-1/8" particleboard is nearly double the stiffness of 1" MDF and the spans are smaller than later bass reflex cabinets plus the internal bracing is end-grain spruce, which is quite effective. CNC manufacturing dictated gravitation toward one material (1"mdf) including shelf braces. This week the EQ and its schematic are going to my tech for upgrade analysis. Anyhow, I'm loving the 3.5s and plan to tackle them after the 1.5s which are simpler as a sandbox. 

George - I am aware of the veil imparted by the EQ, and also the magnificence of the 40Hz and especially the 20Hz bass boost. I am investigating alternatives to the EQ, since it has always been considered a mixed blessing, and is inherently limited in its audiophile values, is un-repairable in its present parts list. I've spoken with Joe who used to make the Golden Flutes for the early Thiel 3s. Hmmm.

Unsound - The wavy drivers are very exciting, and depending on how the future unfolds, I hope to work with them.
Regarding the CS5s, I doubt whether I can help the low impedance, which is baked in with those drivers. A dual input option would sequester the severe bass load to the woofer amp, to free the upper amp. But as you know, that solution requires additional amp investment. What I can address is the two bucket brigade delays for the two midrange drivers which insert 36 components in series with them. That delay can be directly achieved with physical driver offset by re-working the baffle. A veil will be lifted. Given more time in 1989, that 3-D baffle was my preferred solution.  As a historical note, I would have expected a CS5.2 after my departure in the mid 1990s, but that model was not further developed after the driver replacements for the CS5i.

Oblgny - thank you for your thoughts and perspective. The input from you and this group serves as primary guidance along my foggy path. I know that I want to stabilize the brand into the future, and that will take the form of offering owners and upgrade clients a stable service platform, which Rob is now providing, but will need more legs on its stool for ongoing viability.  I know that I want to make selected models better than when new to illustrate the viability of ongoing re-development. I know that I can't accomplish everyting myself, so I hope to inspire some as yet unknown person or team to take up the torch. I began my exploration with the CS2.2s, since I had them since 1990. Then came (5!) more models to compare and contrast. Most recently the 3.5 compared to the 3.6. Obl, I think you're on to something; I feel welcome with the 3.5s. And I am on to some inspiring refinements. This early work will all apply to subsequent models.

It turns out, to my surprise, that as Thiel advanced in time, its sales per model declined. The peak volumes were in my time with the CS2 and CS3.5, followed by the CS2.2 and declining thereafter. It also seems that appreciation and enthusiasm for the brand also may have peaked in the 1990s - I'm still gathering and sifting clues; I'll be visiting Rob in Lexington at the end of November for more insight and direction. 

The truly iconic Thiels surely include the CS2 and CS3-3.5. Their age puts them in most danger of failure and they have ready upgrade potential via multiple avenues, many of which I have been exploring in the HotRodLab. The Renaissance products will not be restorations of the original executions, but rather reimaginings of what the designs hoped to become when they grew up.

Pieces of the puzzle are taking their places, thanks in great part to my interaction with this forum, both here and behind the curtain. The more recent designs will probably be addressed in historic order, or as I gather confidence and abilities. I am particularly enamored with the CS5 due to my personal history developing that product, and that it is the last of Jim's sealed bass designs, and that a reconfiguration of the baffle can eliminate over half of the series elements in the crossover and so forth.

JA, I would not be taking this path without your forum. Here is where I found a focus and so many generous souls who have provided insights, guidance and products for my development work. Thank you all so much.

JA - Happy listening indeed . . . and for the right client in the right environment it isn't very subtle is it? What got us involved in making speakers was stories like yours, except in our beginning times the speaker part of your experience was pretty dim. We loved hearing of young folks, usually shopping for their first 'serious' music system, not knowing much of anything and being dumbstruck that music in the home could be so good. I remember customers who started with a pair of 02s ($200?) and went way up the line over a lifetime of enjoyment. We kept an 'ego file' and some late nights when things were really hard, we would sit in the front office and pull out some love letters describing how they had found Thiel, and what joy we had brought to their lives - and that's really what kept the midnight oil burning.

Jon - I haven't had your experience, but I can comment.Broken leads are usually caused by excessive excursion.Your 2 volts AC at the tweeter and woofer with a 60Hz test signal is improper. Tweeter should be near zero.Caps and resistors usually blow open, not short. Hmmm. Shorted feed cap or resistor would cause direct feed and tweeter destruction.If your XOs are on PCBoards, the short might be in a trace, not component.
Rob Gillum at Coherent Source Service might help troubleshoot

Andy - a couple of points in a very complicated maze. Choice of impedance indirectly includes slopes because slow slopes require less moving mass to avoid self-attenuation on the treble roll-out, and lower inductance for the same reason, and therefore fewer coil winds.

Note that the 1.5 drivers were entirely developed by Thiel in-house; they include under-hung coils, copper shunts, focusing pole and plate geometries, etc. Thiel had a complete driver development lab that was fairly robust by the late 1980s. We developed our drivers and turned the designs over to Vifa for production engineering and supply. We were too small to "own" those designs, so we allowed Vifa to put them in their catalog to recoup their tooling expenses. So, they could be considered "off the shelf", but they got on the shelf by our putting them there.

I have an eccentric theory about why phase-coherent x time coincident speakers are "harder" to drive. I believe that coherence allows our auditory perception to be accepted as "real" and not "reporduced". We take that "real" input much more seriously, and many aspects become sonically important, whereas they were not even heard in the incoherent speaker. I have previously described how I came to form this theory - it is an important leg of my sonic understanding stool. That phase-related amp misbehavior from the amp driving the Thiel low impedance load is obvious, but if driving a high-order low impedance load, I believe it might not be obvious.

There is a real limitation to dynamics in first order filters. On the one hand, there is no time smear or phase rotation in the crossover range for very realistic immediacy, but on the other hand, the drivers, especially the tweeter, do more work, generate more heat, and experience anomalies coupling to the air beyond their 'natural range'. Thiels are actually small-signal machines which do pretty well in normal rooms at normal volumes. But they are not and were not intended to be screamers.

You are on to something. I always notice more edge on higher order filters, which can be perceived as greater snap and slam. We had a drum kit and xylophone in the listening room at the original farm house. I vote that the first order filters sound much more like real drums and percussion. That direct comparison helped form our decision to go first order.

Andy - The Infinite Slope speakers are very good. His approach sequesters the phase misbehaviors to very narrow bands, so they are noticed very little, and the time misbehaviors are managed for least damage.  Previously you noted " how high frequency affects our hearing". The Infinite Slope treble reaches the ear ahead of the rest of the signal, and even though it does so in a well engineered fashion, that pre-arrival is radically different from natural sounds, in which the wavefront arrives simultaneously. The ear-brain is very good at reconstructing those elements into the alleged sound, but that takes work to do. Some folks prefer the wavefront to arrive intact. I am one of those folks, as are most of you on this thread. Thiel took the philosophical position that waveform integrity was important, regardless of whether the majority of listeners assigned importance to it.

By the way, source material, amps and ancillaries that work best with Thiel also keep those relationships straight. The low to no negative feedback designs such as Ayre, address the same goals with the same priorities as Thiel.

Jon - your pic shows PC boards which are all too easy to burn.PM sent

Andy - you are puzzling it out. In the beginning, when we were wrestling with all these variables and making prototypes to compare A with B with C, we had some aha moments that led to the decision to tackle the 'real deal' . . . keeping the phase and time information intact - creating a coherent source. All the other approaches define ways and methods to minimize the importance of the time aspects of the hearing riddle.
Addressing time/phase serves to multiply the burdens of developmen -, it all becomes enormously more complex. Previously on this thread someone (?prof?) mused whether the Thiel Rightness is actually caused by phase coherence or rather the result of all the other details that were paid attention to in the designs. I say it is both. Coherence adds its own complexities and solutions, and it also demands that other details be very right or else they sound very wrong.

Andy - the term of art: Phase Coherence is used in various ways by various people for various purposes. Thiel was a pioneer in Coherence along with Richard Vandersteen and John Dunlavy. We all use the term to mean Phase Coherent and Time Coincident. In other words all the elements of the signal launch and add together to produce the step response of the input signal with no negative pre-shoot at any frequency. Very simple and very difficult!

I’m on my way to visit Rob where we’ll audition lots of models including cs7 vs cs7.2. Rob says concrete vs composite are the same. I’ll form my opinion Wednesday.

Preliminary report from Lexington. Rob is well equipped and stocked with knowledge, experience and parts. A truckload of spare parts was delivered from Audio Consultants - what a score! 
We spent some time comparing CS7 to 7.2. More discussion can follow; don’t hesitate to score a good deal on the original 7s. 

Guys - I am traveling and booked solid. I’ll report early next week.

JA - our time was quite limited, a wreck on the expressway cost us over two hours late arrival. Most of our time covered tech. Rob has original driver and product testing rigs and much deeper stock than I had imagined. He has all the development prototypes from Cs3.5 on. At CSS he has a showroom/office, adequate for the job of comparison. We didn’t make it home where is we’ll hi Fi-Ed. At CSS he has Bryson cubed series. ThielRules might know the gear particulars. TR was there as well as Tammy, long-time driver-maker, and Jeremy Kling, son of Walter Kling, co-originator and Mfgr Director after I left. We all compared 7s to 7.2s and considered them co-equal. Concrete and polymer baffles are equally inert- deader than CS5. Sound is preferential- only. 7 is gentler,7.2is more incisive. We heard equal detail. Jim was working on a 7.3 and I suspect a goal was to mitigate the “in your face” tendency of the 7.2. Best I’ve heard.

Asturias - I am spitballing for thoughts on cs5i vs cs7. Never heard direct comparison. 5i has older drivers, but with similar motor technologies to 7s. 5 has no coax, so vertical loving is greater. Baffles are similar tech, but 7s are thicker and quieter. 5 has big bucket brigade delays on mid ranges. I haven’t seen 7 schematic, so don’t know its timing mechanism. Generally newer models are superior to older. I imagine the 7 specs better.
BUT the 7 has a sealed bass to under 20 Hz which I know to be glorious IF you can drive it. I consider the 5 as a prime upgrade candidate due to direct upgrades with clear advantages. Physical time alignment is low hanging fruit and biamp capability lessens impedance stress.

Correction: I meant the 5 has sealed bass

I see.  Small trace sounds suspect. I've exhausted my clues. I have a PP XO schematic - same drivers, if that might help. Try Rob.

Andy - I'll try to fill in some blanks. And this piece of history may interest some of you.The 3-D baffle is a rather difficult item to produce one-off. The 03 had a flat baffle and the 03a added wool felt around the drivers which worked very well, but seemed somewhat inelegant with our high-WAF furniture-presentation. So I devised the contoured wave-launch baffle which, I believe, was unique at the time. The CS3 (1983) was made by hand, assisted with power tools. First the laminated baffle was beveled on a tablesaw to remove excess waste. 45° across the top and 45° x perhaps 15° up the sides. The CS3 was then shaped with a hand-plane and random orbit sanders to its final contours. The CS3.5 (1986) was contoured with a form tool in an inverted router as follows. The excess waste was removed by tablesaw just like the CS3, then the CS3.5 was machined on an inverted pin router: fixtured in a jig, back-side up, where the jig interfaced with an overhead pin centered on the underslung form tool, which was an 8" diameter arc to shape the rounded edges. That semi-manual method produced the early 3.5s. Later, that forming operation was moved to the Computer Numeric Control Router to the great relief of all. The CS2.2 (1990) used the same tool, but was designed and manufactured entirely for CNC Router.

You might consider the CS3 method of table-sawing the compound angles on the sides and the straight 45° across the top and then shaping with a hand plane, sander and an arc pattern to guide you. Divots or other errors can be filled with body putty (Bondo); and paint covers a multitude of sins.  You might also know a CNC job shop to carve it, probably with a ball-end mill. You then remove the swarf with a sander. Or take the Avalon (etc.) approach and make facets without the smooth arcs. I think I would do that and cover the facets with F11 wool felt to absorb the migrating launch wave. I am presently having great fun with felt on the curves and believe that felt on facets would be a good way ahead.

Have fun. Keep us posted.

Prof - Yes, we assumed grilles to be on, which affects final frequency balance, knocking the upper octaves down perhaps 1dB. Some grille frames are also inherent to diffraction management. Some reviewers and individuals have used them bare and then take pot-shots at that extra 1dB of treble and/or diffraction effects which they have directly caused by mis-use of the product. End of rant. Good to see you here.
Tom

Fabric, even polyester made to be sonically transparent, does have multiple effects, as you say. Many Thiel models use the grille frame to fill the cabinet corner with a rounded continuation of the baffle round-over and as such that frame is an important component of the wave launch. And, as I mentioned, the treble reduction of the fabric is part of the intended balance. But many audiophiles dislike grille fabric. Many have gotten good results by removing the fabric from the frame and using the frame as intended for diffraction control. Pointing the speakers straight ahead puts the listener a little off-axis to reduce high frequency beaming. Thiels are designed for straight-ahead pointing, but it seems a majority toes them in, which puts too much energy in the brightness region. The straight-ahead orientation often requires wall treatment at the first reflection point, which solves many imaging issues, while keeping a flat on-axis and power response.

I hope to hear discussion about sdecker's observations. I have heard neither speaker, but share his thoughts. Bring it on, please.

I don't have full vs 1uF bypass comparisons, but I do have some relevant history. When Thiel developed the bypass configuration, caps were relatively primitive, and smaller value, higher spec bypass caps made a significant improvement. We developed that 1uF tin foil x styrene cap as state of the art and used it for nearly every station. Note that the CS3 had teflon nF double bypasses and the CS2 and 3.5 had styrene ultra bypasses. As caps got better, the ultra bypasses became effectively obsolete. Note that multiple caps help and hurt. They hurt timing precision, since their discharge rate is faster than the larger value, and each station must be tuned. Effectively cost-prohibitive in our particular niche. 

When Jim was developing the SE, Gary and Rob report that they listened to and measured many, many configurations of bypasses and brands of caps including darling  audiophile cost-no-object ones. They chose the then-best Clarity SA for its rightness, both measured and heard. The single (non-bypassed) value was chosen as sonically superior to the bypassed version. Note also that the CSA, with its copper spluttering, is said to be a league ahead of the ESA, which was a relatively small advance over the SA.

Beware that the incision of these higher grade components comes with its own set of potential perils. In Lexington last week I saw Rob's 7.2 XOs in which he had replaced all series feeds with ESAs to compare with stock. He (and his cohorts) preferred the stock parts. The point is that the whole thing is a pot of soup, and "improving" something may require other compensations. When converting Beetle's 2.4s, I was relieved that every upgrade resulted in upgraded performance. As he mentioned, his SE's were late Chinese manufacture with room for improvement by reverting to old methods, from which we also upgraded layouts and coils. All his parts were very carefully selected, and synergy ruled. My own path is with Mills resistors and Clarity caps on old-style point to point boards with new layouts and heat managemnt. I am considering those electronic upgrades as end of project decisions. My focus is on re-bracing and re-baffling which is progressing well.