Thiel Owners

Rob - Thiel's homegrown boards were all masonite. The show boards were masonite sanded and painted to look cool. Asian imported boards were all fiberglas. It is possible there were transition boards, but I've not heard of them.
The 2.7 was made entirely in-house in Lexington under management of the first New Thiel team who sought to return to roots. New Thiel ownership changed leadership 5 times in the 5 years they ran it. Team 2 abandoned Jim's design approach and took all manufacture to Asia.

As far as I know, if yours has a masonite board, it also has components refined through time for best cost-performance.

sdecker - from my perspective, I appreciate your thoughts very much. This thread is my sole involvement with the world of internet opinions. Because I am committed to Thiel speakers as my life-long transducer, and because Thiel performs in somewhat idiocentric ways with some equipment, I find it quite helpful to hear people's opinions about the system performance with various gear. Also, in this on-going community we gradually develop a sense of various contributors' tastes, values and approaches to their sense of rightness and musical engagement.

I don't consider such reviews and commentary in any way improper or off topic. Of course it's not my call to make, but I did think expressing my opinion might encourage discussion around the issue of what is or isn't on or off topic. JA has consistently encouraged broad musical dialogue.

As an aside, my Sony SCD-1 was delivered by a friend today to Bill Thalmann in Virginia for his assessment and possible resurrection. It will spend a few days in his audio ICU and I hope to have a report next week. I'm really fairly in the dark regarding its performance relative to other spinners - I got it because it was Thiel's and had been upgraded and it was, after all, the first of the genre and a heroic engineering effort, especially for 1999. I'll be open to hearing how you guys think it stacks up with more modern gear.

Prof - Ken made the production molds for the CS5 marble baffles from my mahogany master pattern. We had built our purpose-built music room at Thiel Audio and I was able to pass on what we had learned and decided. I haven't heard his system, seen his room, nor seen the video. I'll check that out.

What a nicely made video of a magnificent sound system. Ken extended an open invitation to me to experience his place. But we really were in different worlds. For those of you who know Thiel's history, we were a nose-to-the-grindstone / little engine that could with the principals all typically working 12-15 hours x 6-7 day schedules. Jim and Kathy did more shows than I did, and I returned right after strike to tend to production. Free time was a foggy fiction; it was inconceivable for me to visit Ken's gig, no matter how attractive. But, I didn't realize how far he took it, what a monumental effort.

I note Vance Dickason, speaker designer and author, in the credits. That's serious design chops. His crossovers would be informative.
And there's the matter of cost. Imagine the price tag on that rig! That illuminates a piece of Thiel history. Ken ran a commercial/industrial mold-making operation. He claims he made our patterns, masters and molds for the CS5i baffle. Mostly true. He quoted them, which induced serious sticker shock in yours' truly. I opted to refurbish the pattern that I had made for the initial baffle runs via a local Lexington marble shop - patterns are somewhat damaged in the mastering process. We pushed and tugged awhile, and I ended up making the quoted $20K pattern for a couple days' work in my project shop and original CNC. Similarly the production molds - we had made two in the original marble shop. Ken's price was unrealistically high, and I opted to upgrade our originals myself (in my bootstrap style.) That fact cut Ken's price in half and added a highest-quality renewable wear-coat. He still made plenty on the job. His work was exemplary of a highly skilled craftsman. And his prices were wrangled into the zone of accessibility due to competition with Thiel's vertical integration. Such is the nature of small business with options to self-supply technical elements. 

johnhh - I concur with Scott and can add some tangentially-related personal experience. Not Sansui, but Yamaha of the same period. Our non-profit performance venue here in NH was given a Yamaha P2200 @250 w/c and the workhorse darling of its mid-70s era. I wanted it to somehow become a great sound-quality amp for our Thiel 1.5 small venue system. We had caps upgraded, replaced fuses, modified its output impedance, and generally souped it up as best we could. It sounded fine . . . until I substituted my Classé DR-9s or even the Adcom GFA-555II or 5300. These hi-fi amps 20-25 years more recent and designed with sound quality as a goal were in a far higher league.

I'll add that my experience with Japanese design(ers) of the 70s oozed conventionality. Japanese culture does not permit innovation. If any design element could be scrutinized, even by a high-school textbook, then the designer would risk his reputation by promoting it. I have some actual stories of Japanese companies buying avant-garde American designs and then subverting/ diluting them until they were "normalized", and thereby losing their specialness. It's no accident that Japan did not break into the high end in the 20th century.  

John - there's a parallel phenomenon happening in the pro world around tape recorders. A great recorder can be a holy grail for a recordist, especially one with analog vinyl as their target playback medium. Recordists post their epitome tapes for comparison against modern signal chains, often to significant embarrassment of sloppy performance. In 1980 the tape machine might have been king. But it can really suffer in comparison to best of form contemporary technologies.

Some folks here have studied integrated amps and have good advice to give.

Dan - that is quite a system you’ve got going. I have no direct experience with your prospective amp models, but note that they all look great for current delivery and the brands pass the sniff test. But you knew that. Good work.
I have two general thoughts: power supply and the 1.7. I’ll address system power later.

I have observed that the CS1.7 may not be as advertised. Jim had his CS 1.7 nearly ready to go when he died, which then simmered behind the 2.7 in the last few years of Original Thiel outsourced product development. New Thiel changed the 1.7 design significantly, replacing Jim’s first order crossover with "high order" (probably 4th), and introducing a 2db/decade slope down from bass to treble - a standard PSB-type crossover. I also read that the tweeter sports a "new voice coil and re-designed motor". That sounds suspiciously like a normal overhung tweeter, again standard industry fare available from suppliers, whereas Thiel’s underhung drivers were made in-house. The underhung motor exhibits an order of magnitude less onset transient distortion and a completely symmetrical push-pull linearity - albeit at the expense of greater manufacturing cost, lower efficiency requiring a larger magnet, and greater hazard of burn-out due to voice coil geometry. I don’t know whether the woofer also opted for a "normal" motor rather than Jim’s design specialties. I have queried Rob for further insight. In the past he said to me of the 1.7 "It’s not Jim’s work".
The 1.7 caused a parting of ways for the first team of New Thiel which included Steve DeFuria, a long-time industry insider who had collaborated with Thiel over its entire life-span. Steve objected to New Thiel calling the 1.7 a "Coherent Source", the application of a house curve, and allowing reviewers to represent it as a first order, phase coherent transducer. Notice that Stereophile among others never reviewed the 1.7 - I’ve been told that they didn’t want to play New Thiel’s game. I am not disparaging the performance of the 1.7. I am told that by conventional standards applied to normal speakers that they have been called "more refined than previous Thiels".
Do any of you here in Thiel-land know how the CS1.7’s step response looks?

To correct the record: Rob says the 1.6 and 1.7 use the same moving system in the tweeter and the only difference is that the 1.7 has an additional bucking magnet.

Dan - following up on the power part of your amp query. It has always fascinated me that Thiel speakers seem to bring out the worst in many amps. This weekend I interacted with a very knowledgeable long-term Thiel user who stated he had amps that worked with every other speaker, except his Thiel CS2.4s. That rings a bell - it echoes the experience of thousands of users, reviewers, fans and detractors. Why is that? Thiel impedance is low, but it is also highly resistive, which amps love. There are speakers with far more brutal load characteristics, which don’t have Thiel's ’hard to drive’ reputation. Go figure.

I have been figuring for decades and have an idea that makes sense to me. Back to some context. I was the set-up guy for many shows. Most of that electricity sucked, and sometimes we even changed rooms to be away from elevators, etc. with very dirty power. Pin that. In the late 80s Thiel built a dedicated music room of some renown with lots of care to proportions, materials, treatment, etc. It had fluorescent lights, air-conditioning and office computers all on the same service, since its footprint was in a 40’ wide industrial space that all shared a single service. Even though we isolated the hi-fi gear on one leg, there were still many reactivities on the line. Our electrical engineer said that the problem was intractable - just live with it. Long story ensues. The solution came via borrowing hi-fi circuits from the next bay, which was used almost exclusively for finished goods storage. That power was cleaner and we hung a sub panel with short runs to the hi-fi system. The system performance was improved more than normal people would imagine.

I am presently studying the mechanisms that might cause such mischief. My working hypothesis is that various design(er)s manage the various line distortions more or less effectively than others. Those design(er)s who understand the effects of dirty power can minimize the problems within their chassis. It seems a major component of the problem includes the differences between the high (+) leg and the neutral leg at or near ground potential; but it rarely actually is at zero due to losses and interactions in the feed wires.

I have always been puzzled by how some / most equipment sounds so much better when run from 240 volts rather than 120. One explanation is that 240 volt is a real balanced feed with each leg running 180° out of polarity and the ground centered between them. Circuits see symmetrical power. Some snooping has led me to investigate recording studio power. High end venues often mix, master, etc. on the graveyard shift. Audibly better recordings result. As I rebuild my studio in a new location, I intend to run a sub-panel with both 240 volt and 120 volt outlets. Performance can be easily compared by swapping the IEC cord for 120 or 240 volt feeds. Can’t wait to learn.
A seemingly optimal 120 volt solution is to use balanced power as is used in many labs and studios. An isolation transformer is tapped for 60 volts plus and minus rather than the usual 120+ and zero. I like the concept, have heard good reports, and would like to have that solution in my bailiwick - but it seems quite expensive.
One bottom line from these musings is that buying an amp that can be tapped for 240 balanced power input might give you a leg up in bypassing a lot of power problems and getting good sound more reliably. These explorations feel right to me; they address many unsolved mysteries about why or how Thiel can sound bad when other speakers sound good.

The other half of this equation is my personal belief that Phase Coherent transducers invite the audio cortex to scrutinize sound as though it is real, rather than an electronic facsimile. Our felt judgement is far more acute for ’real’ sounds than manufactured ones. My first exposure to this phenomenon was in developing the 03 in 1978 - we wrestled with this phenomenon for a year and a half and learned that we had to "fix" problems for coherent presentation that weren’t even on the radar when a higher order crossover was substituted. I know I’ve addressed this idea before, but I’m repeating for those new to the forum or to raise some issues that might connect some dots for folks beyond myself.

James - about your perceived distortion. I'm guessing that a recording that is good enough to consider as a reference does not have the type of distortion you are reporting. If it emanates from one channel, swap channels. If it swaps, it's somewhere upstream. Of course I'm guessing, but it sounds like a loose tweeter voice coil. Do you know Rob at Coherent Service? He can help troubleshoot as well as rebuild the driver. 

naimfan - Some thoughts on the CS2.2 (where I’ve reinserted the rightful decimal point). The model 2 was a smaller-scale version of the original 3-way model 3 - less expensive for smaller rooms and moods. The model two often benefited from trickle-down tech. The 2.2 got the UltraTweeter developed for our flagship CS5 in the late 80s. Pretty good tweeter. It also got some driver motor technology including custom motor machining and copper coil and pole-piece shunts, new technology at any price at that time. The CS2.2 was Thiel’s first passive radiator which went on into all future floorstanders. On the manufacturing front, the CS2.2 followed the CS5 for which we got our feet wet with all CNC machining and serious Finite Element Analysis of the drivers. It got our first open lattice interior cabinet bracing and first acoustically frameless grille. It’s a very sophisticated product, especially considering its modest $2K class. (In fact many dealers and reviewers told us it would make more market sense at twice its price.)
As often mentioned: "they’re not for rockers", and so forth, due to bass output limitations. I have an update for that. The tweeter handles the load, the midrange does quite well, its low end being crossed over an octave higher than the model 3. By the way, Rob has a good drop-in replacement for the midrange. The woofer is only 8", but it’s built to boogie. The main limit is a hard splat of the passive radiator on bass transients. At least that’s what early reviews (John Atkinson in Stereophile) said and we all acquiesced. I never settled with that idea. I had developed that passive radiator which has only a foam plate and two compliant surrounds that glue directly to the front and back of a lip in the baffle. In other words there is nothing for anything to splat against. Hmmm.As was our practice, Jim was headlong into developing the CS3.6 and didn’t look back at the 2’s bass problem.
No secret that I’ve been working on updating these classic Thiel speakers for the past few years. My explorations led (via a long spiral route) to outboarding the crossovers while eliminating all magnetic elements (such as mounting screws), leaving only the components themselves on a masonite board; and paying closer attention to stray fields, wire routing and dressing and so forth. Guess what? No splat. I can’t overload them short of cringing. I can’t say I know exactly what was causing the problem except something in the crossover feeding the woofer and possibly midrange. I know the problem was there in spades and now it is reliably gone. And by the way, that big, wooly quality of the bass. Gone too. The CS2.2 can’t move as much air as the CS3 models, but it’s no slouch.
So, if you have a chance to get a pair and if you’re interested in making them substantially better in the future, there has been lots of work done toward that end. It’s hard to imagine you being disappointed.

Tom - your hypothesis is of interest. One observation is that all Thiel speakers mount their XOs to the bottom or back near the bottom with lots of fiberglass between the XO and the drivers / moving air. Plus the sound is hard, like a driver bottoming. When I put the redeveloped XO inside the cabinet, mounted on nylon studs and rubber feet, the problem had vanished. My best guess is some sort of electromagnetic saturation and/or possibly the bottom end of the midrange bottoming out. (But I didn't change the midrange driver, only some specifics of the crossovers.)
As I've mentioned before, getting the XO out of the cabinet does worlds of good. My in-cabinet solution is different from yours - I am suspending each of the 3 small crossover boards in free air, rather than clamping it the the wall where it must absorb the cabinet wall vibration.

Input accepted. What I know for now is that this rubber-damped / no cabinet contact method performs considerably better than stock Thiel. And that putting a crossover inside a cabinet is an inherently compromised idea. Further improvements always await.

The CS5(&I) was the exception. It was too big for that. That huge xo board slid into a pair of grooves near the back of the cabinet, abutting a rubber bumper at the top. The cabinet base with a rubber bumper was screwed up against the bottom of the board. I think the long edges were just a light friction fit. Long ago indeed. 1988 introduction = 33 years ago.

Probably more than a third of those components were time correction for each the upper and lower midrange drivers. Bucket brigade analog time delay to fine-tune the arrival time. Those ubiquitous yellow 1uF styrene film x tin foil caps were developed for the CS5 and persisted to the end in various forms.

Prof - at your instigation, I’ve spent a lot of time with the 02 during this past year. It is small, shipable, simple and inexpensive. It provides a good development platform and test bed for many ideas which can be inexpensively iterated multiple times. I presently have 6 units for comparison and beta use and would buy more if they came along.

I don’t yet have a way to measure impedance. The woofer is 6 ohms and the tweeter is 8. The Stereo Magazine review states the bass double peak at 18 and 17.5 ohms (quite mild) with the saddle at 5.4 ohms and absolute minimum of 4.75 ohms at 170 and 10kHz - averaging 6 ohms, which is what we called it. Those loads are mild and easy to drive compared with modern Thiels. Several of us here agree that a minimum of 4 ohms would make Thiel products so much friendlier to such a wider range of users.

That said, the 02 has no Zobel or other corrective networks, so it presents a less resistive load than modern Thiels. More reactivity is harder to drive. The crossover is a straight-forward second order x polarity correct design. So, the tweeter signal arrives before the woofer, but the soft diaphragms allow a sweet, seamless crosspoint transition. The Gefco woofer has a proprietary pulp cone which is quite exceptionally smooth over its entire range, gently rolling off up to fully 10kHz. The Peerless fabric dome (we called it silk, but I don’t know) was the darling of the day, used by Polk and nearly everyone else. Nice, smooth tweeter with no real misbehavior. All that said, the drivers use conventional motors, coils and materials, unlike later Thiels that reduced distortion by more than a factor of 10x. Similarly, the cabinet is unbraced albeit veneered both sides on industrial particleboard for greater stiffness than MDF. And the crossover components preceded audiophile components. We self-wound the unpotted coils from standard CDA110-ETP wire. Resistors and caps were "better" grade parts bin quality and hookup wire was standard, stranded in PVC or vinyl. In other words, nothing special except attention to simplicity and longevity.
My work this year would take an essay to summarize, but let’s say that every change I made created a clearer, tighter, more Thiel-like presentation. Potting the coils in varnish - then replacing with CDA101, solid CDA101 wire in teflon, etc. etc. etc. But nonetheless, the original stock speaker has an easy, friendly presentation, like coming home to love. I hear it, I get it and I even have developed some theories about whys and wherefores. One summary thought is that the stock 02 sounds like what we (especially from our long-ago youth) expect to hear. Audio neurology is very synthetic and wired to clues, cues and expectations far more than we generally realize.
As a reference, my lab mule has migrated from stock 02 to hot-rodded 02 with original drivers > time aligned at second order. (That is the SCS format.) Then I substituted CS.5 drivers which are stellar.The woofer is cellulose / wool pulp with all the Thiel motor juju. I compared first vs second order crossovers executed (eventually) with CS7 style components, and landed firmly on the first order x time-aligned configuration. The workhorse value of this hot-rodded version is far superior. It no longer has that ’old shoe’ familiarity that you appreciate, but it shines with joy. The present iteration has rounded (1.5" radius) grille edges, braced cabinet, and lots of new technologies. Since it's not coincident, its coherence varies with ear position. I made an easily adjustable stand for seated through standing tweeter height. And the fun never ends.

The form-factor is an original 02, the execution incorporates some beyond-Jim elements. But no coaxes or x.7 style drivers. Presently I'm looking for new lab space while liquidating my tonewood enterprise.

Naimfan - you can use double wires (star quad, etc.) and connect both runs to the same terminals. 4' from wall to tweeter is a good distance if you can listen at least 8' away from the speakers.

Dan - use the 14-4. There is some disagreement around best configuration for star quad. Conventional configuration is opposites as plus and other opposites as minus. Also try adjacents as plus and other adjacents as minus. Please report back what you hear.

I hope you jest re 1.7s. Just reporting back-story.

Prof - I get it. One of my pairs ended up in the living room of the house I'm sharing, driven by PS Audio's Sprout II from a decent CD player. They consistently delight and make me wonder at times if and why most folks would need more to enjoy music. They were retired in 1984 by the SCS and its successors. For all the technical improvements, new solutions, advanced materials of the SCS4 - the 02 really holds its own.

One of these days I'll get you a pair of Renaissance 02s for your critique.

For those interested, the original 03 EQ was similar to the 01 as a fixed point boost at 30Hz, enabling the speaker's -3dB point at around 27Hz while maintaining 90dB sensitivity at 8 ohms (6 ohms minimum). Those early EQs had less sophisticated circuitry and implementation, with electrolytic caps in the signal path and so forth - creating a transistory veil on the music. Each generation got better with the CS3.5 becoming direct-coupled.

Jeff - you're not alone in preferring the direct path and sacrificing the deep bass for more transparency. You might consider the CS3.5 in your replacement search. The EQ is getting upgraded well above stock and the speaker is in line for significant driver and crossover upgrades.

All - this might be an opportunity for me to expand on unsound’s response to dmac67 regarding the ’improved’ CS3.6s. Indeed Jim’s approach to design was to specifically control any system misbehavior in pursuit of absolute fidelity to the input signal. In the case of drivers and their crossovers, all those parallel XO circuits serve to correct and control any anomalies and resonances so the drivers can span their 7 to 8 octaves in net 6dB/octave atentuation as required for true phase-coherent response. Additionally he sought for an absolutely neutral tonal balance (as he interpreted the term). So, any change in any driver would require reciprocal crossover changes to keep the outputs loyal to his design intentions.

Let’s unpack the "Rosewood 3.6 offering".
"Rosewood" as used by Thiel is the wood of commerce in the Dalbergia genus, specifically Dalbergia nigra, Brazilian Rosewood, considered the most classic and valuable of the lot. The offered speakers are what Thiel called Morado, which is the lowland Bolivian variety of the genus Machaereum (Amberwood being the upland variety.) It’s a beautiful wood, many people refer to it as a ’near Rosewood’, etc. but it is not the upcharged wood that Thiel calls Rosewood.
The "rebuilt crossovers" is of interest to me. It would be of great interest to learn what was done to those crossovers to compensate for the different midrange and modified tweeter. Seriously, if a schematic could be gotten, I’d love to see it. In Thiel-land the final months of critical listening would be spent fine-tuning the XO to prune performance of the specific drivers in their cabinet toward the design goals. These drivers are different. 

The ’original tweeter, but with soft domes’ are "better sounding" than Thiel’s aluminum domes.That is a matter of opinion and will be true for some listeners. However, the differences in the soft diaphragm will include mass, compliance, breakup and resonance behaviors. If the seller compensated for those differences, I’d love to hear and study that work. If not, the output will vary from the design intent. A historical note re that tweeter. It was developed for the CS5 and trickled down to the 2.2 and 3.6. It was a massive joint undertaking between Thiel and Berger, the lead development engineer at Vifa. It incorporates the underhung motor and copper shunts that became Thiel trademarks. Copper shunts for the voice coil and pole piece were not in use in the late 80s when Jim discovered their efficacy through Finite Element Analysis. We investigated patenting the solution to learn that Faraday had discovered it, and it was in the public domain, although not in use for loudspeakers at that time. Note that other manufacturers call their rings "Faraday Rings", but Jim is who introduced them to audio.

Let’s look at the soft vs aluminum dome issue in terms of our relationship with Vifa, our long-term development partner until we outgrew their tolerance for our eccentric requirements. We co-developed many drivers with them from about 1980 to the late 1990s when we gradually took driver manufacture in house; not because we wanted the extra work of making drivers, but because our solutions were too difficult, expensive and niche for them. With this UltraTweeter as an example, we co-developed drivers with Vifa who retained rights to sell them (and all our joint drivers) to any and all their customers; how they amortized their tooling and development costs. Berger implemented and sampled Jim’s aluminum dome and simultaneously the soft dome for a broader customer base. Both are in the same structure so that the Thiele/Small parameters will be very similar. But the soft dome will have individual variation far exceeding the aluminum. That’s the principal reason Jim chose aluminum, so that his optimized crossover could apply to all drivers, rather than some lesser percentage using any soft diaphragm with their greater divergence from the norm. That story goes on and on, but I’ll take pity in this late hour. Either UltraTweeter will sound "good", but the aluminum one will consistently produce the designer’s intention.

The midrange gets far sticker because the chosen replacement uses a different motor and cone. It may very well be an excellent driver, but the 10mm voice coil length suggests a normal overhung motor, plus most all the Thiele/Small parameters also change. So an XO redesign would be required to achieve Jim’s level of accuracy. Again, I’d love to see the XO schematic to see what this designer has done.

The woofer remains stock.They are bullet-proof and rarely fail. I wonder what XO upgrades, if any, the seller applied there.

None of what I have written here is to disparage the seller or the potential care and expertise applied. However, I do point out the large undertaking, fraught with pitfalls in arriving at a speaker similar to Jim’s product. I suspect that every person on this thread would pick it blind as being a different speaker from the original CS3.6.

Mac - thanks for this opportunity to expound on what makes a Thiel a Thiel. If you choose to get this pair, I would love to work with you to document their particulars as they relate to my growing stable of upgrade ideas and directions for classic Thiel models.
Cheers,Tom

JA - my interest is academic. I want to know who made what changes for what purposes with what outcomes and what seems "better" for what reasons.

Since namifan asked about screws, I'll peek into the rabbit hole with you.My EMI scans show that the steel screws are active in the driver fields. Early hypothesis is that they matter. I have replaced all screws in my working prototypes with brass. Fiberglas would be better, but unavailable in small quantities at affordable prices. I invite any of you to swap your screws for brass and post what you learn.

Tom - your missive runs parallel to my experience. In the case of my outboard crossovers, and mounting my internal crossovers and back-plates - I am using all nylon. Drivers need more strength so I use brass.

naimfan - stripping is an issue with screws directly into MDF. I use "Wood Hardener" from Minwax available at the hardware store, or you can use thin super glue dribbled into the hole since the internal threads are already formed via previous screws.
Take care with brass screws; they're not as strong as steel.

Thiel migrated away from direct wood fastening to inserts sometime after I left. There are advantages of reliability over multiple uses. However, metal on metal under vibration will (reliably?) vibrate loose. Breakaway Locktite helps as does Mortite onto the threads. But over the years, inserted fasteners seem to come loose and direct to wood fasteners seem to remain tight. I'm going with plan A and use direct fastening. Of course castings need inserts. Brass is a good solution, or other non-magnetic substances - non-conductive is even better, since electrical fields are affected by any conductive material, magnetic or not.

Hi guys - an update from the hotrod garage. Recently I was lent a pair of CS2.4s for experimentation. Initial auditioning was disappointing, but it turned out that one coax was dead and the other rubbing. Rob rebuilt them, and yesterday I reinstalled and listened with John Pries in his fairly good room with fairly modest equipment. The room is decently proportioned, but has full glass on both short walls. Speakers are centered on the long wall with listening position about 8’ back and 3’ from the back wall, with an open door to the attic stair. Not bad. The source is Tidal hi-def to a Prima Luna 100 integrated @ 40w/channel and bluejeansish cable.

John had started with Klipsch (copper skinned) monitors, then 1.6s, which is what we compared with the 2.4s. The 2.4s are stock @3729-30, mid run with unmarked caps, decently wound coils on glass boards. Tidy work and better looking than Beetlemania’s later Chinese units. All the caps with design and early implementation as PolyPropylenes (Solen) were here executed as PolyEster (MKT), and the coils were not as tight as original, both a step down.

Impressions included lack of bass control and considerable bloom (small tube amp) and a slight upper midrange scoop. Otherwise, the detail, harmonic structure and resolution skunked the 1.6s. The treble was delightful. From out in the garden two stories down his wife Barbara announced "that’s a lot better, more realistic". I was surprised at the improvement over the 1.6s, which I had thought held their own pretty well.
We know from Beetlemania’s and a couple others’ work that there is room for improvement. First pass will be some physical things like my (for now) proprietary wire, thermal management, laminar launch baffle treatment and non-ferrous fasteners. These changes can be made without my listening lab, which has not yet re-materialized. I'll wait for XO work until I have some measuring capability around me.

I am pleased and grateful to have these gems for an extended time, to develop and refine some upgrade mechanisms.

Enjoy the music.

Tom - I would be the thermal culprit. I think thermal management is more important than generally believed. I am systematically addressing the mantra of shortcomings of Thiel speakers over four decades. Among those complaints is "they’re fine for vocal or folk or light jazz, but don’t try to rock with them" and variations thereof. It’s true, our development regime included primarily those mild forms of music. Of course we did what we could to manage heat if it didn’t cost much. When we began using aluminum cones, we went to aluminum voice coil formers. Those coupled cones could easily get hot enough to vaporize spit. The 01 and 03 were sealed boxes and we routinely got burned-out voice coils and resistors and melted caps if too close to a resistor. (Aluminum VCs introduce eddy currents and had to be abandoned.) The stories go on.

A big aha was when I got some Thiel subs. Nice subs. One solution is a thermistor on the voice coil to feed more power to overcome the higher resistance of a hot VC. Hmmm, not sustainable. And on and on. I want to dissipate, not overcome.

Now, consider the excruciating detail of Jim’s reciprocal circuits to correct driver behavior, control impedance rise and smooth out resonances. Those models are temperature critical. And those selected, ideal temperatures rarely exist in real life. A cold speaker will not sound good nor measure well. A hot speaker’s measurements are far outside of acceptability, and sound bad. In pro speakers many strategies are employed to keep the temp right. Ever see the fans blowing on the ribbed magnets of stadium drivers? Ever see one whose fan fails? You get the picture.
So, I have put thermal performance way up the list of necessary solutions. I’ve developed some techniques that stabilize temperatures while doing no harm. Regarding parallel resistors, sometimes yes. A modeling program or a thermometer or your finger can identify which resistors would benefit. Also, coils generate heat. Stand them on standoffs and consider air-flow around them. (They get hot enough to melt the varnish holding them together.)

Regarding boundary turbulence, my aha moment was hearing a particular harshness when I got some CS3.6s. We’ve addressed it here before. It turned out to be boundary layer perturbation, primarily on the flat areas of the baffle, but also on the cones and surrounds. By definition there is no motion at the fluid (air) / solid interface, and laminar flow increases as a function of distance, depending on properties of the fluid. Air isn’t linear per amplitude. Nano models built for aerospace research conclude the boundary layer to act essentially chaotic, like whitewater rapids or worse. My approach includes study of theory and extensive experimentation. My goal is to manage the propagation from the high-impedance sheer (source) wave to the uniform low-impedance transmission medium of the room air. I am using micro scale textures, some fixed and some mobile. Douglas Pauley has patented some solutions which we are jointly developing via our 02 workhorses. The resultant ease, naturalness and articulation are improved to a surprising degree.

Your paint is probably addressing the same phenomena. In early Thiel days we discovered that a particular light spatter coat on the baffle sounded better than flat-smooth. I see that these present 2008ish 2.4s still have that spatter pattern on the flat baffle, but are smooth on the curved edges. My speculation is that smooth passed marketing with higher marks, and the curved sections are far less problematic than flats. I like my results with cellulose micro flocking on those curves.

Your room treatments and active laminar flow device sound intriguing. For now my situation constrains me to concentrating on the speaker itself. Someday I hope to be apprised of what you’re doing and how I might get my mind around it.

The question always comes up whether enough listeners care enough to support these exotic tweaks. I don’t know, but it sure is fun learning about how things work. Einstein said something like "relativity is easy, fluid dynamics is hard." We’re wading in fluid dynamics.

thoft - I'm not an electrician, but do have experience with circuitry and ground loops. I highly recommend you fix your house grounding, not only for the amps but for safety and right functioning of all your gizmos. Electric circuits routinely bleed to ground as part of their design; a significant part of the designer's art revolves around grounding. Additionally, you may be in breach of code and insurance requirements.

Rob - It is likely that the two coils have the same inductance and simply vary in their aspect ratios - a minor difference. Speculation is risky.Using the coil wire (like yours), and not a jumper wire, is best practice.

Tom - that guy knows his wire, and consider his background at Belden. I'd love to hear his wire.
He reminds me of my first aha speaker cable experience at CES around 1978. (This post repeats some earlier references, but 200+ pages deep and some new folks might like hearing . . . ) Kimber and Thiel were fairly new companies, before "wire" had become a component. We were using 00 welding cable for speaker wire. A bunch of serious guys had congregated in our CES exhibit room the night before opening. Ray had this braided stuff to submit to critique. Its sell price would be around $1000/pair foot due to all its exoticacies. (in 2021 dollars, about $3700/ pair foot.) Wow! the sonic upgrade was stunning to everyone in the room. We exhibited with it, and took some flack from reviewers and dealers for using $30K speaker cables with our 03 being introduced at around $900/pair. That experience focused our attitude that each upstream component bears its own responsibilities for signal purity. We generally fed our speakers a much more expensive signal than norms would suggest.

Does anyone here use braided flat wire? either Kimber or others? I know it's extremely expensive to make well, and I'd like to hear if anyone is succeeding with it.
TT

Tom - that's the real deal. There's always plenty more to learn than most of us can absorb. That's beyond my scope, but if you dive in, I'd love to hear your summaries as time might allow.
TT

Unsound - I don't know much about particular products. What little I do know says that for braided wire to work as designed, it must stay extremely quiet internally, and that's no small task. Straightwire's Bflat (I believe) preceded Wireworld's entry and was abandoned due to those difficulties (near impossibilities) of keeping flexible strands functionally rigid. Kimber filled the gaps with silicone which had to be applied to each strand and then rolled to form a filled web with no thick or thin spots. Sounds tedious.
I know that Jim ended up with Goertz flat wire (late 2000s) as his best, affordable and reliable solution. Please tell me more about Alpha-Core's approach.

Tom - I prefer to consider the room as a separate entity due mostly to how our aural neurology manages the inputs. Signals less than 5 to 10 milliseconds are conflated into a single onset transient. The longer that initial signal emanates, the more slurred the transient sounds. Longer delays are clearly understood as reflections and the geometry and contents of the room are associated with those reflections and carry relatively low weight in synthesizing the sonic event.

Rooms contribute much more than most people give credit. Room acoustics and treatment brings far more value than most folks think. But, the speaker-maker must draw his line somewhere. In the early days, we spent lots of effort deciding between directional, omni, bi or di pole radiation, etc. We landed on a broad polar pattern that mimics (fairly well) how a real singer or acoustic instrument radiates into the room. Thiel's relatively omni-directionality makes nearby room objects more important than many other polar patterns. But we believe it produces the most natural presentation. YMMV. So, for Thiel, the first side-wall or ceiling/floor reflection is more important than with a pro speaker that limits dispersion to 120°. Beyond that, I have found that diffusion techniques solve lots of problems without the down-sides of absorption.

I don't know how much laminar flow would help the room because of the low energy of each of the wavefronts that meet room objects. At the speaker, especially at the driver / source, the wavefront energy approaches the sheer strength of air - so propagation management matters a lot there, but less-so the farther it gets from the source. At least that's my layman's understanding of the territory.

Stefano - Here are some thoughts from having owned, listened to, appreciated and messed around with CS3.5s for decades. You are correct that anything you do will have sonic consequences; and also that some of those consequences will take you closer to the speakers’ original performance. Regarding the tweeters, unless you are hearing problems, I recommend you follow the advice to keep the 28/2s as backups. If you swap, see if Rob at Coherent Source Service can renew your ferro-fluid in your originals to save them as backups. Those VersaTronics caps are high performance, long-life caps. I don’t know of a single failure. However, 40 years is considered their estimated service life, and you are getting close. Some of those electrolytics are in signal paths where their failure would wipe out their driver; so I would replace those for safety. Rob or I can coach you; A’gon disallows sending schematics, etc. Note the 3.5 was the last product with the ultra-bypasses - styrene .015uF around PP 1uF. Great caps, keep them. Also keep your hookup wire.

Caps: better caps exist today and caps are an expensive upgrade. Your biggest bang / buck is to swap the 8uF tweeter feed for a ClarityCap CSA or PUR. These caps will not alter the ’house sound’ whereas other brands will.

Resistors: Jim developed those non-inductive ceramic resistors and they’re better than normal sand-casts. At the time we considered better resistors, but budget prevented their inclusion. I highly recommend swapping at least the series resistors, especially in the midrange and tweeter with Mills MRA-12s. Pretty short money, same circuit performance, sweeter sonics.

Binding posts: If your plastic-cap binding posts work, keep them. They are better than later big, brass posts which were Kathy’s capitulation to market perceptions.

Note: XO values were weaked end of 1987, you want the revision. What are your serial numbers?

Grille frames: This suggestion is just that, offered for general understanding. Those frames cause diffraction, but the fabric was considered in final voicing. IF your room is well damped (soft stuff, especially at wall reflection points), the difference in frequency response is often OK when bare. In that case, best performance is to create a grille frame that functionally fills the baffle edge voids, but eliminates the outer frame members. Conceptually, the new frame would keep the base perimeter and chop off the aerial parts. The long side struts would be rounded over to finish the curve of the baffle. Affix in place with Mortite or BluTac. You improve the anti-diffractive base function and eliminate the diffractive aerial elements.

Equalizer: The equalized bass was fundamental to Jim’s vision - it was abandoned due to market forces. It produces more integrated, better performing bass than the later reflex system. The EQ can be substantially improved (I am close to an available product.) Resistors replaced with metal film, Transistors replaced with lower noise, higher performance, maintainable versions. Caps upgraded as appropriate. Power supply redesigned as regulated rather than present unregulated circuit. Original all-discrete, Star Darlington, direct-coupled design remains. All in, big step up.

There are other hot-rod tweaks which we can discussed via PM if you wish. I posted all this detail for all you who might have been wondering.

Tom - can you tell us more about your Sonoran Plateau wire?
For the record, in his final products, Jim threaded the driver lead wires through silicon tubes suspended between the braces. Can anyone tell me if those tubes are affixed to the cabinet walls, or are they stretched in space? My proposed wire augments Jim's 18-2 solid in hard teflon twisted pair with an additional pair of 18-2 stranded in foamed teflon. The 4-wire bundle adds both mechanical and electrical damping for a cleaner signal at the driver. I'm still invistigating flat wire and other ideas. Again Tom, can you talk about Sonoran?

I remember here awhile back that some of us messed with parallel feeds, like you are describing. Seems all agreed it sounded somehow "nicer" and "more alluring". Seems that at least 2" and preferably more separation is needed to minimize capacative coupling. Such a jungle!

FWIW: That's real Brazilian Rosewood.

JA - it's still in the queue. Either a long queue or slow going.

Please remind me which Class´e amps replaced which Adcom amps.

JA - it seems that the SCD-1 is old business and I allowed him to back-burner it since I'm not in urgent need. He is also going to look at possible upgrades for my classic Classé DR6 and DR9s. I hope to pick up the Sony later this summer, if he can fix it.

markmike - please post the pix and state your model and serial numbers. I might provide some clues.

JA - I'll post a report after my visit there late August.

Agreed - except Thiel's custom 1uF bypasses were beyond polypropylene. They were tin foil x styrene film, which is close to teflon in performance. All the CYC clones were a level or two down from classic Thiel standards.

Beetle - one of these days I'll open up a CYC 1uF bypass on your old board to see if I can tell what it's actually made of. 

Perhaps someone on this forum knows the subtleties of polyester capacitor markings. MPT and MKT show up on lists as metalized polyester. I don't know anymore about it.

Yes - all Lexington boards are masonite, including all the CS2.7s. I speculate that the masonite board contributes to the reported sonic advantages of the 2.7. There is a bit of mystery regarding how the 2.7 can sound so good with that huge 400uF electrolytic cap (albeit bypassed) in the midrange series feed.

Beetle -  good memory. Those 1uF bypasses had been styrene / tin foil. Some sleuthing would be required to find out if, when and why they became PP - and on which models.

marqmike - thanks for the photo. I'll chip in later.

I'll make some general comments from my study over these past few years. I wasn't at Thiel during the transition to China, but have learned some from first hand operatives. Here on this form there was surprise and disappointment that their Thiels weren't wholly "made in USA". I shared that surprise. In my time we made everything - unusually vertically integrated for a small company, using best of form american and european parts. We also valued affordability - we wanted our products to be affordable to folks like us. Those elements contain conflict, especially as the world marketplace gravitated toward lower cost elements from far-away places. As time went on, increasingly greater part count could no longer be sourced in the USA. Add to that fact, Jim had cancer for at least 5 years before his 2009 death. During that period the company outsourced more and more to simplify its operations to attract a potential buyer to carry on the company.

Evidence shows that during that time outsourcing crossovers went to China, and it also went to lower-spec components. Rob says that Jim tested and approved all changes and Kathy's marketing opinion was that performance wasn't degraded to affect the market niche. But nonetheless downgrades were brought into the mix.
For example, the 2.4 original specs include all polypropylene caps with polystyrene bypasses, built on point to point masonite boards. Beetlemania's SEs were late production FST/China with sloppy coils, polyester caps, PP bypasses on glass boards - definitely a notch or two down. He has shared in these pages the summary of his detailed, painstaking and thorough upgrade and the sonic outcomes. I have evaluated his FST boards and judge them to be inferior to classic Thiel. They include solutions we superseded in the late 70s with the model 03. It looks like Marqmike's boards are transitional between Lex and FST. MM's caps are Chinese production when Solen (France/Canada) took their bulk production to China. The coils are decent, probably from ERSE production moved to China. Audiophiles consider masonite point to point superior to 'glass. Note these glass boards are a hybrid with point to point layout but with copper solder pads for terminals. That's a step down from Thiel's physical twist joint stabilized with solder. The coil and hookup wire are CDA102, a step down from best of form CDA101, but still above the normal CDA110. I judge that these downgrades taken collectively, audibly reduce transparency and precision, but not by a lot.

I now have a pair of 2.4s of similar vintage and style as Marqmike's. I can't afford the time, but would be interested to hear a comparison of this execution vs original blueprinted Thiel execution. I clearly project the winner, but I don't know the particulars.
Beetle's execution takes original performance to another league. I have since gained access to a yet further refinement of the ClarityCap CSAs. The CSA is a game-changer to me, using foamed copper rather than zinc for the capacitor end caps. The subsequent Purity Cap increases the CSA copper cap thickness from 1mm to 10mm for significantly further enhancement, especially in high power usage. Presently I'm coaching a 2.4 upgrade which is likely to include CSA vs Purity comparisons for your and my edification.

My hotrod garage is presently closed, looking for a new location; but the project is alive and well with advancements being made on multiple fronts. One of these days we hope to offer various plans or kits or services to upgrade your 2.4s etc. to the limits of your budgetary courage. Meanwhile there are available known upgrades with no down-side risk.

Jazzman7 - sounds good to me. Shipping is problematic. I got a pair of CS3s shipped from Minneapolis to New Hampshire. Crating and freight cost around $700 combined, without insurance which would have added another $200.

Let's envision a gypsy caravan that covers the country by degrees doing rebuilds, offering advice, providing seminars and (why not) recording live music for distribution to our fans. I love it! I'll leave the details to you to work out.

Dreams are what the future is made of.

Woody - something that hasn't been mentioned is that Thiel, especially pre-coax models) require a listening ear height of 3' in order for time alignment. That's a baked in constraint. You can tilt the speakers to compensate for other ear heights, but do pay attention or you are likely to be disappointed.