Thiel Owners

I'll throw some thoughts into the mix.

Critical listening in an unknown space with unfamiliar gear and recordings is a challenge. In this case, the room had been designed to be modifiable for feedback from FuzzMeasure analysis, with which I am quite familiar and fairly adept. The room features a structured masonry / brick front wall and solid, infill masonry side walls. Such a rigid room makes for exceptionally clean sound, but carries downsides of pesky modal behavior. We borrowed 2' behind the listener from an adjoining room for a tunable chamber. Room side, right behind the listeners, is a unique porous wood-sliver panel, floor to ceiling, wall to wall. The acoustical panel drop ceiling at 10' has a 2' x room-sized chamber above to generalize the floor-to-ceiling modes. That sealed chamber still needs two short dividing walls to isolate it from a very large attic space. Chasing slap echoes, tuning modes and identifying further work to be done on the 4 deep-set windows consumed much of our time. We used music along with test tones and sweeps for this work - any music with broad-range content would work - and did. By Sunday morning we could compare and contrast the playback system elements under consideration. 

Reference cables included Morrow and Straightwire as well-known to me, and some serious foreign contenders that Max knew and loved. We settled on the Morrow as baseline to demonstrate my TRX interconnects and internal hookup wire. TRX is Thiel Renaissance X (24th iteration) - of major conparisons, not all TR. We benchmarked via Benchmark, Iconoclast, Cardas, Audience, Straightwire, Kimber, Morrow, and many home-brews to try out ideas as they surfaced. Suffice to say this wire dive was way deeper than the original late 70s exploration, the 1988 CS5 re-examination and the 2005 survey for the CS3.7. Incremental progress was made via input from the above-mentioned collaborators as well as advanced interaction with MIT operatives. 

I am aware of wire skepticism. My approach is not fundamentally skeptical, but is strictly evidence-based as well as requiring theoretical foundations. This family of analog signal cables embodies those requirements. Most of all, it must perform demonstratively well on all fronts. I have employed a hefty handful of listeners along this trail. This trip expanded to include duramax747's assessment. Feedback/ input has been extremely consistent and positive. Disappointments and dead ends have mostly been attributable to available materials and compromised methods of the prototyper's constraints. 

If I were to propose adjectives, they run toward clarity, detail, and depth of structure. All support Thiel's core performance attribute of 'vivid solidity'. The measurements support the subjective sonic experience. The outcome reflects the lessons learned along this deep-dive journey. We are especially encouraged because among prototyping limitations is working with available materials and methods. I use magnet wire with polyamidimide insulation - far from audiophile standards. My positioning filler is fibrated polypropylene, which is first-rate, but my hand twisting loom is far from the precision of real planetary machinery. A technical description of the cable will have to wait for decisions by the working group as to IP concerns.

Back to Max's room near Charlotte - we had various iterations of the SCS4 from stock Thiel to fairly hot-rodded - TRX internal wire to outboard XO with serpentine layout honoring waveguide principles with elements arranged for optimized field interaction, upgraded cabinet edge treatment, custom proprietary multi-segment caps from Reliable, and fluid dynamic wavelaunch technologies from Douglas Pauley. This wavelaunch tech is the most surprising and exciting aspect of this redevelopment work. I would have never thought that Thiel's already excellent anti diffraction wavelaunch could be qualitatively improved. The improvement allows the soundfield to hang in space. Max's first response after demonstration regarded 'projection'. Among the multiple causes of that 3-D spatial projection, Doug's wavelaunch tech leads the pack. 

I believe that we have developed synergistic, meaningful, cost-effective methods to take my brother's designs to a higher performance plateau without modifying any of his work and achievements. I am encouraged.  

Back to music. Max had a variety of big, bold, well-made recordings. I used two for orientation: Patty Larkin's 'A-Go-Go' which is a composite of live, on-stage recordings / one woman with her guitar with minimal artifacts. Straight-forward, well-done music making by a master whom I know, who's guitar and maker I know well, and that I have heard live in my small village venue. I believe I know her musical intentions. The other was Dana Cunningham's premiere 'Dancing at the Gate' which I produced from a simple music-school stereo recording to DAT and mastered with no compression, EQ or other artifacts. Contemplative piano on the half shell. My post-Thiel life includes lots of piano involvement including an innovative bridge and soundboard bracing system. These artists are interesting, involving and careful with their presentation. Subtle playback system mods track better when we care about the music. Last weekend was a wonderful musical experience for me.

tmsrdg - it’s weird stuff, and addresses something where Thiel already excels. The terrain is the integrity (or not) of how the driver energy translates into the limp air-mass for you to hear detail. Thiel’s bezels, waveguides and curved baffles with pebbled finishes all address this issue. But further research demonstrates the audible effects of wavelaunch turbulence. In short, the moving driver diaphragm moves air up to its modulus of deformation - it misbehaves as it tries to establish a coherent wavefront in the air. By modifying the surface to better manage the sheer forces and turbulence in the boundary layer between the non-moving baffle surface and the moving propagation wave, we define wave-fronts with considerably more integrity.

Douglas has patented two interlocking technologies: one addressing the propagation surfaces and the other managing the boundary layer behavior in the port itself. I was exposed to this nascent tech in 2018 and have been co-developing it for these past 5 years. A surprise is hearing a type of sonic improvement where you wouldn’t expect needing any.

tmsrdg - patience is the key.

The 3.7 is late model classic Thiel. Its ’needs’ are far closer to ’wants’ than earlier models which have far more real problems, upgrade potential and far more head-count in the field.

But to address your question, there will be a multi-pronged development and delivery. Each model will have its own upgrade protocol. Your 3.7 drivers are ’settled’. Your crossover will have some user installable upgrades. Specifically some of your Yellow Styrene 1uF bypasses will have replacements by way of custom Reliable multi-sectioned caps, in either aluminized or all-copper, as well as various non-inductive resistors. At some later date those upgraded parts can be re-used in a new serpentine XO layout, either inboard or outboard.

This first-offering upgrade will also include a set of wavelaunch films which can be applied by an adept user / or via some other arrangement - as yet unknown.

Similarly a new internal wiring harness will be available to be routed through the existing soft tubing, possibly with additional hold-downs. The new cable has less structure and depends more on mechanical support from the cabinet.

These examples are illustrative of future vision. At this time we are starting with training wheels with the SCS4, either extant, or new higher-grade specimens. Lessons learned there will be applied to next-step offerings, addressing older models with larger problems, such as replacement drivers and generally more room to grow. CS1 through 5 models will get early attention, starting with the CS3 / 3.5.

This information is a peek behind the curtain. There is ongoing work being done and many moving parts.

Cheers, Tom

I apologize in advance for many of you who have heard this before . . . I am a big fan of the sealed bass in the CS3 and 5 series. In fact, Thiel was philosophically committed to sealed bass in its upper models, while allowing reflex bass in its lesser models with the cost savings reflected in the lower sell prices. The main advantage I see in sealed bass (whether equalized or not) is the phase / time coherence all the way down to the low limit, and then 12dB/octave to  subsonics. That is radically different from reflex bass which attenuates the woofer at 24dB at its bottom and the port at 24dB at its bottom and top rolloffs. That puts the foundational deep bass a full wavelength behind the upper frequencies. That's 20+ feet, depending on actual tuned frequency. Philosophically that runs contrary to Thiel's committment to wavefront integrity.

When that direction was coming to be in the 3.6, the mood was more relating to what others' products (even with 6-figure prices) were getting away with. Some of us are more sensitive to the philosophical integrity, and others notice the sonic discontinuity of reflex bass. Note that Thiel reflex bass is executed as well as I've heard it. Look at the clean graphs and the tight tuning. But . . . it's still not the real thing.

Cabinet leak is a good idea. External is unlikely unless there has been cabinet damage. Loose driver screws could cause a leak. Also between the large/woofer enclosure and the 6" midrange tube enclosure. Check that by pressing the midrange and seeing if the woofer / passive radiator move.

Tom D - theaudiotweaktheaudiotweak

All the factors you mention led us away from ports except for small / inexpensive speakers. A passive radiator obviates most of the problems you cite. For those ported models, my new work introduces both surface flow and film resonators to make the port very much more ideal. The difference is significant. Those technologies are suitable for any / all ported speakers. I hope to have beta-ready trial kits this year.

Jim - You are correct, the EQ circuit boosts the bass starting at 300Hz to 12dB at 25Hz. Our initial intent was to EQ the woofer and leave the midrange and tweeter unaffected, but the slow XO slopes required grouping the woofer and midrange to keep the frequency response correct - resulting in far less efficacy for the bi-amp idea.

Note that any perceived differences in the upper range caused by the EQ are indeed real. They are artifacts. In that regard the CS3.5 EQ was a performance advance, and is a drop-in swap. 

Note that the bi-amp configuration caused more problems than it solved. For one thing the tweeter-only amp is a lot of cost for little return. Also, to maintain proper frequency response, the two amps must provide equal gain into these particular loads, which is far from trivial. Likewise, the cables must be the same as well as same lengths to maintain proper time alignment.

Our rather extensive testing confirmed that a single amp and single cable always outperformed the bi configuration. Going forward, we swam upstream with that single input approach.

Andy - I have no inside information. Rob is in a difficult time with his failing mother. CSS is viable.

Hi Tom - Lots of experimentation and solid results. No viable manufacturing capability yet.

The wire / cable journey has been a very deep dive. My solution is unique, effective and manufacturable. I have benefited from Ray Kimber, Mike Morrow, Steve Hill (Straightwire), Cardas and Galen Gerais at Iconoclast / Belden. My bottom line is that stranded wire is inferior to solid wire. Extruded insulations do harm, even the revered PTFE. Dielectric effects and signal propagation integrity are far more important than metallurgy. Connectors matter a lot.

Since my cable contains proprietary elements, I must be careful and incomplete in what I say.

The hookup wire is gauged per driver. Bigger than big enough is too big.

Each polarity leg consists of 2 half-sized conductors to raise the skin-effect saturation point. Cable is star-quad with a core drain to ground static charges, especially on drivers. The core drain provides some shielding (shading) without the tunnel effects of an external shield sheath. Drain and signal pairs are decoupled via counter-rotation and spacing.

Individual wires are enameled at 0.00075" film thickness for qualitatively lower effective dielectric constant than any insulation, including PTFE.

Cable layup uses round spacers to isolate wires with minimal (tangential-only) contact. Therefore most of the isolating is done via air-space.

Having no conventional insulation, the cable has very little body. Its integrity is supplied by fastening to cabinet walls and braces via adhesion to natural cork and cotton string tie-down. This exoskeleton wire becomes part of the cabinet and therefore is connectned to the crossovers and drivers via concentric / axial connectors.

Routing is more carefully away from driver fields, as are crossovers which are now on multiple boards to decrease field interactions between components.

There is a family of analog cables. Signal interconnects as RCA and XLR, speaker cables terminated in either locking bananas or GR Research Electras (Hint: Propagation field integrity matters most) as well as internal wire harness.

Obviously my 2024 introduction date is wrong. But, continual progress is being made.

devinplombier

The role of audio wire/cable has been routinely underestimated by mainstream engineering. However, in aerospace and now in high-speed digital technology, wire has risen to the status of a system-limit - and therefore importance. There are some design engineers who state that audio cable is a large multiple more complex than digital cable. Audio must support radical changes of voltage and current with back-forces. The problems are real and the solutions difficult.

The speakers you cite are taking their best guesses and engineering solutions within their defined parameters. Those parameters, understandings and blind spots vary widely among designers. Welcome to the murky world of audio engineering.

My recent deep dive builds on Thiel's history. In 1977-8, we faced the daunting task of whether we would tackle the impossible dream of coherence. We found that a coherent source showed glaring problems which vanished when reverting to normal high-order, non-coherent topologies. A near deal-breaker was a persistent gritty, homogenized haze in the coherent iteration that defied our attempts to understand and mitigate. Our aerospace-physicist (non-audio enthusiast) cousin made a visit, heard the problem and suggested we all (everyone) were hearing the effects of inter-strand cross-talk as learned from deep space probe image retrieval. Lesson learned included how our audio neurology processes coherent audio input at qualitatively higher scrutiny than non-coherent input, which it categorizes as artificial and less important.

I suggest that you would hear significant, qualitative differences between your two, or any, cable scenarios if using a coherent source such as Dunlavy, Thiel or Vandersteen, whereas those differences would fade to near meaningless with any non-coherent speaker.

My present work is, of course, with Thiel speakers. I've used the CS2.2, 3.5, and now extensively the SCS4. But the lessons apply to all speakers, and become meaningful for all coherent speakers. I thought that Thiel's 1978 model 03 brought the use of solid, rather than stranded wire, but we later learned that Dahlquist (from aerospace) used 18 gauge solid in the DQ10 darling of the day. Our 18-2 solid twisted pair in teflon spanned Thiel's whole timeline with critical comparative re-evaluation in 1988 for the CS5 and again for the 2007 CS3.7.

Regarding variable gauges for frequencies - the differences are subtle, but again, become meaningful for coherent sources. The math that describes propagation in wire diverges in the lower audio frequencies. Below 1kHz the rules get squishy and below 100Hz the rules diverge. Bass frequencies are supported with greater propagation integrity with conductors with lower surface to core ratios. So larger gauges are 'better'. Happily, skin effect saturation frequency also decreases since woofers are attenuated in the upper audio octaves.

The soup of interactive ingredients is extremely complex. The cable that I have developed makes sonic and measured advances without harm in Thiel's coherent topology.

Yep - those 48 strands work out to 25awg good to 85kHz saturation / 2 for layup loss. All good except. My quibble is that a tweeter works around 5 watts for negligible current demand. Meanwhile that entire conductive mass must be charged and discharged at every cycle. A single pair of 25awg per leg would allow simple helical layup, carry the required current and produce greater, audible immediacy. You could read that immediacy as a shorter up-ramp of the onset transient with a cleaner initial onset and peak reversal. 

You're 'one great voice' is a good visual. My mantra is 'vivid solidity'. You know it when you hear it. The dozens of cables I've tested all sacrificed that vivid solidity enough that I felt compelled to take Thiel's original solution to the next level. Obviously that's a different place than the vast majority of brands have gone. Indeed some Thiel upgraders have gone to various stranded solutions and liked the outcome. I prefer and am perpetuating Thiel's pursuit.

I'll vouch for the 1.2 as being the first second-generation Thiel speaker. It preceded the CS5 and as such was the working trial for our first copper motor shunts, and aluminum dome tweeter. This was the first product to benefit from our investment in Finite Element Analysis. And it shows.

The CS1.2 is one sweet speaker as is, plus it lends itself very well to XO component upgrades.

devinplombier

I don't know much about the Apogee line across the board. There are other high-end brands and models - Wilson Watts come to mind - that are have treacherously abrupt impedance swings with sub-1 ohm bottoms . . . and advertising that claims them easily driven. Thiel speakers present extremely resistive, non-reactive loads. But since their impedance is (too) low across the board, many amps just can't deliver the juice to drive them well.

Duramax's advice is spot on. Listen to demanding material at high volume and assess the enjoyment / naturalness of your response. I find it less than productive to try to analyze and describe the nature of the failure(s) because there are so many interacting factors and such academic understanding isn't the goal. Finding worthy ancillary equipment is your goal.

duramax - I'll be happy to ruminate on the CS5 baffle. The CS5 followed our 1987 CS3.5 which was a runaway hit, and our most expensive product to date at $2450/pair. Times were good, the market was hot and many manufacturers were presenting $20,000 statement speakers. Jim was extremely uncomfortable presenting anything over $5000/pair. So the CS5 actually represents quite an internal struggle. Contrary to statements that the CS5 was a 'cost no object' design, it actually carried compromises to keep its sell price well below $10K for a design with a natural price around $15K. Let's talk about the baffle as an example.

The CS3.5 had a machined 2" MDF baffle with braces. We also fastened the drivers with 2" screws through the entire baffle thickness into hardened pilot holes to best distribute reactive loads. An equivalent baffle for the CS5 would have been 4" thick and still fall short on impulse integrity, and add significant size to the enclosure. To develop a relatively thin-wall baffle with significantly higher stiffness and hardness, I engaged a local marble shop (sinks and bathtubs) to make samples from shop-built molds. By using 3 marble grits we reduced the polyester binder resin to half of typical and increased the stiffness by nearly double while nearly eliminating ringing. That baffle was (I think) about 20mm (3/4+") thick providing considerable internal enclosure volume.

We kept the overall size of the CS5 to no larger than an average woman. Jim invented a bass alignment that paralleled two subwoofers with a full range woofer such that all three were producing low bass at a declining impedance to draw additional power, but in a range that rarely has much musical content. It works and I hoped to apply the concept to further products, but we changed course to reflex bass due to its far lower cost and problems.

Riffing on the baffle: we designed the cabinet around the shrink-factor of that 'dry' triple grit baffle pour. We also developed a dye process for the marble so we could use a clear gel-coat showing off the marble grain. A next-generation concept (we always looked well ahead) enhanced the triple-grit concept to include granite, basalt and marble for even lower resin content, higher stiffness, greater internal damping and a more stunning look. But from a more conservative viewpoint, Jim and Kathy re-budgeted the CS5 at $9300 retail, so we had to lose the dry triple-grit x clear nicesity. The 'normal' pour gave the baffle more shrinkage and the cabinet dimensions shrank a little for it to work. Somewhere around the first 50 pair of CS5s had black gelcoat polished directly off the mold. But mold maintenance had also been deleted from the budget and all the rest of the 500 pair model run were cast with a primer gelcoat and sprayed in-house with emron (aircraft epoxy). My production records showed that those 'budget' baffles cost considerably more than the original, superior baffles would have. Such things happen when non-manufacturing executives step in.

As was our habit, we accumulated ideas for the CS5.2, which never came to be. The home theatre market was steam-rolling much of the esoteric high-performance marketplace. Jim wanted to go where he could design / develop more products faster, which was not a CS5.2. To a similar point, the sealed bass with its natural 12dB/octave in-phase and time aligned output was jettisoned for reflex bass in future products more or less because everyone else was getting away with it. That's a discussion for a different day.

Another point is that Jim wanted the CS5 baffle to follow the tilted flat plane of our other products. That necessitated electronic bucket-brigade delay on the upper and lower midrange drivers accounting for nearly half the component count in that huge crossover. Those coils and caps are all in the signal-feed path, which creates most of the sonic reticence experienced in the CS5. Now, close your eyes and imagine an arc-plane baffle that eliminates that electronic delay in favor of proper physical driver placement. Can anyone visualize a CS5.2?

As it stands, the cast marble baffle of the CS5 is enormously effective and beautiful. That product is a testament to a time when people of fairly ordinary circumstance could afford a product that still holds its own 35 years on.

devinplombier - we did explore low level crossovers. Net result is that we knew our niche which was all analogue and purist. Other folks have done versions of what you suggest. We always looked at overall cost effectiveness and separate amps, cables, etc. are not cost effective solutions, in our opinion. Also, the time of the CS5 development was 1988 - early and crude digital performance.

Of additional interest may be that the level of control needed for net resultant 6dB/octave slopes over approximately 7 octaves is far from trivial, even with digital modeling. We committed to minimum phase x time aligned performance with the 1978 model 03 and never looked back. Note that one would need to model, in addition to the driver roll-off slopes, the various reactive circuits to cancel resonances as well as the impedance correction shunts to maintain Thiel's resistive (non-reactive) load profiles. Most folks are very surprised how difficult that is with off-the shelf digital filters.

The approach we were intrigued by is low-level, active analog circuits before the power amps. In fact one of our first, pre model 01, trial products in 1975 was just such a speaker with 3 built to order amps and custom active crossovers in a 10" 3-way in a large bookshelf format. That product was unfeasible for a new company in a farmhouse. The technical aspects were manageable and indeed prototypes were impressive. But we lacked the stuff for market education and penetration. Indeed, powered speakers never became very popular.

To your question: we envisioned a lot. Line sources, spherical globes, di and bi poles. Active equalization was our first market entry and contained enough uniqueness to create more demand than we could meet for the next 30+ years.

Cheers, TT

lars - a few thoughts. I have rather recent acquisitions of both CS5i and CS2.4, but not much actual experience with either of them. I’ll jump to my conclusion first which is that the 2.4 might well suit your needs better due to what may be your fairly small space.

The CS5 has individual drivers, which at 8’ don’t integrate their soundfields very well. Also, the primary advantage of larger speakers is to fill larger spaces with deeper bass and higher amplitude. The 5 will do that only if you have the proper amplification to drive them, which as you’ve read here is a very big deal. The 2.4’s upper coax tolerates listening at any distance without compromise. The listener is freed from the triangulation necessary for individual drivers to integrate.

Another trajectory is that as Thiel and in particular Jim learned more he developed more sophisticated drivers. The CS5’s only driver from the ground up is the UltraTweeter. The ’i’ designates ’improved’ and adds copper motor shunts to the 3 woofers for significant bass improvement. The Focal lower and MB/quart upper midranges are very good stock drivers having none of Jim’s innovations. I’m saying that the 5 is earlier on Jim’s journey and as such the drivers are more ordinary. But, if you are filling a large space at full amplitude, the 5 puts out much more sound. Unless, of course, you can’t handle the power in your room.

There is another aspect that matters a lot to some people, myself included. The CS5 has true sealed bass. The transition from midrange to bass (down to 10Hz) is true first-order phase and time coherent bass. It acts like a real acoustic instrument in your space. The 2.4 is a very well executed reflex bass system using a passive radiator. The transition from the upper bass (woofer) to the lower bass (passive) at 24dB/octave (4th order) introduces a full cycle of delay in the sub bass - more than 20 feet behind the upper bass. That alignment has become acceptable in nearly all speakers, even costing $6 figures. But it is less authentic than first order sealed bass.

My studies in audio neurology suggest that the brain builds audio understanding from the bottom up. And therefore inserting the timing discontinuity between deep and mid bass consumes effort to decode and suspend judgement regarding the error. Again, the 2.4 bass is about as good as reflex bass gets, but it’s still reflex bass.

The 2.4 is easier to drive than the 5. Reflex bass eliminates current draw in the deep bass because the bottom octave is supplied via tuned mechanical delayed resonance, not a driver motor.

A pair of 2.4s might be found at under $1K. And a stellar and affordable upgrade path exists because there is no electronic crossover between the midrange and tweeter. The 5, on the other hand has 4 electronic crossovers, each far more complex than any of the 2.4 circuits because the later 2.4 drivers behave better, needing less correction than those in the 5 / 5i.

Tom T

Beetle - our project was good both ways. Thank you for your careful and complete evaluations and reports. Even though we both think that you got seriously upgraded performance, there are actually big gains to be attained with less expenditure. One of these days I hope that you'll try out the next round.

I know it's been a long time - but things are coming into focus.

devin - Replacing like with like components is certainly the safest. Many folks underestimate the intricate co-dependence of all the elements. Even layout changes affect the subtle outcomes. Keeping values at original levels is a good start, but various parasitics also matter. The frequency response is rarely changed in significant ways; but time/phase generally is.

You may remember the maddening months to years delays between Thiel product announcements, or even live displays, and their actual release. Most of that time was tweaking, mostly in realms more subtle than most designers would pay attention to.

In our DIY / rework world, the largest trap that I’ve found is subbing foil for the original wire inductors. Foil is indeed a more pure inductor, but its parasitics are so different from wire that considerable xo tweaks would be required to re-settle the circuits.

The safest bet is subbing in Mills MRA resistors for Thiel stock. Another safe and surprisingly effective tweak is replacing any electrolytic cap with 2 x half-value cap. The ESR and other anomalies are reduced by half, plus if you turn one cap backwards, end for end, other anomalies are cancelled. If there is a bypass cap (like Thiel’s yellow 1uF), that can be stacked to make a triangle for a single field.

Replacing caps, especially larger values is quite costly and sometimes demands circuit tweaks. It’s not hard to throw as much cost into upgrades as the entire production budget of the original speaker. Not for the faint of heart.

TT

Devin - Thiel crossover schematics state the max tolerance of each cap. Some are 5%, some 10%. Among the strategies, I prefer to try for close to equal values when splitting. I've learned end-reversal via trials. E-caps are fundamentally symmetrical, but in fact the small differences, including lead-wire directionality, are audible and supported in the measurements. Music signal is AC, so optimizing for either direction produces mixed results; splitting directionality is always superior. 

By the way, Thiel's long-term cap brand was Solen. Their house sound is somewhat dark and smooth. We later went to ERSE for a more neutral, open signature. When they became unavailable, M.D.L (Taiwan) was chosen. and the final FST (China) cap was CYC. I use M.D.L from Madisound for classic late-Thiel signature.

My new layouts honor propagation waveform integrity. When splitting any cap, build a bundle that makes sense. If 3 caps, I build a triangle separated by cork pads, with the 1uF bypass in the middle of the bottom row. Air circulates through the whole stack. A unified field is floated around the 4 cap bundle. Trials prove this arrangement superior to other layouts. By the way in series feeds, the yellow 1uF Thiel bypass cap is replaced with our new multi-sectioned cap with multiple coaxial windings that start around 0.015uF and add up to the target value. Those will be available in 2025.

gmosley - first pass is to try Rob @ Coherent Source Service.

Next pass is to know that a group of us are working on upgrade packages of parts and implementations for existing classic Thiel products. The CS3.5 is on the A list. New grilles will be made similarly, but out of much stronger materials and incorporate better diffraction reduction techniques. At this time, you might consider simply filling the top and side edge steps of the baffle with plumber's putty, play dough, etc. to reduce diffraction and listen without grilles.

Of, of course, scare some up from Rob.

Did you get the Equalizer?

thieleste - I concur with vair68robert - Cardas is a good, albeit expensive, solution.

Background in brief: Thiel discovered its 18-2 x tight twist in teflon via aerospace, and solved otherwise insurmountable problems developing the 1978 model 03.  Phase/Time coherence brought many system shortcomings into bright focus and ordinary wire was a big limitation. That 'Thiel' wire withstood all comparisons over the decades: 1988 CS5 development where we compared all the available contenders, and again Jim’s 2006 CS3.7 Swan Song design. From Thiel’s perspective and value assessment, it couldn’t be beat.

I began my own Thiel assessments in the mid 20teens and got serious in 2018. My scrutiny operates a little differently in that my value equation assumes product ownership, so further investments become incremental and more affordable than in new product bill-of-goods. I was aided by all the wire/cable heavy hitters and have learned more than I thought possible, especially in the domains of wavefront propagation integrity, and who understands what and how. My studies and progress landed with physics and quantum states, in which company ’it’ all makes as much sense as one’s skills can digest, interpret and apply. Deepest, longest rabbit hole of my life.

In our present context, the only ’stranded’ wires that pass muster for me are those that somehow insulate strand from strand - which are typically called 'solid' even when very small. Those include Cardas, some Kimber, Morrow, Anti-Cable and Iconoclast. (And possibly others I don’t know.) Many of the geometries are quite complex and expensive, and very good. New Thiel's 16awg - stranded is a big backslide, which often happens in the Dr. Toole frame of reference of ABX and things only matter if provable under their parameters and testing protocols.

My (upcoming) Renaissance wire family incorporates everything I’ve learned and painstakingly proven in the lab and listening room over the past 7 years. From my perspective of ’do no evil’ and ’preserve everything’, it nails it. And it is relatively simple and accessible - and therefore affordable.

Summary: Wire matters. There are more ways to go wrong than right. Low frequencies act differently than high frequencies (beyond current capacity). It you want to replace wire, investigate the brands I’ve mentioned. If you want to wait for mine, I really expect it to be available this year as ICs, Speaker Cable, internal hookup, chassis wire and coils. It stands on the shoulders of Thiel’s classic wire, which in many ways is already best of form.

gmosley - Room placement is an art of its own. This forum can guide you to literature / information on that subject. Thiel products are wide-dispersion transducers, so try to keep them away from side walls or treat the walls more so than with many other speakers.

Thiel’s grilles are multi-purpose. In addition to the aesthetics, they also fill the edges of the baffle to reduce diffraction. Better if you use them.

The EQ extends bass to about 25Hz - worth the extra pair of interconnects.

beetlemania

- I’m brewing the tea we started those several years ago. It’s getting richer and stronger.

Among the many lessons learned were real developments in wire and cable. You and I both like Cardas; very well conceived and executed designs. But with considerable complexity and cost. I learned from senior wire designers that to work optimally, the design must be worthy, plus everything must be executed to perfection, and the connectors to even higher perfection.

I suspect that a big part of what you like in your Cardas conversion is the PBP posts. Any of you who have those big-lug Thiel binding posts like on the CS2.4 can run an experiment. Put locking bananas on your cables. Try the inexpensive WBT / angled, stackables from Parts Express. Remove the knurled post nuts. Plug the bananas into the banana jacks. Listen and report back.

Ruminate on that and consider the GR-Research Electra Tube Connectors.

Small progress every day.

Tom

beetlemania -

Biwire is a special case. There are theoretical and empirical advantages. I've gotten good results with internally-biwired cable (in the same jacket). Problems arise with separate runs that the speaker designer can't control. An individual can evaluate including their own preferences. Also, there are upsides to single-wiring including avaialbility of all the conductive cross-section for big transients. Etc. etc. etc. Wire is a maze as well as amazing. Tom

Thiel Non-Model History – products that never were -

Prof – thank you for this question. Indeed we’ve covered a lot of Thiel history here over the years, but little about developments and decisions behind the curtain. This story would fill a memoir, which sadly, has not written itself. So, I’ll shine some light, while limiting the scope and depth for manageability. There’s always more.

Context -

Perhaps somewhat oddly, I’ll start at the end. I was informed by an insider that at the end Jim was working on an omnidirectional full-range driver. Such an omnidirectional driver would fit nicely into our global orientation of the speaker as mirror-image of the microphone and belief which we shared, that the omnidirectional mic captures sound most similarly to the ear, and far more faithfully than any directional counterpart. Although directionality in mics and speakers is a sad necessity in stage and professional arenas, the requirements can be more well managed in dedicated playback situations with less sonic degradation and higher retention of sonic information via room treatment and tuning while retaining  inherently superior omnidirectionality.

I can’t speak directly to that late idea-in-development. But I can recount early non-products based on first-hand knowledge of my brother and my twenty year lived history with Thiel Audio. What follows is a summary sketch of some of the experimental non-products during the early years of Thiel Audio.

Let’s venture back to 1974 at my Georgewown Road homestead. As a rung on our ladder to self-sufficiency, my Conceptions Design Studio had taken on Walter Kling as co-designer-craftsman, and secured early success in the high artisan-crafts marketplace. Our informal community sought a venture that would utilize and engage all the willing members beyond Walter and myself, in an enterprise with enough breadth, depth and horizons to support us for our forseeable future. This was the 1970s when autonomy and self-employment were hallmarks of the emergent counter-culture. We decided to fund Jim for a year to investigate whether his electronics knowledge could be responsibly applied to this task of right livelihood for a group of friends seeking meaningful co-employment.

Sidebar: At this time there were no computers outside large institutions, no internet, and only nascent knowledge of how loudspeakers really worked. The Thiele/Small Parameters were barely a decade old and not widely in use. Jim was inclined toward electronics with a first-interest in circuit innovation. More sophisticated amplification is where we first explored. Loudspeakers were seen as necessary tools to prove and improve amplifier advancements. Our survey of available loudspeakers revealed competing limitations and trade-offs, and no particular solution for accurate, revealing laboratory monitors.

Non-Product History -

In that first exploratory year we discussed, explored and studied what was needed for a really accurate and transparent research transducer. What floated to the top was a spherical globe around 1.5 feet diameter, fully covered with small (1/2” diameter) full-range dynamic drivers. More extended bass response could be achieved via greater sphere size and driver count, and/or by crossing over to a powered sub-woofer or folded horn. We built both a powered subwoofer as well as a 6’x 6’ folded horn to Jim’s specifications. The long story can be summarized to our realization that we were in over our heads with far more questions and considerations than our resources and scope would accommodate.

Lets count the most significant trials to date:

Non-product 1: Spherical multi-driver hung from a wire

Non-product 2: Powered subwoofer

Non-product 3: Folded horn subwoofer

All were built, tested, evaluated and set aside as exhibits for the Future Non-Museum.

Next stage was a distillation of contending technologies. It was clear that powering individual drivers with individual amps held extreme promise. Each amp could be tailored to the particular demands of each driver, and low-level, active crossover circuitry before the amp could produce better results at lower overall costs. The prototype that emerged was a tri-amplified three way, small format speaker with active crossovers including bass boost.

Sidebar: Note that Meridian had not yet come to market and to our knowledge there were no such products in the world. We determined that despite our collective enthusiasm for the concept, we as a self-funded fledgling enterprise could not support market penetration of such a product. To reduce further temptation, we burned the prototype on the pasture pyre.

Non-product 4: Self-powered speaker

Further distillation led us to what became the model 01, a 10”x 1.5” actively equalized, high sensitivity, full range speaker covering 25Hz to 18kHz. That product was fully developed along with its manufacturing engineering and feasibility studies. We began selling that product to local markets in 1975 with encouraging results.

By this time, our founding team included Jim, myself, my wife Kathy, Walter Kling and Fred Collopy with talents in business design-development, and emerging personal computing. I considered that team of 5 as essential for critical skills to take the plunge of full-time commitment to this business undertaking.

By request from users, Jim next developed the conventional 6.5”x 1” ported bookshelf model 02 to higher popularity than the model 01.

Our third market product was to be the model 03 a floor-standing 10” 3-way. Through its development we discovered that time-phase coherence was an important missing ingredient normally traded-off as not important enough for its trouble. We sidelined the conventional model 03.

Non-product 5: Conventional tower 3-way floorstanding model 03

We struggled for the next year and a half before deciding to accept the impossible challenge of producing a coherent speaker. The actualized model 03 with its sloped baffle time-aligning all the drivers, and first-order slopes maintaining phase alignment and impulse integrity was introduced in late 1978. It utilized the active bass equalization of the model 01 which persisted through the 03 conventional, 03 coherent, 03a, CS3 and CS3.5 (the fifth generation model 3, including the seminal non-coherent original version.)

The founding team survived only a few years, and the loss of Walter and Fred were nearly catastrophic to the business. I consider that upheaval as a fundamental loss.

Non-product 6: The fully functioning involvement of all 5 founders

As we gathered experience we learned to evaluate and accept/reject potential products in the hypothetical phase with minimal commitment and expense. That’s the highlights; let’s stop here for today.

Marqmike- PM sent

Massimo - The early 3.5 tweeter feed used a 6.5uF feed cap. It was tweaked in 1987 to 8uF, which is what you want. Only buy now if you need it. I will have a much better replacement soon.

The 3.5 mid and tweeter are front-burner projects. We are incorporating late-stage design elements into the 3.6 midrange platform - dual cone, advanced motor, to fit the CS3, 3.5 and 3.6. The 3.5 tweeter is obsolete and our replacement will also incorporate late-stage advancements to fit the CS2, 2.2, 3, 3.5, 3.6 and 5. Completion of midranges and tweeters is necessary to re-work the crossovers between them. For now, get your advice from Coherent Source Service.

The ScanSpeak 10F8424 is not robust enough; it will burn out. SS’s recommendation for drop-in is 12W/8524G00. CSS may have that driver or another of their choice.

Tom

Prof - I know very little about this topic, only what was 'leaked' by an insider and not denied by another.

But, as I said, the approach harkens back to the very beginning, before we settled on dynamic drivers in an enclosure; and it attends to our shared first-principles of point-source, freely radiating energy without close range diffractive interference. And as I recall (from about 1974) our reasons for rejection were based less on technical appeal than on costs to market such a radical approach . . .

Sidebar: At our first 1977 CES, we twenty-somethings showed our model 01 and 02 and a static 03 prototype, and offered an unheard-of 10 year warranty - all to much interest. I remember being teased by an industry somebody that being from Kentucky, weren't we supposed to be barefoot and pregnant and sipping moonshine . . . ? Cognitive dissonance.  Imagine if we had showed up with seriously radical products beyond our active equalizer.

Since this is all water long under the bridge, I invite anyone to comment that might know more than I about this mythical omnidirectional project.

unsound - thanks for this input. It makes sense, including the context of return to a foundational idea before the company's launch - but with a life time of experience. Jim built every design on all the accumulated knowledge gained along the way. That product might have been truly awesome.

I routinely hear that the CS5 was Thiel's only cost-no-object offering. It was a niche-stretch at the time, but not cost-no-object. In fact Jim and Kathy chopped thousands of dollars at retail off the sell price due primarily to insufficient confidence of a price jump from under $2500 for the top of our line CS3.5 to what wanted to be a $15K CS5. They kept it under $10K by the loss of some product content plus damagingly thin margins, especially for a factory stretching its technical capacity. Jim emphatically stated that he would never design a speaker better than the CS5.

Anyhow, I'm pleased that he broke through that ceiling. And wish the world had it.

Thanks again.

Tom

ronkent - Regarding your mod. Be aware that the 3.7 mod that Rob suggested applies to only some speakers, not all. There was a QC problem with FST that required Thiel to test and characterize all incoming 3.7 coaxes and apply padding resistors as needed to bring the amplitude to standard. There were also harmonic distortion problems, which were failed and rebuilt in-house. When reworking drivers, CSS would add the resistors directly to the input terminals, or as in this case specify where to put them in the crossover.

Later, the coaxes were built to proper spec and noted with black trim rings. The 3.7 has an octave-to-octave balance within ± 0.5dB. So, you don’t want to pad it down unless it is out of spec, which is possible.

Note also that the mod as described here (padding both mid and tweeter) would effectively raise the woofer level a couple of dB. Moving the speaker closer to a boundary wall should accomplish the same thing. Also, any perceived brightness could be caused by harmonic distortion rather than simple excess amplitude. In that case padding down would reduce the gross amount of distorted output, but not cure the cause.

Roxy54 - for the record, with the 1987 CS3.5 we converted from 60/40 solder to silver solder for all our speakers. Our choice was the aerospace standard Alpha SAC-305 which is 96.5% tin, 3% silver, and 0.5% copper. It is not only permanent, but is technically superior. The improvements are audible. It requires higher melting temperature, but achievable with a 140 watt soldering gun.

spacebird - the number before the decimal point is the model. Higher is bigger. The number behind the decimal point is the iteration. Higher is more recent. More recent products incorporate all the learned advancements (that can be afforded).

Next up the ladder for you, if you have a larger room or want bigger bass or louder playback might be a CS2.2, 2.3,2.4. The 2.4 is a stellar sweet-spot.

I'm working on a product summery / timeline to help make the product journey clearer.

Welcome.

mchan888

Hello and welcome. This past summer I site recorded piano, bass, percussion and drums with my single-take, minimalist 2-mic documentary system, direct to SD at 32bit-192kHz. Earthworks QTC50s flat to 50kHz, no compression, eq, limiting, etc. Live to playback tweaking in the same space at the same time. We migrated from 3.6 monitors to 7.2s, my first up-close comparison of those models. I agree with your assessment. The 7.2 are the pinnacle of Jim’s work.

The 7.2 treble is not reticent. And its dispersion pattern is quite similar to the 3.6. So, if they sound dissimilar in your room, I suggest you lightly touch each tweeter while playing music to learn whether they’re both playing. If so, we’ll dig deeper. I have the schematics and welcome your PM to sort it out.

Andy2 and all -

I say that neither Jim nor Thiel Audio would espouse valuing ’"time coherence above other aspects of sound reproduction". Our company was built around addressing and honoring ALL aspects of sonic/signal/musical reproduction as a whole. Most designers - products minimize the importance of the time-phase aspects of fidelity, especially in the day that we did it. Only a handful of brands made time/phase behavior important - including Thiel, Vandersteen and Dunlavy and Quad, and possibly some smaller attempts.

Note our attention to time-phase was not above other aspects, but as one among several necessary ingredients for faithful representation of the musical signal.

Having paid attention to this stuff for half a century, my perspective is that keeping time-phase correct allows the ear-brain to pay attention to the playback signal as though it were real - thereby permitting a more holistic, immersive experience of the music. Although we rarely admit it, we humans do not possess unlimited brain-power. Work is required to reconstruct a musical signal that is missing its time domain content into an interpretation that makes sense. That effort subtracts from the state of consciousness that is possible when experiencing real music, either in its un-recorded state or its time-phase correct played back state.

Among the most frequent comments re Thiel/Van/Dun/Quad, etc. are ’naturalness’ and ’image density’. These are psychoacoustic attributes facilitated by the addition of phase-time correctness to the other realms of dynamic and tonal correctness.

I assess that designing for all of the musical aspects rather than discounting or fudging against the time-phase aspect requires an order of magnitude more effort. Everything becomes extremely more complex and difficult.

I can only afford a summary overview, the details took a career to address, and the work is still not finished.

Indeed.

Thiel's requirements of passing a square wave / exhibiting a single, proper step response, etc. increase difficulty so much that most practitioners consider it impractical or even a fool's errand. We went for it despite the difficulties. We were young and idealistic, plus we wanted to make a mark and improve the art. Over the years, we invented improvements that managed the inherent problems. 

The biggest problem/ limitation is dynamic range because each driver covers 7 octaves rather than 2 or 3. A driver acting outside its sweet spot has larger excursions, must dissipate more heat, enters break-up, etc. All those must be counter-acted with considerable difficulties. We began inventing new driver technologies withing the first several years.

It is also much harder to get smooth frequency response in a coherent system. Thiel's elaborate crossovers create complimentary circuits to correct driver anomalies that steeper slope filters would make less obvious. Also, we migrated to stiff diaphragms because their mis-behaviors are simpler, more predictable and therefore more manageable with circuitry. Those components introduce their own sonic degrades and add cost. 

We addressed these issues as cost effectively as possible. Our performance per cost was extremely high. As manufacturing director and later consultant to other manufacturers, I know that our output/cost was a multiple of average and our margins were a fraction of average. We tried harder.

Thiel's results were often less than best in some respects, but generally first-rate if one values over-all high performance on all fronts. Our speakers addressed everything quite thoroughly rather than a few things brilliantly. We believe our products supported a more musically authentic experience than conventional approaches.

I can say that Jim / we might not have gone there if we had known how hard it would be. I suspect that hind-sight and insight would have led us to the later-stage insight that we used in home theater products. We could maintain respectable phase coherence and proper time alignment while considerably reducing difficulty  with the fudge of keeping first order slopes for an octave on each side of the crosspoint and then migrating to second order symmetrical slopes beyond that. Out of band excursion and erratic behavior is greatly minimized while keeping the critical advantages of single step response and coincident time arrival. But you can't solve the puzzle until you know enough to solve it.

Most brands didn't and still don't even try what we we did. I'm glad we did.

Andy - you’re on it, and the CS5 is the best example of thoroughness competing with directness. There are multiple reasons for that ’reticence’. A big one that Atkinson and Archibald underestimated is amplification. The amps they used were current starved, delivering more anemic sound than best of form. The underlying root cause there is the excessively low impedance. I think Jim could have done better there - for another discussion. But to your point there are lots of components. My count from the schematic is 106 including 16 that are 1uF bypass caps. Note, we developed that 1uF (yellow) bypass with a European aerospace supplier as state of the art Styrene film x tin foil. Note also that Jim was fundamentally an electronic circuit guy. He worked as if the well executed electronic circuit did the job as well as could be done. That is debatable.

Of those 106 components, 32 of them are in the conventional signal path. Not particularly over-excessive for a 5-way design.

42 are in resonance (shaping shunted to common) circuits. Quite a lot.

And 32 are in analog bucket brigade time delay circuits in the  upper and lower midrange. This is a big deal where some history might shine some light.

We had gotten Japanese representation in the mid 80s which opened up the Asian market to us - quite rare for an upstart American company. Those importers pushed us for a ’statement’ product in a time where many such products were coming to market. Our best product was the CS3.5 at $ 2450/pair with EQ @ US retail. They wanted something at 10X that price, and we had 150 pairs pre-sold regardless of price. Jim was very cautious and less than wholly confident at that time that we could deliver high value in that class. In fact he didn’t want to make the leap and Kathy eased him into it over many months time. My (over-ridden) opinion was that this product-under-development had a ’natural’ retail price of $15K, but it was introduced at $9300. The other huge deal was time-line. The market wanted it now, and its natural development cycle was at least a year out. All that is to say there were significant stresses in the cycle, and stresses show in product strains.

Our signature straight-plane x tilted baffle format accommodated up to 3 driver set- backs by adjusting tilt and driver to driver distance. The CS5 had 6 drivers. It wanted a concave curved baffle for properly time-aligned driver to listener ear geometry. We solved the bass by placing the woofer between the two subwoofers (fore and aft the woofer launch plane) for a net single-point launch. The tweeter was time-aligned at the top of the stack. The upper and lower midrange had to be recessed about 5/8” and 1/2” respectively. The direct way to do that is geometrically. That would have required a thicker and more complex baffle, which could have all been accomplished with additional lead time, which the market wouldn’t grant. So Jim provided the electronic answer with analog delay, which he considered more elegant. This baffle fit our signature tilted flat plane. This long tale is to provide background behind how such products come to day.

I always pre-visualized new products as we developed our factory.  A CS5.2 could use a shallow cone upper midrange instead of the 2” dome to solve much of the timing issue, and applying our patent-pending double cones to the midrange drivers would increase basket depth options. The more complex 3D baffle shaping was feasible if Jim and Kathy could be swayed. I wanted to compete in the more expensive arena; whereas Jim and Kathy were not enthusiastic. Home Theater was rearing its head and Jim wanted foremost an arena to invent new products. HT became that arena.

So, Andy, back to your point that more complexity constrains a design. I agree. I also add that working to surpass such constraints is the business of innovation. We worked ahead of the curve, inventing many solutions that gradually became part of the industry playbook. The CS5.2 could have put many of those future innovations, such as motor shunts, rare-earth focusing magnets, formed double cones, etc. to good use. Such a path became reality with coincident / coax upper drivers, reflex bass (which I discouraged in our statement products) and so forth and so on. It’s a long tale of a complex stew with an outcome that the CS5 was never revisited, much to my disappointment.

But, as time went on and lessons were learned, there were never any more time-delay circuits, and metal drivers allowed simpler shaping filters; and the speakers worked toward greater clarity and dynamics – and reticence was reduced.

JA - awhile back you commented on the sonic improvements from the CS1.5 to 1.6 and larger improvement to the CS1.7. I would love to hear your particulars about the 1.7 and how it seems better to you.