Thiel Owners

Unsound - my apology. Indeed adding an amp and cables is a huge expense. My thought is with used market prices such an upgrade has potential to raise the performance level of the 5i a lot. And more to the point, the CS5i's difficult load practically limits its performance to where it rarely sings. Jim used a Krell FPB-600 - and it sounded beautiful. But doubling your existing favorite amp might be an attractive alternative to finding that Krell or other beast to drive them. Most amps buckle under the load.

Unsound - Regarding impediments to CS5i ownership, I am confident we can address the primary shortcoming quite easily - and that the amplification problem would be reduced significantly. The impedance drops to 1 ohm in the deep bass and doesn't rise above 3 ohms till 200 Hz. That severe bass load drains the amp of ability for transient response in the upper bands. By adding a second input for the 3 woofer/ subwoofer drivers, we can sequester the "amp problem" to the bass.
From history: Jim was aware of the problem and the solution. His personal make-up caused him to avoid the fix due to the potential chaos of cable and amp interactions in the field. However, if you used 2 identical amps with identical cable runs, you could successfully vertically bi-amp each speaker with its own channel for each band. As was mentioned earlier, each amp could sit close to its speaker so the 4 cable runs would be less expensive.

While in the XO we could easily upgrade some caps and resistors if desired.

Unsound - agreed on 40Hz, and even 20Hz at moderate levels in a moderate room with moderate music. I like the 20Hz 'sound' better than 40Hz, possibly because of the relative lack of low end phase shift, to which I am sensitive. The bass response effectively starts as low as there is program material. There are ways to successfully aid the bass against overload ie subwoofers, etc. I use an SS2 behind each main speaker, whatever model. In the Thiel passive XO mode they contribute only below around 40Hz (plus overlap), and there isn't much music down there. But nonetheless, that supplementary deep output provides local air pressure to aid the main woofer's room coupling. I'm surprised how much more authority the bass gains. 

Please say more about 2>4 upgrade path.

My CS5 comment was market-based. And as you allude, most of that was self-inflicted, in my opinion. The product is good, especially the 'improved' version with Jim's driver designs. It's not common knowledge, but I consider the CS5 and the squandering of its potential segue into higher range markets to be classic self sabotage. The product as designed wanted a $15K (some said higher) price and another 3 months in development. Rather than settle into that league, Jim and Kathy were adamant about keeping it under $10K. It didn't 'fit' the market there. There was much too much speaker there to make sense, and its amp requirements etc. took it into the higher league anyhow. (Note that we had 100+ carte blanch firm pre-orders with no price stipulation.) You might detect my flabbergast these 30+ years later; this rant has never taken words before. One example of 'what happened' is that I had designed the baffle to include 3 stones (marble, granite & basalt) in 3 particle size ranges for an inert - extremely well-damped, good-looking baffle. The stone mix had enough variation and visual texture that the gel coat would be clear, to see into the stone matrix - all quite subtle since ground marble takes dye well. The casting molds required higher maintenance, and it would be 50% heavier, but what a baffle! I found the right supply partner in Atlanta and settled on $100 / each with quantity guarantees, etc. (Note that a company like WA might add $25K for such an element). I was elated. That is until I learned that Kathy came behind me, capped the price at $50, reduced the spec to ordinary bathtub cultured marble, and gloated over the cost reduction, all on their way to a $9300 / pair introductory price, which caused market confusion against the $15-$20K pre-marketing by distributors and dealers. End of rant.
It's called family business. The J&K alliance was internally infamous. They abandoned the CS5 platform and stated that it was 'too expensive' to sell through. I suppose polite professionals are supposed to keep such things under their hats. But I think that you fans might benefit from some leaks whose potential harm has long-ago elapsed.  

Welcome back brayeagle.

From the other end of the upgrade lens, I just snagged a pair of 02s in near mint condition for $210 delivered: 0687-0688 (around 1980) from one-owner in Michigan. Cabinets never opened, parts all look new, the improved-cone woofer; they sound great. Thanks for the nudge Pops.

In the day, the BBC and therefore most "sophisticated" speakers were using thin walls and diffractive edges and "tuning" it all for a pleasant response. The 02s fall in that camp. The 5/8" walls sing and the sharp baffle edges add some zing and the stock caps and carbon resistor (only 1) all add up to a nice, musical presentation. Part of the magic is very few crossover parts and a well behaved driver complement. I'm looking forward to comparing them to some modern audiophile monitors in the area, just for fun.

Prof - "well" is such an interesting and loaded term. For years I have been pondering what makes some systems sound so "right" and others, perhaps technically more advanced, sound less right. One idea that keeps popping up is at the core of what we are doing in recorded music.

In real, live, acoustic music, there are many distortion mechanisms in play: all the reflections of the site, the compliance of air, the bouncing of and off of the floor and/or ceiling, the audience . . . and so forth and so on. Live hearing is extremely multi-dimensional.

In recorded music, we generally capture each sound source very close-in and pure with little to none of that other stuff going on. Even in distant-mic'd orchestral recording, we go for high-up and clean capture.

 Much of mixing and mastering is to introduce distortions to make the sound "better". But it is at best doctored and at worst pretty artificial on the disc. Yes, we play this doctored sound into our headphones or rooms and hope for it to sound "natural".

Many design(er)s of playback equipment accept adding compensations as part of their job. The BBC approach nods to that "sweeten it up" on playback idea. I believe that the ear-brain actually likes distortions, as long as they are harmonic, such as might plausibly have been in the original musical event. The 02 has something like 15% harmonic distortion in the port and a couple of 5%+ areas in the upper bass. Add some diffraction around the drivers and at the cabinet edges and the different panel resonances to the squirmy driver diaphragms . . . and there's a pretty rich soup. And it sounds quite comfortable, kinda friendly, maybe seductive.

At the beginning of Thiel Audio, we actively struggled with this dilemma: do we make a transducer that sounds inviting, or do we make one that replicates its input signal as faithfully as possible. As you know, we chose the latter, which is far more difficult to pull off plus the risks are large. Especially in the early days of digital, but even today, there are far more poorly made recordings than well made ones. And prior to that, there is no real agreement that I can find concerning where the responsibilities lie for what goes on the disc. At Thiel, the winning strategy was that the speaker should accurately translate its input signal into moving air. (Period.) Everything upstream is someone else's responsibility as is the downstream room. There are dozens or scores of aspects and technologies to manage in the chain; for the speaker to try to referee that whole kettle of fish is untenable, actually impossible. And therefore the purer philosophical stance is to make a clean speaker, which we attempted.

My observation is that the recording process often doesn't produce a signal that sounds good when played back neutrally - it requires or assumes various distortions being re-added through playback. The Thiel 02 is by far the highest distortion speaker Thiel ever produced. And it had in its day a very loyal following. I like it surprisingly more than I think I should. My experiments here with bracing the cabinet, upgrading caps and softening edges all serve to clean up the sound and make it sound more "Thielesque", but at the risk of losing some charm. This is like deja vu all over again. The end of the 02 form is the SCS4. I am shopping for a pair of those to help sort out what the Renaissance 02 should be.  

Prof - I'm out of touch with today's market, but there do seem to be plenty of very expensive speakers out there with knowledgeable folks wishing for better performance.

About the next room phenomenon - that addresses an open question in my mind. I don't know if the industry agrees about radiation patterns. I do know that we at Thiel decided in a technical vacuum that our speakers would act as a quasi point sources with the 30° off-axis radiation being as close as possible in power response to the on-axis response. Definsible guess.

I now understand from sound reinforcement that directionality is part of their formula, based on very sophisticated Green's Functions governing wave propagation through distance. I don't know what pattern is considered ideal or if home playback should follow the same rules. But I do know that from the beginning Thiel took some hits because our speakers had "too much" high frequency energy in the room. J. Gordon Holt, founder of Stereophile, is known to have commented when hearing Thiel 03s from the next room "it sounds great from here, but music doesn't sound like what we're hearing in there".

I do know that recording engineers differ widely regarding the playback conditions assumed in their mixing and mastering.

Unsound - on a related note, I just read the 2003 thread with your active participation that included lots of very lucid input from Roy Johnson of Green Mountain Audio. He alludes to adherence to the Green's Functions in his speakers, along with other very advanced assertions. Smart Guy. Their website seems to have truncated at 2011. Do you know how the Green Mountain products fared in the marketplace? Or what critical acclaim they got, or agreement that their limited dispersion formula was "right"? I would appreciate whatever you can share.

Prof - thanks for the links. Kevin and Floyd are gurus. I've seen some of this stuff before and I should study it closely - oh time. But I do take issue with a core element which paraphrases something like: adjust the outcome to match what more people like. That subjective preference projection is less than clean, no matter how many people prefer it. I would certainly prefer a solution that put 2dB less information between 1-8K (or whatever) onto the disc such that the transducer would measure flat.

The puzzle is pretty deep.

Thanks again for the links. I'll find time this weekend to do further reading.

Unsound - thanks for the recap. You might remember that I use PowerPoint 1.2s with an unusual 45° wall launch pattern, which I have placed on adjacent walls of a corner.  There definitely are merits to the engineerable predictability of boundary propagation.

Tom

Beetle -  I suspect our Renaissance Series will have their own oversized backplates for the double-post upper levels.

Beetle - I have not yet made a direct comparison of single vs bi-wire.My working assumption from multiple inputs is that there will be an improvement; the degree and cost-effectiveness is my concern. Additional to your experiences and those of others, my personal experience is from the development of the CS3 in 1982, which was bi-wirable. Those development experiments were conclusive and cost-effective.
Thiel's reversal to single-inputs was one of those cases of market consideration, and control. Bi-wirable speakers are viewed, especially in the day, as "requiring" double cable runs and/or dual amplifiers to meet their design goals, which significantly increases consumer cost. The other rub was that some users chose radically different cable runs which degraded system performance. Jim could not tolerate that system degradation and Kathy felt that adequate consumer education was not achievable. 
All of my cable runs are bi-amp in the same jacket, which costs very little more than the same wire with single terminations. I'm getting close to building some crossovers to test.

Prof - at that time I was not tuned into Thiel Audio nor Rolling Stone, so I heard nothing of that "test". It would be interesting to know of any follow up or results.

espirits4s - I cannot address more than the CSi, since I have no personal experience with the CS7.2 or your particular amp. But that amp doubles twice to 1400 wpc into 2 ohms, implying that it should do fine into the 1.2 ohm load at 10 hz. The 7.2 has the opposite characteristic, a rising impedance going low, so a long cable run would act differently between the 2 models. But, it seems far-fetched to me that an amp or cable interaction would be responsible for a significant bass difference. 

I have considerable experience (in the day) with the CS5 - and I wouldn't call it bass-shy with a good amp - which you have. So, I would look for a speaker problem. First-off you can test viability and polarity of all the drivers with a single 6-volt lantern battery rigged for intermittent contact to the input terminals. Plus to red should make all drivers push out - a second or two won't hurt the drivers.  A dead or reversed-polarity driver would make trouble. The lowest and upper 8" woofers are the subs which would reduce bass power without affecting tonal balance noticeably. Please report your driver test as a base-line.
Tom

Esprit - have the 5s and 7.2s been tried in the same position in the same room with the same driving source?

Waveguides are attached with a 3M plastic adhesive and usually come off readily. I can't recommend any tricks. The rubber goo surround fix is good. The original goo was silicone.

Gary had been Jim's lab assistant and audiophile interface for at least a decade. He was part of Home Team Thiel for the 2.7 development. When New Thiel took the reins, Gary held out for Phase Coherence in the new products. He was over-ridden by the New Team, including Steve DeFuria and Bob Brown who had been long-term Thiel supporters and associates. They said that Gary had "drunk the kool-aid". Gary left New Thiel when Mark Mason decided to pursue Canadian-type, non-coherent designs.
Bryston and Thiel had collaborated for decades, sharing and trading equipment for the lab and for shows. Good fit for Gary.

There were knowledgeable insiders and observers who collectively judged the new ownership's approach to be built on "appalling arrogance and ignorance".

In times of change, some people go where the wind blows to keep their jobs and carry on how best they might. Thiel employees were unusually principled. Most stated their principles and committments and parted ways. Rob found a way through the maze to carry on well by remaining conscientious without addressing policy. Most of the Real Deal Thiel Team evaporated before the Lexington factory went in the dumpster. None made the move to Nashville. The Thiel Audio we knew was over within the first year - 2013 - which is getting to be a long time ago.

Eagle - I think the Bryston speakers had been development a long time, but Gary may have been in on final tweaking. He has a good ear and did lots of listening with Jim during development over the years. BTW: Gary set Rob up with a pair of 28Bs. If they are Gen 3, I think that's Bryston's top of the line. Their amps have gotten more and more refined over the years. I'm planning to hear Rob's this summer.

Thanks guys. My working assumption is that it needs service and is a candidate for upgrade. Do you guys know anybody who is doing the serious upgrades that were being done awhile back?

The bankruptcy auction got a lot of participation. In addition to Jon's observations, I noted no single big buyer for all those under-priced drivers. It seems that most finished goods sold at pretty significant prices compared to actual cost.

I dropped out of Jim's big Krell amp at $2500 - it sold for $2700 compared to around $3500 on ebay and 'Gon. I bought the Sony SCD-1, so I can play SACDs and CDs in higher style than my Philips CD80. I'll be interested to compare sound quality.

Jon - that stuff is only lost 'till it's found. Still working . . .

Jon - I use the PowerPoint 1.2 extensively. I am told on good authority that the PowerPlane is the same driver and that the XO is modified to account for the different wave-launch environment. The speakers are said to be virtually the same except for the unique wave-launch of the PowerPoint.

cabinlife - I offer my assistance in documenting your CS2s for the purpose of educating us all in their particulars. There was a XO refinement in May '87 and original electrolytic caps are up to 35 years old. And so forth. If you like, you can post your serial numbers and whatever else you wish and various folks here might help guide you toward optimizing their performance.

Tom

CS2 guys - welcome to a remembrance. The CS2 is among the most successful and most formative in Thiel's history. The 3-series launched from near the beginning; the CS3 was 4th generation thinking in 1983. The CS2 was a fresh start in 1985. It was devised as the little sister to its big brother CS3 - smaller, less expensive, less bold for smaller rooms and music, delicate and refined. It was the largest count and longest running speaker in the history of the company. Rojacob's fall at the end of the cycle. Some dealers encouraged us to delay the introduction of the 2.2 because sales were strong for the 2. In fact, the two products over-lapped their production cycle in 1990.

I want to tribute Tim Tipton as part of the CS2s success. Tim had succeeded in his own enterprise and came to us as a seasoned manager - he ran our purchasing department for 20 years until his retirement. Tim brought concepts such as progressive forecasting and commitments. The CS2 concept required parts cost to be 1/3 lower than the CS3, in addition to cabinet materials and labor 40% lower. Let's look at drivers and crossover components. The Dynaudio D28 tweeter was an expensive unit, used in many of the leading brands at that time. The Vifa midrange and Seas woofer were substantially customized and therefore priced at premium. Tim negotiated prices based on our forecast of 10,000 drivers with annual commitments a year ahead of need on prenegotiated monthly releases. In a volatile marketplace filled with small, transient, unstable startups, Tim positioned Thiel as an anchor around which driver suppliers could plan. The importing distributors could bring in thousands of units for us on a predictable schedule. He negotiated every ounce of cost reduction from the situation, and with price-freeze protection 2 years out. Tim's local reputation was that he could squeeze and dime out of a nickel.

The crossovers were expensive - including 6-9s aerospace coil wire which I had sourced from ITT for the 03. Tim applied the same long-range forecasting to that wire for substantial cost reductions. And caps and so forth. He also introduced more sophisticated accountancy, such as the CS3 drivers, caps and so forth continued to bear their previous cost burdens and the negotiated reductions were applied to the CS2 bill of materials instead of averaging the shared parts costs.

In my cabinet planning, Tim served as an intelligent participant. He helped isolate the profound cost savings of concentrating the wave-shaping mechanism into the grille board rather than the 3-D baffle augmented with a complementary grille of the CS3. Remember, this jujitsu machining all preceeded CNC technology and required serious dedicated methods with high-skill workers. The grille-boards became an independent operation, untied from cabinets, wood finish prediction, run sizes and so forth. CS2 cabinets had all flat panels with only the single angled baffle.

Some of you might enjoy a marketplace anecdote. Tony Cordesman was writing for Stereophile at that time and gave the CS2 an astoundingly positive review, taking the Quad ESL-63 as his comparison against which the CS2 stacked up quite respectably. Tony specifically cited the elegant success of the CS2 grille for dispersion control and diffraction mitigation. Stereophile's new publisher Larry Archibald took on the CS2 as his extensive long-term reference and always niggled an upper midrange edge . . . It turned out that Larry "never used grilles" and therefore had negated the carefully engineered tweeter wave guide and rounded edge boundary. He later claimed that Thiel had solved the edginess problem via crossover changes. Tony was an extremely astute listener, migrated to The Absolute Sound, and all these gyrations stay behind the curtain.

Back to Rojacob's CS2s. That historic model doesn't turn up much these days. But considering its inherent strengths I suggest they are well worth reviving. At nearly 30 years old, the electrolytic caps are near expiration. Storage is especially hard on Ecaps and failure of series feeds would endanger the midrange and tweeter, which are out of production. Your coils and wire are state of the art. Your schematic is tweaked and final behond #4900. If you want to delve, I can recommend caps and resistors as well as some cabinet tweaks. The CS2 began life at $1250/ pair in 1985 and were always compared to products costing some multiple. You could make them better than new with very little investment.

George - I use PX02 and PXO5 crossovers rigged for each model. They create an inverse of the roll-off characteristics of the speaker, therefore creating an ideal match by definition. That solution was patented. It seems to my ear to work very well. I have never heard better integrated bass, and it seemed to get high marks from reviewers. There are issues of amp noise and repairability. 

Performance issues remain. One function of subwoofers is to increase the speakers' output. We run out of bass excursion first because we seek to tune the enclosure as low as possible. The PXO doesn't increase the system output limit, but only adds the missing low frequencies. Someone on this forum has an Integrator, which does increase output by allowing a higher crosspoint to be chosen. I want one.

There are also questions begging in the phase domain, since the reflex bottom rolloff approaches 4th order. (The Stereophile graphs look more like 3rd order 18dB/octave.) My admittedly shallow understanding is that the high-order crossover will create ringing, which Thiel speakers go to great lengths to avoid at the higher crosspoints. I don't hear artifacts and wonder if the ear might be less sensitive to phase anomalies in the low bass. Perhaps others might comment on that problem; ideas welcome.

I am not familiar with the SmartSub 2.2, and haven't seen any reviews of x.2 subs. I don't know much about the development history of the subs, being after my time. Might the .2 designation be New Thiel? My SS1 and SS2s have Jim's (again patented) room boundary interface, which works very well to set the proper level based on proximity to walls. I have experimented with 1 vs 2 subwoofers and found noticeable improvement with 2. The 2-series crosses over in the high forties where directionality is thought to be hard to hear. But it seems to my ear that directionality is an issue. I would choose 2 SS1s over 1 SS2. 

George & prof - I think the low-pass 3dB down point should be @ 45Hz, the natural 3dB down point for the 2.7. There is no high pass / low cut filter in the XO, so the sub crosses to the woofer at the natural bottom rollin of the woofer. 

J.A - Thanks for asking. Good progress re comparative listening and measuring tools. Great appreciation for Jim's component choices within his budgets. I'm presently testing various mechanical upgrades: cabinet stiffening, stiffer driver mountings, XO panel isolation, etc. Advantages of hindsight include location of cabinet resonances via reviewers and more sophisticated instrumentation. I can address some of them, others are baked in. Many electronic components are chosen, but no crossovers yet. pm to you. 

George - thank you for this post. I used the 2.4 spec in error. The germane point is that Thiel's Passive xo does not attenuate the main speaker's low end - in this case a little below 35 Hz.

rwmeditz - I am interested in your further comments and comparisons about 3.5s vs 2.2s.  I am familiar with both, but don't have 3.5s for comparison and hope to align my analysis of pieces and parts with your sonic / musical experience. Please elaborate as you might wish.

Integrated amps have evolved indeed. What strikes me about rwmeditz's experience of "holographic image" is how the impedance difference of the 2 speakers might be operating. The 3.5 bottoms around 5 ohms from 100 to 1K Hz with rising lower impedance. It presents a rather simple / orderly load for an amplifier. The 2.2 bottoms around 3.5 ohms from 100 to 200 Hz with another dip just under 40 Hz, presenting a more difficult load. I wonder what you would hear if both speakers were driven by a big, high-current amp. Comments are very welcome.

On a personal note, I will be traveling and out of contact for a week. I am attending a master class in soundboard tuning from the Australian physicist Trevor Gore. My primary occupation is designing stringed instruments and Gore's pre-retirement workshop will share his lifetime of understandings, tools and methods regarding what actually happens as the string excites the soundboard of an instrument - and how to  optimize the outcomes. Many of you know that I was building guitars before we diverted to speakers - stringed instruments continues as a first-love; and there are many parallels and mutually enlightening aspects shared among the disciplines. 

RW - your story strikes the core of our inspiration to "do speakers". Thank you for sharing. That emotional connection you reference is what we identified at the core of the minimum phase presentation. I have subsequently studied aspects of auditory neurology and psychoacoustics to enrich my understandings. Most of the world and nearly all of the technologists dismiss that very element (which you feel) - or at least its connection with phase integrity. You get it.

My present task is a rather difficult one - so many technical aspects go into creating what you have experienced; some are known, some are surmised, some are mysteries approached through trial and error, and some are unavoidable or changeable. I must unmask the limiters without reducing the successes. Of special help is your comment about the added volume needed to unlock the 2.2s. I will contemplate clues around that idea . . .

Of interest to users is that a large part of the puzzle resides in amplification. The 2.2 presents a more demanding load, and your listening experience sounds a lot like signal veil from a straining amp. That load is baked in to the 2.2, and gets worse with many subsequent models. I hope that you can borrow a "great" amp and report your findings to us here.

Rosami - no firm news. The bankruptcy is still pending, so I am behind a curtain. I can say that I am enlisting talent to augment my limited abilities.

I can answer your question regarding hot rod kits. Some general recommendations will flow (and have already) from the work. But the nature of the beast is extremely sensitive to particulars, and all aspects interact. So the "kits" will be very specific. Since custom parts are involved, purchase from Rob will be required.

No news as to timing. But yes, news regarding progress. I am working.

Rules - zero phase rings before and after a transient event. Minimum phase does NOT ring. It's worth some study

Hi guys - just back from the soundboard voicing masterclass in Denver. Physics rules the technical arts. I'd like to affirm these recent comments about the 3.5 and the progress of Thiel's models in general. As this inner circle understands, the goal did not change. Each model approaches the same goal of translation of the input signal into acoustic output without degradation and within its constraints, which are primarily bass and power. The 3.5 was the 5th iteration of the model 3, and with it came a leap in magnet geometry in the woofer and midrange. Subtle distortions were reduced by an order of magnitude. The 3.5 also saw the final form of the styrene x tin foil 1uF bypass cap and refined the ratios and form of compound bypass nano-caps. These technologies were, to my knowledge, not used by anyone else at the time.

The 3.5 was a clean, settled design which would have run longer, but for the advent of peak-intensive music. Wilson's Audiophile Recordings, complete with uncompressed cannons, along with other exhibition-style recordings simply over-drove the 3.5, well beyond what vinyl / analog ever could. The 3.5 drivers were by far the most robust we had ever used and we replaced, by far, more of them than any other model. That led to ported bass in upper models which, I believe, Jim developed extremely well. His alignments produce proper augmentation without the B&W bump and dump. But, as 3.5 fans know, the ported alignment introduces more phase shift and steeper low-end rolloff for a less natural result.

Each generation of each model makes incremental progress, led by its new driver development and then shaped by its enclosure and crossover implementation. The 3.5 was a breakthrough product and still holds its own today, nearly 35 years on.

I thank all of you for your continued appreciations.

JA - the disciplines are related and knowledge usually cross pollinates. But my goals in Denver relate to acoustic guitar and piano design. Making the music is my heart's delight.

JA - one dovetail of the XO project with my musical instrument work is that I use the same listening, mic, software and equipment setup for both purposes. So, I am taking measurements while making the guitar, and making archival recordings for later comparisons. I can compare those progress measurements against playback measurements in the same place in the same room, for a very fine-tuned reference standard. 

Hello Sandy - good to see you here. This forum is my sole on-line activity. My experience tracks yours. In all the years of demonstrating and exhibiting 3.5s, we (Thiel Audio) NEVER blew a driver, even when purposely tempting fate with high levels for extended time. But, news from the field mounted, we warranted many drivers, especially midranges with burned coils. The story was almost always moderately powered amps - 50 watts is enough to fry voice coils, but not enough to stay clean. Program material was almost always soundtracks or other oo-wow exhibition pieces. Rooms were usually large - required power rises as a cube function of size - it escalates very fast. And the "felons" often had compromised hearing - think rock concerts. A core liability of first order slops is vulnerability.

The first replacement driver was free under warranty, with a lesson and warning. If they blew another driver, the dealer had the option how to charge - we supplied the driver to the dealer at our cost. After that they were on their own. Under normal use with normal program material, a driver should last many years. But, "normal" has drifted to much higher demands than the old days.

Sandy - thank you for your remembrances. The 03a was the real beginning of it all. It upgraded the 03, which was Thiel's shortest run in history: one year. The 03 was our first product to address time / phase with first order slopes. We struggled nearly two years to minimize all the problems which coherence made glaringly obvious, but which melted into obscurity with higher-order crossovers. We launched the 03 in early 1978 to an encouraging audience at the Winter CES in Chicago. But back home, we were still working around the clock to identify and solve "problems".

A root problem turned out to be diffraction, which wasn't much on the radar in 1978. We nested the drivers in heavy wool felt, and brought them closer together, virtually transforming the speaker, without changing the drivers or the crossover. By the end of the year we had the 'a' iteration which we introduced at January CES 1979. The illustrious Harry Pearson, editor of Absolute Sound and to many minds the founder of "high-end", gave the 03a a glowing review, citing its many unique strengths and only its ultimate power handling capability and his unease about the equalizer as his mild reservations. He finished with something like "Thiel's 03a is a tour de force; I expect to hear more from this emerging designer." That is a paraphrase from memory.

I suspect you're right about the early 80s at Havens and Hardesty because the CS3 was introduced January 1983. Here's a quiz for the group. Has anyone ever heard of the 03b?

Correction: the CS3 was introduced at the Chicago Summer CES in June 1983, which is part of the 03b story.

Sandy - I hope that name is OK, since that's your hi-fi identity to me.

I credit Harry Pearson with creating high end out of its hi fidelity foundation. Hi Fi had been generally academic, formal and engineering driven. Hi End was generally young, entrepreneurial and music driven. Thiel straddled the two worlds more than many high end companies. But Harry was the first to take live music with all its nuances and psychic / emotional / intangible hooks as the Absolute Sound, the real and final reference for our work. He and his team created much of the vocabulary which we still use. Many related industries lack such vocabulary or frame of reference and therefore struggle with having to prove their work, satisfy the textbooks, and so forth.

The radial wavy plate midrange of the x.7 presents a better launch-pad for the center-mounted tweeter than does the previous midrange cones. The cone double diaphragm allows a much shallower front profile than standard, and is therefore more benign than some other coaxes. But off-axis listening does change character. The x.7 wavy plate seems immune from those effects of geometry in addition to being an order of magnitude stiffer.

I'd like to share my thoughts regarding spikes; I performed the investigations way-back which led to our incorporation of spikes in the CS3 in 1983. As Sandy said, the chief problem is recoil. The effects, however, depend greatly on the floor system and are generally more evident in odd ways.
On carpeted floors the speakers sway and those motion effects become quite significant at high frequencies. A firm carpet may exhibit little to no symptoms, but a foam pad or bouncy carpet may allow considerable movement. On the other hand, coupling to a bouncy / resonant floor may introduce greater problems than allowing the speaker to float on the carpet. Some floor systems can be stimulated into resonant modes by spike-coupled speakers. Those vibrations can often be felt via bare feet in the listening position, offering clues to what's going on.

The sonic effects seem to congregate around image stability. The subtle spatial cues that convey image specificity can be scrambled by a moving speaker. Vague imaging, especially front-to-back depth, can be caused by unstable speakers.
Early Thiel spikes , (up to at least 3.6) were 3-point, non adjustable - explicitly defining a plane of contact. Those spikes had 3 lengths for various tilt strategies to arrive at 3' launch point aimed at your seated ears. Later models adopted 4-corner adjustable spikes. Caution: if those 4 spikes are not very carefully adjusted, problems could result from the insecure foundation.

This afternoon I tested my 2.2s under development. My floor is glued 5/8 + 5/8 plywood on 2x12 joists on 16" centers. That's stiffer than many domestic floors. My covering is commercial (old ski-lodge) tight, hard rubber backed carpet squares. My spiked speakers transmitted a little more vibration to my bare feet. I sum to mono in my preamp and pan left or right speaker for comparison using self-recorded material - in this case Dana Cunningham's Dancing at the Gate - with lots of detail. The spiked speaker produced more subtlety, nuance, detail, complexity. Highest single notes sounded more dimensional and more ambience was apparent across the range, with bass decay remaining more musical for longer times.

The improvements could be reliably noted in my setup. These are the kinds of performance particulars which I hope to increase with upgrades. However, I must note that this incisive precision is not always appreciated. Many manufacturers purposely make cap bundles to spread out such transient information; and I suspect that many listeners would find the added detail to be a negative, especially with recordings that lack the spatial and ambient clues that add enjoyment when present.

To batmanfan's original question: outriggers will do more than plain spikes. And I don't know the configuration of the bottoms of his speakers. If they are flat like the 1.6s, then some kind of feet are almost certainly an advantage. (I have seen marbles set into the corner sockets.) Threaded feet allow more precise bearing. And if your floor is resonant, then some kind of isolator pad might help decouple from those resonances.

Have fun.Tom

Sandy - I'm having fun sharing this stuff with you all. You're on it - most speakers scramble time / phase, sometimes by many hundreds of degrees. The ear-brain does a commendable job of reconstructing those wave-forms into intelligible music or sound. But in Thiel, Vandy, etc. which keep the time/phase intact, those sway effects are certainly audible. I notice that I listen at lower level (-2dB) with the spikes in place, probably because the transient detail is satisfying without needing extra volume.

FYI: I am working on a whole package of mechanical stuff in the 2.2s. When all added together, I expect the mechanical upgrades to make an appreciable improvement.

Sandy - Here's a recap of model strengths and upgrade strategy. I hope those interested find value and that others can allow an old man his remembrances. The model 3 was our first best guess of addressing the most things required of a loudspeaker. A long-excursion 10" 3-way to fill a decent sized room at decent sound levels. On the heels of the success of the CS3 (1983), there was an expressed interest by supportive dealers for a little sister at lower cost for smaller spaces, larger than the 04 / 6.5" 2-way. The CS2 was developed in response and in communication with dealer feedback, much like the 02 / 04 had been. Now, in the used market, cost is no longer a differentiator, but, the personality is distinct. The smaller diameter drivers produce a more delicate, intimate presentation, and tend to excel in low-level conditions.

As to which generation of each model to upgrade, there are real issues of driver availability to consider along with technology maturity considerations. Rob and I discussed driver repair and replacement options. The CS2's drivers are no longer available, plus they were prone to destruction by over-driving, and they were basically customized, ordinary drivers. Also, the solutions developed for the CS2.2 system were qualitatively more successful than the original CS2; each generation stands on its predecessors' shoulders. The 2.2's drivers are quite special. The tweeter was developed in-house for the CS5 and includes the first generation of shaped magnet structure, copper sleeves, custom dome geometry and so forth - a breakthrough driver in its day. The midrange also contained significant customization including its fibrated, poly-coated cone. The woofer was the first iteration of the dual cone which graced future drivers up to the x.7 star geometry, as well as other leading edge motor developments. As the production designer, I love the 2.2 as the first product designed in conjunction with our in-house CNC capability. Previous products had been partially migrated onto CNC, but the CNC-native 2.2 went all-out with interior bracing, in-house milled passive radiator, flared mounting rings, interior driver relief machining, and matched grille / baffle contouring - all things previously unfeasible via manual methods. I also gravitate toward the final iteration of separate (non-coax) drivers for their place in history and wave-launch characteristics. For the seated listener the discrete drivers produce wave-forms free of the "moving waveguide" effect of the coaxes. And, personally speaking, I own the pre-production prototype CS2.2s and have used them in my work of recording evaluation for nearly 30 years, in comparison with other mixing and mastering systems . . . I know the 2.2s very well, allowing subtleties of changes to be readily apparent to me.

To summarize, the 2-series is preferable to some listeners over the 3-series, noted by the long-standing audiophile affection for the 2s. And, there are a lot of them out there. And the 2.2s are the oldest (1990) product with reliable driver availability. (I had dismissed the older 3.5 (1988) due to driver non-availability, but would love for suitable replacement drivers to emerge . . . but lots of ground to cover there.)
Each generation developed its technologies as the company developed its capacity to implement the products of Jim's creative mind and relentless experimentation. I think you would be pleasantly surprised to hear an original CS2 from 1985, or any of the further developments. I hope you'll be blown away by the 2.2s now in the works. The foundational knowledge and solutions developed from their upgrade process will apply to all the models we might eventually address.

I know this post is long and some of it repetitive of previous postings. But I hope it adds some value to the conversation.
Tom

The Dynaudio tweeter was a D28/2 - 6 ohm. Rob has been looking, and is continuing to look for a drop-in replacement. Keep us posted on what you find.

As candidate drivers are identified I and/or Rob can vet them. Rob is learning to use Thiel's Klippel and I am learning to use my SpectraFoo to make critical comparisons between the design driver and potential replacements. I recognize that this task is many years past due, but Rob has only had Coherent Source Service for a year and a half and I've been onboard (very part-time) for about a year. Slow progress is being made. We are committed to you having your classic Thiels for a long time.

The difficulties of T/P are substantial and the prevailing wisdom from Canadian Research Council, Harmon, etc. is that the ear-brain doesn't care. We here lean toward the opposing opinion. And I am personally convinced that T/P as executed with first order slopes is worth its considerable effort.

Andy pointed out that the coincident driver eliminates the upper frequency lobing, plus Jim managed that inherent dispersion problem quite well via very low cross points. A seated listener at 8' or more distance gets a properly integrated wavefront, although the summed energy into the room has some potential unevenness. But I suspect that Thiel's polar patterns would hold their own against Von Schweikert and others who manage T/P via high order slopes.

I can't decipher whether Magico and other high-rollers are really dealing with T/P. It seems that such measurements aren't often included in reviews, or such products don't submit to reviews which do consider T/P. Lots of folks are using 4th (24db/octave) slopes to produce zero phase, non time-aligned hand offs. Zero phase means 360° (one cycle) revolution, so the signal harmonics are separated from the fundamental, but not completely scrambled. 4th order, whether active, DSP or passive does introduce pre-and post ringing and time smear at the XO point. Minimum phase is a term of art to describe phase and time in natural alignment considering deviations from flat amplitude response, as in crossover slopes. The lead and lag components add and cancel, producing no ringing or time smear, producing an added output with flat amplitude and phase with no ringing - allowing time alignment via physical offset.
Vandersteen and Thiel were chasing the same goal. Meadowlark seemed similarly aligned, but the only product review I saw was an early model, which didn't quite do what it claimed. Does anyone here know whether Avalon or other highly regarded speakers worked in the T/P domain?  

Unsound - thanks for the links. Indeed direct drive could be awesome with today's technologies. As I've mentioned, active was our starting point in the mid 70s. Wish I could find that first crude research speaker: 10" 3-way with 3 home-brew amps and low-level active XOs. We thought we could create a technical success, but judged the market to be unreceptive to the idea, especially from an unknown start-up.  

Beetle - Thanks for the reference products.
Jim did control out of band problems as you say. However, his major tool was to create mirror circuits for the relevant troublesome resonances - electronic anti-resonance circuits if you will, which cancel each problem in the frequency & phase /time domain. Most of the parts in those huge xo networks are series and parallel shaping circuits to tame the driver misbehaviors through many octaves. Beyond just having the chops to pull it off, a big problem is the parts cost, and the juggling act of what to spend to reach what performance plateau.

Rules - keep up the good work and keep us posted. Yes, the variability, especially with operating temperature, but also with elapsed time and use all require assumptions of what is average. We called that "average operating condition" as 1/2 hour into fairly vigorous playing. Note that the flat impedance curves created by the Zobel filters do a lot to minimize impedance variables.
I suggest more study on minimum vs linear phase. Minimum phase characteristic is where phase follows amplitude as each driver rolls off. The trailing phase and amplitude of one driver is inverse to and cancels the leading phase and amplitude of the other driver and nets to zero. Linear phase does not zero out and results in ringing in various ways depending on the type of filter.

Prof - I noted in their website talk that they are using high-order DSP filters, which do not produce phase coherence, although all the talk implies that they do.

In my music production consulting work, I routinely picked out edits in mixes by the pre and post ringing of high-order DSP filters, while listening through (coherent) CS2.2s. The producer would invariably respond that "you can't hear that", usually because of his $7 figure equipment investment and grammy awards. But I and my client could independently spot the edits and other anomalies by separately noting time-code.

I don't claim to know what Meadowlark is doing. The review I found of the Swift shows wildly ranging impedance and phase swings - but it is time-coherent. The Shearwater qualifies as a T/P coherent design.