Thiel Owners

Beetle, I am glad to be here. My hi-fi life has been fairly well buried until recently. Nice to find you guys.

Regarding capacitor bypasses, we discovered in the 1970s what has now become fairly commonplace knowledge, that small value high-grade capacitor(s) in parallel with the workhorse will keep the upper frequency leading-edge waveforms, etc. intact at a higher performance level per cost than a single type cap. My brand knowledge is a couple decades old and lots of progress has been made, so I will be finding my solutions just like you guys. We don't have Jim's lab and expertise. Nonetheless, Thiel's development MO was to experiment with cap configurations via blind listening to rank sound quality among the combinations. Then the highest performance / cost solution was identified and subjected to rigorous analysis regarding waveform integrity, ringing and so forth. The winner was always a bypass pair or triplet. The rank of (affordable) cap types (at that time) from best downward was: Teflon, Styrene, Propylene (foil or metalized depending on current requirements), Polyester (Mylar) and electrolytic, with some variants such as Tantalum, oiled, etc. Mylar was a bargain because propylene was very expensive due to production losses, which is now cured. Notice that older Thiels contained multiples-in-series propylenes or mylars to make the value in order to reduce costs. (Eventually a wound film cap will fail in testing due to a thin spot in the film carrier, so cost rises geometrically with cap value.) Another thing that has changed is that Styrene caps (our favorite bypass) has been obsoleted in the marketplace. Teflon over propylene and/ or electrolytic has a high likelihood of success on those large value caps. Rule of thumb is to bypass at 1% to 10% total requirement, more or less. More for high current requirement. Example: 100uF EE + 10uF PP + 1uF Tef > 111uF value at much lower cost than eliminating EE. 

This whole upgrade enterprise runs contrary to the rigor we applied at Thiel Audio. But I am fairly comfortable flying blind since I do not have access or time to test with instruments. I am also fairly confident of my approach due to serious personal experience. I also will employ a few twenty-somethings including two young women with extraordinary hearing to supply observations beyond my present hearing acuity (I am 69 and ears suffer with age.)

That's all for now. I'm expecting 4 XO kits for my PPs this week from Rob. I'll open my cap investigation and layout next week. 

unsound, thanks for the leads. The SW1 in California has unworkable shipping costs and the SS1 doesn't match my 2-driver unit. Best fit is SS-2. Thanks again, T

J, the hidden problem that arises with an amp that doesn't double-down to 2 ohm loads is that the distortion increases dramatically when driven into overload. That dirty signal is what toasts drivers, especially the delicate tweeter voice coil which is still in the circuit below 200 hz in a first-order XO. Unless you have a small room and quiet listening habits, I suggest holding out for more beef.

FWIW - I spent about 6 months in intensive wire learning and comparisons last year. I felt connected to that exploration since I had researched and selected the original wire for the 03, Thiel's first phase coherent speaker in 1978. Thiel stayed with that wire and its successors throughout Original Thiel. It is CDA101 (5-9s, long crystal, oxygen free) in teflon, tight twist (originally 2 / inch, then 3/ inch. Originally we bought from ITT, the aerospace developer of the wire. When StraightWire began, they took over the audio marketing of that wire. Its successors served Thiel onward as well as many other manufacturers. Note that FST boards have a similar-looking wire, but it is CDA102 (next grade down) with sometimes less tight and tidy twisting in a less than teflon jacket.

In my re-investigation I sampled Cardas and various offerings of various configurations in various jackets. We listened blind, and I compared a suite of 8 measurements of each wire. Although I don't have definitive tests, our listening paralleled what seemed to be superior measurements. Most of the artifacts of various "lesser" wires occur below 20Hz and above 20kHz. But a sonic signature persists in audibility. Across brands and insulation type, stranded wire has a "forgiving" signature with an appealing bloom in the bass, with a bit of soft tizziness in the upper end. I would call it nice and warm. Solid wire, regardless of other factors, sounds comparatively simple, straightforward and lean. Somewhat counterintuitively, the high end sounds cleaner and more solid than stranded.
I cross-checked my findings with Stevem Hill at Straightwire who cross-checked our mutual observations and hypotheses with his physicist associates. I always incorporate what I study and strive to understand with what I hear, beginning with blind listening and progressing through a reduction of possibilities toward a small handful of contenders that sound good and measure well. That has always been a Thiel hallmark, we never choose "nice sounding" if there is evidence of technical glitches.

My solution uses a combination of 18 solid and 18 stranded in a particular twist. It behaves extremely well on the scope and in the listening room. I am hand-laying my own working samples, but at some time it will become available through Straightwire.

As a general rule, you can assume that anything you get from Rob at CSS is actual Thiel OEM, and that anything connected to an FST (glass) board is an Asian clone. In all cases I have found the Asian clones to be quite good, but a step down from Thiel OEM.

There is a lot more to wire than can be measured with capacitance, resistance, and inductance. Electromagnetic propagation interactions, electrolytic absorption and wire crystaline anatomy all do things. Some among us hear artifacts from those things. Some in the engineering arena think we're crazy. But in all cases that I have gotten a straight-ahead engineer to listen, they agree that there is something going on. Steve Hill thinks they can't admit what they hear because their understandings of the processes don't account for the differences. Therein lies the gate to audiophildom: we believe it matters if we can hear it, and the cognitive understandings must follow from the heard experience.

I believe that a major factor in the popularly-held fallacy that phase and time information don't matter or can be misrepresented in test design is that such measurement techniques are given undue validity by most, leading to most manufacturers implementing designs that will test well against a fundamentally flawed standard . . . supporting a vicious cycle that justifies and validates the fallacy. This statement is philosophy-speak for don't eat everything you're fed.  

Stereophile's measurement design is an example of what I believe to be a fundamental lack of understanding of the ear-brain and its response to whole-goods: that is sound with its tonal, spatial, transient and dynamic characteristics, to name a few, intact, unadulterated, coherent. Aural intelligence is a profound function of human existence and more closely related to the other senses and, in my opinion, the development of consciousness than is generally appreciated. 

Over the years the critics were very supportive of Thiel's designs and products. I believe Thiel products garnered more international design and engineering awards than anyone else in the world. We couldn't have asked for more support. However, most critics didn't really get it. If they had, they would  have designed measurements (applicable and informative for ALL audio products) which illustrated performance in all relevant domains. Instead, the tests support the prevailing wisdom which I believe to be partial and flawed.   

In this forum and elsewhere I have heard people wondering why Jim didn't prune an heir to carry on his work. The answer is complex but it includes the vicious cycle I site. Engineering candidates dismiss the reality of factors that fail or fall outside the measured paradigm. To summarize a very deep matter, all candidates state with full conviction that phase coherence cannot be implemented along with frequency domain success. (that's a period.) And if they could be co-implemented, then it still wouldn't matter because the brain doesn't care about it; ask anybody from the Canadian Research Council or any other proper institution of research or learning. And since it (fully coherent reproduced sound) can't be accomplished and wouldn't matter anyhow, then why would I (the candidate) risk my career and standing in the professional community to swim upstream against prevailing wisdom. Short story: Thiel Audio DID search long and hard for many years without success. The Tennessee Buyers committed to carrying on the work; and no small part of their reversal was based on their research which supported the conclusion I stated: forget it, get with the times, hire Mark Mason and get on with the prevailing paradigm. 

I suppose I'm ranting. I do really appreciate you guys for getting it. Very few people do. There are further involvements and interactions resulting from coherence that relate to the amplifier discussion above and many more aspects of sonic performance. In business it is impolite to blame the upstream signal chain for less than satisfying sonic results. But truth be told, a coherent speaker doing its (Thiel defined) job of absolutely representing in all domains the input signal fed to it has an impossible job. There are thousands of ways that a signal is corrupted from acoustic (or augmented) event, through the recording, storage and playback chain to reach the listener's ears. Almost all practitioners along the way default toward euphonic engineering, trading off against the non-important aspects, which muddles the mess nearly beyond redemption. I think that any, perhaps all, of you Thiel aficionados would love to have witnessed the thousands of hours spent ferreting out the contributing factors involved in creating an honest transducer. When the speaker is coherent then EVERYTHING matters. That's how we identified wire, magnet eddys, diaphragm propagation moires, diffraction and so forth and so on from the O3 development and onward year after year making progress toward an impossible goal of authenticity and integrity, knowing we (the speaker) would be blamed for exposing problems not of its making. Since the history is ancient, you may not be aware that Thiel put those elements (wire, etc.) on the table before they were acknowledged in the audiophile community. Many elements are still contested all these years later in the pro and academic communities. 

Back to Stereophile's tests. The 50" (etc.) standard could be circumvented for very little investment. Outdoor measurements are anechoic. Tilting the speaker to a ground-plane microphone eliminates floor bounce, And so on. The problem isn't about accomplishing the test, it's about ignoring the importance of all of the outcomes. Aiding and abetting falsehood is an entirely invisible and unintended consequence. 

I should get back to work. I hate end of year bookkeeping. Thank you for the opportunity to vent. 

An amp with zero loop feedback must be extraordinarily well designed to work for music. Feedback disturbs phase response. Ayre is addressing design from first principles and I would expect all their products to be very clean, especially in the transient-temporal domain. If it sounds clean it is clean.  But, the joys of power should not be underestimated.

I hot-rodded my 1990 Classé DR-9, considered a high voltage / high current amp. In stereo it produces100 wpc > 8 ohms, 200 > 4 and 400 > 2 ohms. Nice muscle and finesse. Strapped to mono it gives 400 wpc > 8 ohms, 800 > 4 and 1100 > 2, which means the power supply runs out of current before voltage. I use them in my studio where they are surprisingly cleaner strapped than in stereo at moderate levels into a moderate sized hard-to-measure space described earlier. Anyhow, I would have thought that the 400 > 2 ohms in stereo would have been plenty, freeing the other amp for other duties. But the improvement was dramatic enough to assign both amps to mono duties.

The current behavior of your Ayre is of interest. But, more power produces an effortless transient attack that really makes music work emotionally.

In my opinion amplification and all other elements of the chain are very germane. Thiel speakers reveal upstream misbehavior extremely "well". This battle never ends because only by ultra-resolution can we retrieve the inner detail of music.

Prof, my detailed response didn't show up. Hmmm. Bottom line is that the height change from feet is very small compared to the sonic integration triangles at normal listening distances. 38" ear height is design target. Baffle motion from recoil forces is substantial for short tweeter wavelengths. Spikes work.

Nice to see all these Thiel pixels! My (memory of) 38" ear height relates to average seated ears. I don't know what height was assumed after mid 90s, but probably the same. Measurement distance is 3 meters with an assumption of that as minimum listener distance. Notice how little tilt would be required to adjust for spikes at 10' to 12' assumed distance. Rob, or anyone, please correct me if 38" is incorrect. One day soon I will have spectrum analysis equipment to verify such matters going forward.

Regarding measurement methods. Thiel did have a SERIOUS anechoic chamber at 20' high x 30' wide x 50' long (approx). Room reflections shelve at 200 hz. We began life balancing for 2db down below 200 hz and above 200 hz flat out to 20K+. That target was modified over the years to conform to general industry practice, AGAINST Jim's wish for FLAT is FACT, not opinion!

Jim had designed and built an interrupted pulse stepped signal generator which fed 1/3 octave pulses to the speaker (6' off the ground), @ 1 pulse acceleration, 1 pulse measured -calculated room decay before next pulse cycle 1/3 octave higher. Pulse sweep was 20Hz to 30kHz.

Jim also used rapid full-range sweep above 200Hz and noise burst / Fast Fourier Transforms, before such stuff was available on PCs. Oh, there weren't any PCs when Jim developed these tools.

All those measurements were correlated with free space measurements (speaker suspended from edge of roof 20' above ground) and half space (ground plane) with speaker firing into open space (woofer bounce) and firing down-angled with mic at position of first ground bounce, and buried in the sandbox for infinite baffle response to isolate edge diffraction effects. All these measurements converged into a well-rounded picture of frequency, time and power response where we knew how the speaker would interact with a room. Thiel believed that speakers should do their job of flat response including edge diffraction, and the room must be massaged to do its job of even support. I don't know where these presumptions landed by Jim's death in 2009. An earlier commentator above shed light on that; Jim may have migrated away from flat to align with generally held expectations. I always wanted to garner consistent assumptions from the recording engineering community, but we couldn't find consensus there. Everybody second-guesses everybody else!

Larry Staples was an early Thiel pioneering dealer in Louisville. He and others like him forged a new path in audio which became known as "high end".  Hello Larry.

JA, Please contact me at tomthiel@worldpath.net
Thanks,
Tom

I am interested in developing a "felt need" list of priorities for legacy Thiel upgrades. The 3.5 with its EQ might be a candidate, but I need more than a hunch. I can tell you that the equalizer was dear to Jim's heart as a good solution to the problem of deep bass and that it was abandoned due to dealer feedback including association with Bose's marginal implementation. Would a souped-up CS3.5 have value?

lrsky, Thiel has used air core inductors since the very beginning. And since the development of the 03 in 1978 and going forward, we used six 9s copper in all chokes and internal wiring. At the time we were an innovator with film capacitors and even in the woofer section we used or bypassed with film types. If Jim were redesigning that product today with its same cabinet configuration, I believe he would use six nines copper in the voice coils or replace the drivers with 3.7 type updates, but add titanium to the tweeter alloy and extend the response past 30K. He would source the best modern caps, not available or affordable then, and (I believe I would and he might) take the crossovers outboard. Jim wanted to have powered woofers in the CS5, if cost were no object and development resources were available. Jim wanted to make powered speakers from the beginning (our early pre-market prototypes were powered, more on that another time). By the way, that cast marble baffle requiring abrasive diamond machining could have been kept at a higher price, probably adding some damping component to the casting process. The CS5 bass drivers (first and third from the ground) could benefit from SmartSub technology, with an outboard amp and crossover to clean up EMF interference for higher pristinity (how's that?)

As an aside I would explore a 6 ohm nominal system impedance (rather than 4) so that more people could get better performance from less than stellar amplification, which was the biggest limitation to that product.  I would add some cabinet damping material to the wood panels where identified via Chladni Pattern testing which I now use in guitar design. Stuff like that. Do you still have those speakers Larry? Weren't they Brazilian Rosewood?

Tom 

unsound - Please note that I am not in a position to actually transform these ideas into reality, although I am dreaming some dreams . . .

Regarding multi-amplification, the signal-shaping was done within the amplification envelope, wherever it made most sense. The line-level signal was handled as two or three (depending on driver configuration) discrete signal paths, each with its own power amp. My vagueness relates to Jim's ability to manage different aspects (voltage vs current, etc.) in different amplification stages. Signal shaping including driver heat compression, etc. is integral to the whole system design. Jim was a uniquely talented circuit guy before we hit on the "let's do loudspeakers" idea.

As an aside, Jim's first patent was for a lovely head-amp circuit. Thiel developed, manufactured and Monster Cable marketed that unit. Variable capacitive loading via faceplate buttons allowed the user to "find" the best load for his particular cartridge / cable set. But higher output MC cartridges and then the digital revolution buried that product.

Back to speakers, thanks for those ideas. Fully balanced operation would certainly make most sense. Regarding high-frequency clarity, you have a point; even though the EQ has no active HF circuitry, there still exist jacks, circuitry, wire, etc. . . .

I could be mis-remembering (only 35 years), but I don't believe any Thiel product had dual binding posts from the factory. We tested extensively and found the waters far too mudied by bi-wiring. Cable interactions with the amp and crossover-drivers are extremely complex. Compounding that complexity via multi-wiring always caused more problems. Investing in better single cables always won hands-down for better sound. Controversial  for audiophiles; unanimously clear for our development team. Do you have a picture or brochure of bi-wired Thiels?

David Fox sent me a photo of the CS3 with dual binding posts. Indeed that speaker, #539,  would have been built months into the product life-cycle, and I would have been the person who made the silk-screen and punch jig for that plate . . . Hmmm . . . Be that as it may, feedback developed from reviewers, retailers and end-users citing problems that we couldn't replicate in the lab, and all turned out to be various forms of cable interaction anomalies. The coherent source architecture shone a light on problems that are just not audible under other playback paradigms. At some point (?) we quit the bi-wire game to mixed reviews.

As unsound says, there is value to splitting the signal, especially the equalized signal. For full disclosure, I am bi-wiring (present tense) my ceiling-mount PowerPoint recording studio room monitors as an attempt to preserve the transient edge of the tweeter from the deleterious effects of current draw from the woofer. Mike Morrow is making a cable where all conductors are braided together while having separate signal paths for the tweeter and woofer feed, so the entire bundle (12' long) experiences the same EMF environment, resistance path, capacitive and inductive envelope, etc. while segregating the signal paths. In my imaginary life, I would bring in my audio engineer super sidekick to measure, document and publish the paper elucidating what is learned in an A-B-C scenario of various forms of wire in my controlled, measurable, recordable situation. But, alas, life is short and priorities sing their own songs. In that song, Mike and I agreed that this solution is worth a try. I'll report my experience.

Beetle, I do recommend coming as close as possible to the original value, since that affects frequency balance and phase addition through the crossover. I am a big fan of double-bypassing and like your idea of adding the third cap to hit 27.93uF. That 0.33 cap can be ultra-quality which keeps the razor sharp leading edge transients from smearing.

I think that your first proposal built on CSAs is a better idea than the 0.01uF ultra with the SAs. CSAs seem to be a big step up from SAs, which were ClarityCap's best in their day.

If you do this mod, I would appreciate you modifying one speaker and perform whatever comparative tests you can between old and new. We can build a knowledge base here on the forum for the benefit of all, including yet to be identified upgraders everywhere.

beetle, the actual combined cap value is important for proper tonal balance and phase addition thru the XO region. I like the CSA +, + with the 0.33 being a teflon or other ultra-grade cap if possible. Ultra bypasses are far less important on the woofer, put more value on the coax feeds with less value on the traps (where the electrolytics are.)

Mono is good for critical analysis, either from a mono source or summed in the preamp.

oblong, Rob might have an upgrade path for those CS7s. Note: 5' from the back wall gives you about 5ms reflection delay where the ear-brain identifies it as a reflection and not smear of the initial signal. That's where the big Thiels come to life.

Tom

7' from the wall is a great luxury. Do you also have a high ceiling? Ceiling bounce is a real issue.

XO layout space is a real issue that will have to be engineered as we go. Thiel used good resistors, I don't remember the facts. You might trace the specs of your actual resistors. But I do know that we chose those carefully, considering costs. There is a theoretical and audible improvement from more linear resistors, but I don't know where the present solution fits in that landscape.

The general answer to the mod / value / cost of caps is that the smaller values must be higher quality to get improvement. We are fudging to get the most gain for the least cost by using less expensive caps for the bulk of the work and tricking out the smaller values which we can more easily afford. 

The woofer cap deserves cost restraint, the impulse transient is far slower for lower frequencies. If the 33uF is a Solen film cap, that's probably OK. Upgrading to an SA seems plenty to me. Note that Jim bypassed 'lytics (etc) with a small value teflon for bigger sonic improvement than swapping the whole thing for a film type. Out of the blue (no real research yet), I would suggest a 32 ± SA bipassed  with 1uF ± teflon as a high likelihood of success. FYI: Gary Dayton worked with Jim on these issues and may be willing to share some insight if you (all) want to rattle his cage. 

As an ultimate 7.2 upgrade, I would imagine outboard XOs to reduce size / layout constraints and take the XO farther from driver EMF and microphonics. I always heard added congestion when packing the XO into the cabinet.

I do not have a rule of thumb for how to account for changes in ourboard XO environment and I think any changes would be small enough to dismiss in the absence of research equipment and process.

0.01uF is very small for these voltage and current processes. It couldn't hurt, but . . .

I like your cascading scheme. I don't know the relative merits of CMR vs Mundorf.

Do not place caps on coils due to EMF field interaction. Keep the leads long enough for a heat sink while soldering. Use 4% silver x 96% tin solder or equivalent. I like mechanical / twisted connections under solder, flux the wire first. I don't think 40cm cabinet intrusion is enough to matter much. Do not change XO board orientation in the cabinet. Pay attention to potential wire buzzes when re-packing the cabinet.

Your room geometry accumulates pressure at the rear wall-floor intersection, and behind your head. That wall opening is great. If you can make some trap doors in those corners, you will clean up bass standing waves and flutter echo enormously.

I like where you are going. It might be possible to move the coil, if there is room in the cabinet. We could move the 0.15mH coil outboard past the end of the panel. That would require crimping and soldering lead extensions - not desirable, but possible. Damp all the caps with putty (BluTac, Mortite, etc.) especially if you stack the caps vertically, to damp microphonics. Note that the power capacity of the chosen caps matters-at least as much as original spec. And if the film-wrap direction is marked, match the original.

This kind of wrangling of details normally occurs behind the design curtain and the buyer gets what he gets with myriad value judgements / assumptions bundled in the mix. It is not for the faint of heart; the process will lead to critical listening / comparison and then value engineering assessment to assure that the gains are solid / uncompromised and the costs are no more than necessary. Eventually I hope to have upgrade paths for various legacy Thiel products that Rob can supply. We're doing this first trial in view of this supportive group. Thank you, beetle, for jumping in.

Pruning a brand niche is a very delicate endeavor. Imagine the confusions that result from differing implementations of various products within a brand. I will share that in the development of the CS5 in the late 80s, I wanted to go farther out on the quality limb and market that product in the $20K market (actually introduced at $9700 / pair.) I believe its ultra implementation would have been more than competitive at $20K. BUT, where does such a move leave the line / other products, niche reputation, etc? Perhaps two quality ranges? etc. etc. etc. As a small company, we couldn't cope with the variables. Toyota marketed their upmarket offerings in the USA as Lexus: same idea: different dealers, different image, different cost/performance ratios.

I have bought some nifty analysis software to guide me in this upgrade thing. I hope to find time to develop some solutions. The learning curve is cumulative, what is learned is applicable to other products. My first, stop is my PowerPoint 1.2s, upgraded via beetlemania's cap direction, coil replacement with legacy Thiel six-9s coils and wire, all in outboard XOs, and vetted with Metric Halo's SpectraFoo analysis software.

I have heard that there were reliability issues with the subwoofer amps and that Thiel was making in-house repairs at one time. However, I don't know the present state of subwoofer support. I suggest your friend contact Rob for advice. rob@coherentsourceservice.com

The modification project is indeed making progress, but there's nothing to report until there's something to report.

Mr. Bill- looking down on the room in bird's eye view, my floor standing speakers are 4' out from the south wall. But my PPs are not along that south wall. The "testing corner" is the south east corner. The right speaker is 3' out from the SE corner, hanging on the south wall with its tail near the ceiling. The left speaker is 3' out from that SE corner, hanging on the east wall. The floor to ceiling corner baffle is in that SE corner with the left subwoofer  under the left speaker, flush mounted through the east wall, extending into a roofed area. The right sub cannot be flush-mounted because the south wall is not thick enough. I face the corner to listen, which makes a very flat, even recording and playback response, due in large part to the non-rectangular room layout. The room is an L, with one leg going along the south exterior wall and the other going along the east exterior wall, providing long symmetrical walls for each speaker propagation. There is no wall behind me to reflect sound, but rather the crook of the L, and those remaining walls are all porous, providing enough air containment for spatial signature, but not enough to support room modes.

The tail of the PPs butts against the 8' ceiling for tweeters at 78" up. L-R distance is 4'. I would have preferred 4' out for 5'8" C-C or more, but stuff got in the way.

Prof - for my explorations of imagination I use the 02 as a path not taken. An entire business could have been built on developing and enriching that simple two way box. My unambiguous progress is in the realm of the acousto-mechanical: grille and baffle treatment, port mechanics, cabinet bracing, etc. Enter electro-magnetics and tings get much murkier for me. In the XO, my straightforward progress has been separating the XO halves and moving away from the driver. The drivers themselves represent middle ground. Substituting a modern Thiel (CS.5) woofer removes the normal distortions of normal drivers. Similarly, higher grade caps and resistors certainly solve more than they confuse.
But the realm of wire is truly deep stuff. I have ruled out silver entire or plated. Entirely silver is unaffordable; plating introduces problems, both audible and measurable. I suspect you would enjoy finding someone who really knows wire. The guy in charge of Belden's normal wire has created a high-end entry. It looks a lot like the Kimber stuff that blew our minds in the early 80s. (At $1,000 / pair foot in the day) it broke down our mental barriers to how wire can matter. I don't know enough.

Wire experiments are cumbersome, tedious and expensive. I have migrated to using the 02 with a CS.5 woofer (very linear full range) with no crossover and 4 sets of inputs connected to 4 lengths x 1 foot long, all wired to the driver. Each input pair is fed by equal lengths of Straightwire star quad Octave II. One cable twisted pair per input pair. A second speaker stretches my test samples to 8 pairs. A second speaker stretches my test samples to 8 pairs. I compare the 4 variants in one speaker and then run a FuzzMeasure sweep, which shows distinct (albeit small) differences. No test is perfect, but this one is fairly streamlined.

Wire differences, both audible and measured, are definite and have become instructive. But, far from definitive yet. More study required as the academics say.

By the way, guys, I'm looking for a few pairs of 02s to hotrod.

Welcome F1. I am speaking from personal experience and generalizations and do not formally represent Thiel Audio. The Ayre and Pass Labs are both extraordinary amplifiers from excellent designers. Each has its sonic signature, but either is considered top drawer.
It is worth noting that you blew 2 drivers with the Ayre. I suspect something is wrong with it or something else in your chain. Those coaxes are very rugged and the only failures I know are from distortion. If it were mine I would seek troubleshooting and repair. Also the power advantage of the Pass is significant. More power always runs cleaner. Definitely consider it. Look at the watts / ohms ratings. A good amp for current hungry Thiels doubles power from 8 to 4 ohms and again doubles or nearly so into 2 ohms. That pattern bespeaks a power supply capable of delivering current, which you need if you listen loudly to lots of bass or persistent dynamics and/or in a large room. Good luck.

The 2.4 and 2.4SE drivers are the same. The crossover has upgraded components plus the cabinet finish and the outrigged plinth make up the SE. Some of us on this thread are developing a Super SE with considerably higher performance parts which will become available through Coherent Source Service when the time comes. 

Enjoy, Tom

tmsrdg, power isn't the most illuminating word, current into low impedance is a subset of power and the operative limitation. You can find out the 2-ohm spec on your amp, which should describe its capability. But, as Larry adds another consideration, current delivery spec can be construed over various time windows, so the spec can lie with demanding program material. Larry is right that Jim was a bit gobsmacked by the subjective experience of CS5 (and all our big products) users running out of bass power. The CS5 came after the CS3.5 with equalized bass, which also drew bass current, but at higher impedance than the CS5. We used a 250W Krell powerhouse which performs as rated, 1000WPC into 2 ohms. Jim assumed the role of the speakers was to perform to its stated specs, and the amp must do likewise and so forth. When manufacturers fudge such claims, the burden-in-fact falls to the speaker for under-performing. All these years later, these matters are still foggy in the mind of the audiophile world. Awhile ago I mentioned that I wish Thiel products had higher impedance minimums - 4ohms minimum as beetlemania (?) corrected my 6 ohm nominal wish. Same animal. We would all be better served via honest claims for our products. 

Arvin - about the missing CS4. By the late 80s, we had gotten well established in Japan, which is (or at least was) extremely difficult as well as being a world market-maker. They wanted a state of the art speaker from Thiel, and orientals don't have "NO" in their vocabulary. Seriously. Those days were filled with newly-minted companies like Hales, Wilson, etc. offering $20K+ speakers. The late 80s were the good old days. Jim, however, was an incremental knowledge-based designer and had an aversion to jumping leagues, even though it would have injected significant and much needed cash. So, he designed his next greatest speaker beyond the 3.5. Of course the model naturally begged for the CS4 title. Well, it turns out that the entire orient considers 4 to be the number of death. Of course, our sophisticated distributor and dealers were above such superstitions, but cautioned that we must be concerned about all those other guys who may not want the sign of death in their homes.

The work-around was that the 3-series was a 3-way. The new speaker was a 5-way and therefore could leap-frog the pesky 4 problem and become the CS5, which it did. Now, the next product under development was the Thiel developed and manufactured coax product. I thought it should be a 4-way: coax to 6.5" to 10 or 12" plus passive, etc. However the 4 slot had been leapfrogged and the orient hadn't abandoned their 4 problem, so the 3-way coax x 12" became the CS6. Hmmm . . . lesser than the CS5. You guys will notice that CS6 driver complement is similar to where the 3.7 landed. Luckily for me, I had left the company during the development of what became the CS6, so I didn't have to lose sleep over these pesky matters.

In my opinion, had Thiel's new 20112 owners chosen to stay on the Real Thiel Way, a CS4 could have been a stellar performer and bridge the way to a CS5.2 as described earlier with SS bass. The CS6's primary limit is the 12" handing off to the 4" coax. Put a 6.5" in between and some serious seamlessness wow could happen. Of course you guys know this configuration as the CS7 series . . . It all got kinda confused by that skipped 4. In my imaginary life, I would have persuaded the new Thiel owners to develop the CS4 as a 4-way plus perhaps an exotic bass system between the passive radiator of the 2.7 and 3.7 and the powered inboard subwoofers of the CS5.2 (remember that from a previous post?). That imaginary CS4 bass got some passing fantasy time, but never made it to development. Imagine a true folded horn nestled between the floor joists, fed by a fitted opening on the cabinet bottom through the floor, and exiting via a floor grille in front of the cabinet . . . Yes, it requires build-in. Real life gets in the way.

Next time we'll walk down memory lane where no one has ever snooped and find the mighty Thiel folded horn heard a mile away.

We used a variety of amps in the lab and listening room. Threshold, Mark Levinson, Krell and Classé were always there. Unsound may be right about the TH12e, I don't remember. The Classe´amps were a pair of DR9s strapped to mono, which I still have, have hot-rodded and use every day. Their stereo rating is 8/4/2=100/200/400 and mono is 400/800/1100. An amazing thing to me is driving a pair of PowerPoint 1.2s, sealed box home theatre monitors rated at 89 db/w-m with 75 Hz cutoff into an easy room at "normal" playback levels . . . have you guessed? Their performance jumped a big notch going from stereo to mono with these amps. MUSIC IS TRANSIENT and impact requires power. These PPs are my mix and room monitors with stereo SmartSubs. (I'm still looking for a beat-up SS-2 or SW1). When people suggest your Thiels are sounding conjested on complex / demanding music, look to your amplifier.

An endemic design problem is second-guessing. If an amp designer assumes that low-impedance loads will be narrow-band / reactive, lasting instantaneous / short durations, then he can sign off on the particulars of his power supply. However, Thiel (and some other) speakers assume that a properly designed amp can deliver its rated output into resistive (easy) loads for longer term cycles (hard). The amp with enough power supply will love that load. The wimpy supply will starve the rails, clip the peaks, generally sound hashy and burn out drivers. I personally examined hundreds of burned-out Thiel drivers over my years (1975-1995) and every single one failed due to heat which means they were fed dirty signal. Conversely, over all the years of our listening and testing at well over 100dB (large) room levels, we NEVER burned out or otherwise failed a single driver. The specialty retailer and review magazines are supposed to do the educating, but that education is quite thin, in my opinion.

Jon, fiberglas is the cost effective standard. I tested and love genuine wool. Its naturally spiraled fiber works magic in converting pressure-motion to heat. We used 2 grades of SAE military pure wool felt on the back cabinet wall of the midrange enclosures to absorb the back-wave. Wool's performance can't be equalled by modern materials, which I call wishful engineering products. Wool fiber has a couple of problems in the cabinet. It acts very differently than fiberglas, so don't just replace the FG with Wool. And wool is inconsistent, so every cabinet would have to be trial-and-error optimized, which doesn't work well in a production environment. It is also hydroscopic (absorbs water in humid environments) and it's expensive. I will consider it if/ when we get to ultra-upgrade prospects, but wool must be re-engineered, not merely substituted.

JA - regarding PowerPlanes, there’s a short and a longer answer. The short is that the PowerPlanes were intended for home theater surround channels. The longer answer is hearsay from insiders, that wall or ceiling plane speakers never seem to generate 3-D imaging cues and their directional nature is strong. Especially in a ceiling application it is obvious that the sound source is from above, firing down. I don’t know which of the PowerDriver products came first or whether they were all introduced simultaneously, but I have heard that the PowerPoint with its 45° launch plane is considered the performance winner. The unique launch allows 3D characteristics. For the PowerPlanes, I suggest front wall mounting as the best chance for a good stereo image. That recommendation is based on knowledge that we find directional information unsettling when it comes from behind, overhead, or is ambiguous, such as front speakers firing inward from side walls. The PowerPoint floats the wave in the room with less source specificity.

I do enjoy the time I get in the studio. But there isn’t enough. I am comparing the wire configurations, but won’t be able to run all the permutations. I am comparing single and double runs of SW Octave II against Morrow 4 against OCOS without Zobels. Eventually I’ll break those out to individual runs for each driver as suggested by Unsound.
Beetle - can you report on parallel vs bundled runs? Your reports are a treat.

I am assembling a high resolution system with all Thiel speakers for developing acoustic guitar solutions as well as actual and consulting music production, in addition to developing potential crossover upgrades for classic Thiel speakers. By the way, all you A'goners, the upgrade project is proceeding, albeit a bit slowly due to the huge labyrinth of considerations. I have first-sample products on order and hope to have a report by the end of the month. I need an SS2. You may contact me at tomthiel@worldpath.net.

Proper imaging requires many aspects to come together, about which much has been written and we won't try to summarize. I will say that there are many ways to lose the information that results in a proper image. Cabinet effects including edge diffraction are a big deal as are lack of good transient response and even-handed timbral structure. But, in my opinion, phase coherence is the underappreciated core element. The ear-brain infers elements of space from time-arrival information along the harmonic structure. Jim's designs address that aspect extremely well. Years ago a respected reviewer took a pot shot at the Thiel (CS3 perhaps?) he was reviewing for its "ridiculous soundstage presentation". He complained that they made the classical recording seem like a birds' eye view, like he was in the rafters looking down. How silly! I sought out the recording engineer and, indeed, the mics were all hung overhead. We took that criticism as strong affirmation that we were accomplishing our goals of accurate image representation. We chose to not respond to the review, considering it bad journalistic manners and politics to call out a senior reviewer. To generalize, proper phase angle information is very spotty in recorded music. When it's right, the experience can definitely take the listener into the recording environment. Thanks for sharing, prof.   

JA, the reviewer was J. Gordon Holt the founder of Stereophile who had gotten somewhat cranky by that time. A company chooses its style. Our stance was one of appreciation for editorial coverage, especially since it was nearly 10 years before we ever placed an ad anywhere.  Stereophile Magazine became the finest editorial support any company could ask for, if the dared dream so well.

As the designer, Jim received such invitations and did very occasionally indulge. Jim was private and work-oriented and did not enjoy schmoozing. He and Kathy attended Harry's 10th anniversary party with the reining who's who designers, and another home visit around the CS5. JGH was present at various home visits at Larry Archibald's home / office after Larry took the reins of Stereophile. I was farther in the background, generally working 100 hour weeks to make more speakers.

Question: What was the outcome of the electrical drain off the driver baskets. Awhile ago someone (Thieleste?) did a comparison. All comments welcome.

Background: I am working on thermal stabilization and that might interface with electrical  draining if it is beneficial.
ThanksTom

The ScanSpeak spec on those woofers and subwoofer is 4 ohms.

Rob - I suspect that the only substantive difference between your 2.7 XOs and KYXOs is the wire. There may be small component value differences to compensate for the wire. Rob G. says all the 2.7s were built in Lex under his supervision. So your Tennessee Thiel's seem to be experimental. Let me know if you got the private email I sent to you.

Hello everyone. Thanks, beetle for the update and I appreciate whatever feedback you guys can give. There are so many considerations and development requirements, and so little time. But I am making progress.
Unsound, indeed our crossover cases would be quite tidy, along the lines of an 8" cube and probably as an option. Heat dissipation and electromagnetic field management are the prime motivators. I will be comparing internal vs external in my PowerPoint-as-studio monitor experiment. Making it up as we go.

Regarding CS3.5s. Indeed they are worthy and among Thiel's biggest game-changers. However, redevelopment of that product should include the equalizer as either unbalanced or fully balanced and with greater sophistication applied. The bass equalization concept rocked, but its implementation could use improvement. Big project for which I am not personally qualified. Any takers?
Also, the drivers were the last "modifieds" for Thiel. They had custom elements, but not Thiel's proprietary motor technologies or diaphragms - from-the-ground-up designs. I have chosen the CS3.6 for its more sophisticated proprietary drivers and more accessible solutions. The 3.5 is indeed a classic and I would consider collaboration with the right upgrade champion.

Ron, I have heard the 3.7 and 2.7 the day the finished 2.7 came to the Thiel music room, September 2012 - a room I had designed and built and in which I knew all acoustics and equipment. General thoughts: Over the years since the early 80s, the model 2 was a trickle-down product with the advantage of sunk-cost technologies developed for the 3 and with lighter loads re deep bass and ultimate sound pressure levels. An easier ride with a most-bang-for-the-buck mission. This time, outside designer(s) stirred the pot. My gestalt impression, knowing the long-term history and learning about and hearing this x.7 iteration of the dance, is that the 2.7 is voiced toward forgiveness. Jim was quite single-minded in choosing his solutions toward ultimate fidelity to the signal rather than "easier to take with most recordings". I found the 2.7 to be sweeter, but less resolving, even with the identical coax driver. Natasha Crane (my secret weapon) attended that session with her bat-ears and no hi-fi experience (a great advantage.) After the lights-out,  presentation she said "the 3 is way better, but the 2 is friendlier". 

My present path is to make a pair each of internal and external upgrade XOs for my PowerPoints and work with beetle's CS2.4s and learn what we learn. This project is front-burner, but shares priority with other demands on my time. The feedback of this group is helpful. Parts are trickling in.

There are substantial differences applied to the coax by the two designers and/or assemblers. I have Jim's 3.7 schematic + layout, but only a designer layout for the 2.7.  The 3.7 was all Thiel in-house and it seems the 2.7 was outsourced to ERSE-USA. To correlate what this community hears with the fact matrix I am assembling, I would appreciate detailed photos of the crossovers of both models to learn parts quality of the various components.  The simplest method for me is via email - tomthiel@worldpath.net. Thanks for your help.

Thank you all (I'm from Kentucky) for your participation here and via PMs. Here is a progress overview.
Indeed a high priority is to avert losses from aging electrolytic caps. New work will have ALL film caps for indefinite product life.
 
In broad overview, all series signal path caps are ClarityCap CSA-630 volt. They have excellent technical and sonic performance at justifiable prices. I am bypassing with the custom 1uF styrene that Thiel had made from best German film, now defunct. We will salvage your old ones.
I will be double-bypassing with 1% fractional value MultiCap styrenes. Thiel CS2, 3 and 3.5 used this trick with MultiCap's predecessor. Jim dropped out as of the 2.2. I'm back in for the quicker and quieter leading edge transient at what I judge as justifiable cost for this upgrade scenario.
Shunts to ground in Thiels (pre 3.7 and 2.7) are electrolytics bypassed via that 1uF styrene. I am going with a custom ClarityCap polypropylene having all CSA factors but with a 160 volt film for manageable form factor and cost.

As beetle has mentioned, all resistors are being replaced with Mills MRA-12, an excellent and cost-effective solution.

Note that I am aware that some might want to go farther afield with cost no object components and others may balk at the significant cost of my choices. My vision is to find affordable solutions that reach the next performance / cost plateau. This project is larger than may be obvious. This relatively major redrawing of the rules will require re-voicing both via measurements and listening. Beetle's 2.4s and my PowerPoints are the first line of that work. Our custom parts will arrive early June, when we'll engage the core work and begin substantive feedback regarding our judgements to date.  

Prof, thanks for the reference to Phil Bamberg. He is part of the picture and solution of the puzzle. Your quandary of 2.7 vs 3.7 shines a light on some design particulars of interest. The 3.7 has no electrolytics in any signal path, whereas the 2.7 has bypassed electrolytic midrange feeds and an unbypassed electrolytic in a tweeter shunt. Our target caps will surpass the 3.7 quality for both 2.7 and 3.7.

Of further interest is that later production 3.7s have Chinese made crossovers which seem to have some polyester caps where polypropylenes are specified. And those are built on printed circuit boards, whereas all 2.7s are Lexington made on point to point boards. I associate sonic "ease" with point to point implementation.

The purpose of this post is not to overwhelm with detail, but to share some parts of the emerging puzzle. I am hopeful of achieving significant improvements via parts quality and implementation improvements.

Best regards,
Tom

Robin, the CS series have enough internal enclosure volume to tolerate the added bulk of the upgraded crossovers. We'll produce upgrades both inboard and outboard.

Unsound, you raise a strong, well grounded germ of an idea.

But in my experience and knowledge, active crossovers are a fairly blunt instrument. Their downfall is that they generally assume a very simplistic model of the driver and therefore give a very generalized net filter effect without the necessary interactive nuances between the signal and reactive driver loads.

Those interactive anomalies can be addressed passively. My study of the progression of Thiel designs shows drivers that are better and better behaved as time goes on, requiring simpler crossovers. But they're still not perfect.

As I mentioned way back, Jim's first purest ideal speaker (never brought to market) included separate amps with active crossovers driving each driver with its Zoebel / Boucherot and other corrections attached to it. That hybrid of active / passive offers everything via control of the factors as well as control over or elimination of all the input variables that cause so much trouble (cable reflections, etc.) With the line level signal presented to the inputs, everything else happens "in the box". We decided, as a young, green, minimally capitalized company, that we couldn't afford to educate our customers, as well as the high risk of failure presenting such a wild-card to the market in 1976. Dealers hated the idea of an amp / speaker system that "just worked". Their very existence depended on addressing all those problems . . .

History validates our decision. Today, someone might pull off that feat. I offered this idea (indirectly) to the new Thiel ownership, who thought it quaint and idiosyncratic and didn't want to talk about it.

If a savvy group could buy Thiel's intellectual property (assuming it becomes available), and if the engineering talent could be assembled, I think this solution could make some real waves, especially in how it harkens back to the roots of the company. What if pro-audio / record-makers had such monitors? What if we didn't have to make excuses for poorly produced albums? What if the vast differences introduced by amplifier output particulars modified by cable variables . . . weren't in the picture at all? Just dreaming. 

Thank you for a walk down memory lane.

With very broadband response with steep slopes, a driver could operate in its sweet-flat-pistonic zone. I would need convincing that higher order slopes could be "corrected" to actually produce minimum phase additions through the crossover. I've seen failed attempts, but never a success to my standards. It must pass a square wave right through the crossover, to be phase coherent. If not, then not.

A complicating factor is that unbelievably small anomalies of resonance, less-than-accurate transient behavior and so forth, are readily apparent in a minimum phase system, whereas they aren't even noticeable with higher order crossovers. There's the two-edge sword that Thiel lived with during its entire run. I believe that acuity of perception is caused by the ear-brain interpreting the sound-source as "real" rather than an electronic reconstitution. 

Remember, I am not an engineer. I would appreciate being educated as to how the hybrid might alleviate the need for 6dB roll-offs.

Most folks agree that DSP is the future. The DEQX site is very impressive as are their reviews. The cost do "do it" actively and especially digitally is a small fraction of analog costs. And precision can be had. I note that ATC gives no phase spec, nor do I see relevant claims from Lyngdorf, but I haven't looked very hard. DEQX may really be doing it. If I were starting a company today, I certainly wouldn't be going all analog/passive! When I find time I'll share the Lexus / Thiel story.

In the late 80s Toyota was building their Georgetown Camry plant and Thiel, IBM, Trane, Square D and Toyota were core members of the University of Kentucky's Advanced Engineering Initiative. What a trip. Also, the CS5 was a big deal in Japan. The chair of the AEI stirred the pot and we began exploring with Toyota a premium audio system for the not-yet-introduced Lexus. Their key people, including Mr. Toyoda, visited our plant, and we developed a proposal. Jim insisted that the best solution was integrated amp / speakers. Toyoda wanted Thiel speakers under the Mark Levinson Premium Audio System umbrella.

The project could have put us in the big-leagues, but the costs and risks of development were beyond our capacity. In fact, we were developing new products at break-neck speed and were production self-limited at 30% year on year growth with qualified dealers waiting in the wings . We couldn't handle it, even if Toyota were interested in our integrated solution proposal.

After Thiel Audio, I was involved as a consultant in a couple of Japanese  co-development ventures, and am confident that we made the right decision. Big Japanese corporations are a different animal than small American technology companies. 

Prof, obviously you need a bigger room. Build to golden proportions at about 12'3 high x 20' x 32' with "vents" near the corners (windows or doors work) and you'll have a room like Thiel's listening / optimization room and you can keep your 3.7s and join the upgrade brigade for even more musical satisfaction. What's a second mortgage compared to such bliss?