Thiel Owners

Holco - the woofer is 4 ohm, then incorporate any zobel network values for actual woofer impedance.

Perhaps talk to Jantzen about unwinding.

holco - unwinding is very common practice. I've never done a foil wax, but it sounds easier than a bonded coil.

I'm not following your math. 30 Hz seems like a big change. The complex effective resistance will be different with foil vs round. Can you measure the actual crosspoint change?

We always wound oversized and unwound to taste in R&D.

Rules - I have a schematic derived from real unit by a DIY. If that would help, I can email a copy to you. Tom

Holco - nice work and thank you. Please remind me of your serial numbers for my reconstruction records. FYI: part of the reason for the "eager appreciation" is that the other parameters are all designed for maximum performance. The driver motor distortions are less than 1/10th of anything else I have seen. And the dual cone driver has extremely low breakup and the cabinet has very low resonsnces, and so forth. In addition, I believe that phase coherence allows the aural brain to accept the signal as real rather than reproduced and as such the essence of the music is accessible to the entire sensing / feeling person. You have removed a veil and you can experience the improvements holistically.

As a historical note, Thiel budgeted $1/3 for cabinet, $1/3 drivers, and $1/3 crossover. That formula gives more budget to the crossover than most companies commit, but our requirements are high. A large portion of that high cost is in high-purity copper and the styrene x tin foil 1uF bypass caps. And the other parts were also carefully chosen within their budget constraints.  A company must second-guess its way to its target market, and this particular cost-performance plateau is where we struck our claim. Also, the ultra audio grade parts we have today were not available ten or twenty years ago.

Here's a Sunday Morning Flight of Fantasy from the memory bank. About 1990 with the CS2.2 and a hefty hand-full of new products in the pipeline, I speculated about an opportunity to reach a different market. The company didn't want to go that way, mainly because of the confusion it might create. But here's a sketch.

Create a new marketing entity. My name was 'Perigee - the closest approach'. Products would be developed based on the foundation-chassis of emerging Thiel products. The CS6 and 3.6 were in development. Products would reach elevated levels of aesthetics and performance. Performance would be via ideas like we are now exploring: better parts, internal or external cooling, bi and/or tri-wire with customer assistance to get it right. The cabinets would incorporate some eased or rounded edges or panels, use of exotic materials, super-selected veneers with sophisticated scraped and French-polished surfaces. Each finished pair would be optimized in the anechoic chamber rather than simply verifying its performance within tolerances as in production Thiels.

A corollary might be super performance professional monitors in service-clad rather than furniture grade cabinets. We had some high-end production and mastering labs using our speakers by that time.

Markets would be upscale. Prices might be double or triple those of stock. Although I call the idea fantasy. Its presentation and discussion were real - but its feasibility was very weak. Thiel was already straining under its relentless 30%/ year growth pressure. Taking on more was  unthinkable. 

Just for grins, I googled 'Perigee Audio' and it seems that an Australian has undertaken a similar idea with another historic line of speakers. However, I suspect such an idea might work for some enterprising young company. Who knows if Coherent Source Service might stretch into such a role . . . Isn't life interesting.

 

I'm loving the enthusiasm I'm hearing from you both.
Beetle - indeed there are great expensive drivers and some including Vandy who took Jim's lead in using copper motor shunts. Jim invented the solution on first principles and only when applying for patents did he learn that the solutions were lying fallow in the public domain for a long time. Faraday Rings.
Holco - your coil wire is ERSE or Jantzen 4-9s round and bonded, so your foil coils will be different in geometry and bonding only. Your comparison of the round vs foil is of great interest to me.  

Pops - It's interesting that the CS6 has not come up in conversation or as an upgrade candidate. I learned that there was an internally designated CS6.1 with an upgraded midrange (?) and XO. Rob might know what you have. And I, of course, would be interested in what you find out.

JA - I want to express my appreciation for the group you have assembled here and its productive and cooperative form of engagement. Life is an interesting journey indeed and some gathering of energy such as here can change its course. As most people know at some level, family business is a peculiar beast; rarely is it simple. This group has provided a focus for me to re-approach unfinished business. I thank you all for being here.

Beetle - what a pleasure it has been to have you behind the curtain, questioning, researching, sharing . . . as we have made progress in this delicate matter of re-imagining this product statement of another person and circumstance - especially one's deceased brother. I am pleased with our progress and our results and now approach my own modifications and further investigation with the concrete knowledge and feedback that you have provided. Onward.

Andy - I wish to honor the classic Thiel naming system where a new numbered generation is accompanied by new drivers, generally all of them and usually a new cabinet - a new product on the shoulders of its forebears. Beetle's product remains firmly a CS2.4 with a twist allowed by changing circumstances of time. I don't know what to call these modifications and I welcome all ideas. The working IDs have been Level 1, 2, 3, etc. Beetle's upgrade is at level 3. Functional but oh so prosaic.

All - the presence of an audience here has encouraged me to devote focused time and resources to this project which simply would not have otherwise happened. This group has produced a pair of CS3.6s and an SS1 and SW1 and another SW1 on the way. I'm picking up a pair of CS1.6s this Sunday and have bought a pair of beater MCS1s for research and parts - these pieces in addition to the 4 PowerPoints which were motivated by the project. And my design studio now has a fairly well tricked out electronics and measuring nook for the ongoing work speaker. And it all feels right. Thank you all.

  

Ken - The CS5i is for Improved. The 3 bass drivers were re-engineered and the two low woofers were loaded with a central plug rather than damping mats. Crossover changes also smoothed things out in the lower midrange and reduced the upper frequency impedance rise for an easier electrical load. At the time there was an upgrade kit. Rob reports that the CS5i has virtually zero problems in the field. With proper amplification to drive the 2-ohm bass load, and a room big enough for proper integration, the CS5i can be wonderful.

The CS5 tweeter was our first completely in-house design with fancy motor tricks and an engineered aluminum dome - on the shoulders of new technologies first developed and tried in the CS1.2 tweeter. (The CS1 series became a trial horse for new technologies, much as the CS2 series became a trickle-down beneficiary of upper-end solutions. The late 80s is when our co-development relationship with Vifa flowered which served both of our companies very well for over a decade. We conscripted the CS5 tweeter for the CS2.2 and 3.6 for a very productive R&D cycle.

Pops - I have not yet made serious direct comparisons, but the two driver scenarios both have trade-offs.

The 3.6 vertical stack makes the listener vertical ear position more critical, the constructive and destructive lobing affects the frequency response. But, each driver gets a controlled wave launch from the stationary baffle.

The CS6 along with the 2.3 and up and 3.7, etc. with the coincident driver greatly solves the first problem. The tweeter is where it is designed to be regardless of the listener position. But, the tweeter sees the midrange cone as its wave launch. Even though its cone shape is engineered as a tweeter wave-guide, that cone is nonetheless moving. I don't really know whether that launch is more or less problematic or just different. I know that the coincident PowerPoints are uncannily integrated. And I also know that the CS2.2 tweeter (3.6 and 5) sounds different and lovely.

Who knows? I hope that by this time next year we will all know more.

tomic601 - I would be most appreciative to learn the nature of the 2.3 upgrades. Beetlemania and I are working on the 2.4 and I have heard rumors that some folks have plugged in 2.4 drivers and XO to the 2.3 cabinet. If you have first-hand information about such an undertaking, it would help me assess the territory.

Jim - please keep us apprised of how the 2.3s do for you. The original 2.3 coax was the pioneer coax. Do you know when its upgrade happened?

My original prototype Brazilian Rosewood O2s from 1976 we given to our parents as an early appreciation gift for their support. They are in storage at my youngest brother's house in Virginia; I'll get them this summer and, of course, soup them up a little. The O2 actually started the audiophile ball rolling. The O1 was a 10" 2-way Rock-n-Roll Boogie Machine with 3rd order slopes and huge magnets for high efficiency and the EQ for 30 Hz bass . . .  but not much finesse. Some early supporters asked if Jim could design something more refined, which became the O2 6.5" ported 2-way with second order slopes. Quite lovely for its day. I will apply the boundary-layer airflow technology under present development to the port and driver circumferences, add some absorption to the baffle to reduce diffraction and upgrade the passive components. And I bet we'll have a sweet little bookshelf monitor.

Thank you,
Tom

Beetle - patents are in the works, so no substantive news yet. I am beta-developing the technology using the CS1.6 as my test mule. The 1.6 / 1.7 slotted port is unusual in its airflow dynamics. Ports care where they are placed in their enclosures. The 1.6 long slot side-steps some of those placement issues by being placed in the lower chamber at large. So its airflow dynamics are more complex than usual, since the broad placement creates a gradient of interface parameters with the enclosure resonances.

I can report first-hand that the slot works.
I can also report that the interface technology will improve it as well as improve speakers in general - a fluid dynamics solution for fluid dynamics problems.

Plenty more to tell when I can.

Beetle - I have heard A-B comparisons only with Tannoy speakers in a good large room with decent sources. These demos were professional, not high-fi. I would not say there was any particularly tangible imaging involved. What everyone including myself heard was a cohesive solidity in the harmonic structure of the sound field - natural detail and realness entered the scene, akin to the proverbial lifting of the veil where none had been evident. I would call the effect extraordinary and surprising.

Within the month I expect to have a customized kit of materials to apply to the 1.6s for direct A-B comparison and SpectraFoo analysis in my studio.

Beetle - the airflow technology addresses any and all airflow situations. A high-impact situation is port flow; in a sense that is low hanging fruit. There are many other situations that apply to all speakers. I am looking at back-wave implications for the passive radiator and all drivers. Bass frequencies move more air, but all frequencies have potential. The flair around driver circumferences may come into play, especially at coupling frequencies where the driver transitions from constant velocity to constant acceleration. Air viscosity and turbulence are prevalent in such transitions. We're looking at the rim between the tweeter and coax mid in my newly acquired MCS1, which shares the 2.3 coax, similar to the 2.4. The 1.6 has a significant tweeter wave guide in conjunction with an acoustic loading device to suppress the oil-can resonance peak. Hmmm, what if that interplay might be settled. 

Short answer is yes, many potential applications. The only audio implementation to date is a round port in a Tannoy monitor and augmentation of a B&W golfball textured port. My first application is the 1.6 slot to see what we learn in the higher resolution Thiel CS platform. By the way, the 2002 CS1.6 is better than I thought possible at $2K retail.

Holco - thank you for the report.
Regarding leakage - ideally you want none, but practically a small woofer enclosure leak doesn't do much. Are both channels the same? Of serious concern is if the woofer enclosure leaks into the midrange sub-enclosure. Push both the woofer and passive and see if the midrange moves. If so, troubleshoot.

Regarding foil coils. I await your comments on that change. News from the research front: foil alloy is considerably purer than wire alloy. As I've mentioned, Thiel used 6-9s ultra aerospace wire from 1978 onward. By 1995 that wire was no longer available and we used 4-9s specified as C10100plus. That seems to be the best wire currently in use, except for specialty boutique producers such as Cardas. I have learned that ERSE, Jantzen and other high-end suppliers now top out at 3-9s C11000 @99.9% with higher oxygen allowances. Jantzen specifically disclaims audible benefits beyond this C11000 grade, etc. etc.
My personal experience is to the contrary, having spent the summer of '78 comparing wire. Some auditory brains hear the difference. I am currently exploring a solution strategy.

Beetle - your FST coils and hookup wire meet the C11000 spec. But your FST coil winding was less than best. I await your experience with Cardas hookup wire.

Audio inductor foil meets better than C10100. It is as good as the original ITT 6-9s aerospace wire that I secured in 1978. So the generally accepted higher performance of foil coils might be attributable to some combination of  foil skin effect reduction, tighter wrap and purity of alloy.

Lotta mystery.

Rules - if an engineer with today's tools put his chops to it, the differences could probably be verified. It takes a lot of doing and my experience is that measurements only verify what we already learn through listening. I suspect the Klippel analysis system could show some relevant information. I don't know who might be working at such a fine level of analysis.

For the record, what Thiel would do is discern the improvement by ear, then subject the change to objective measurement. If anything could be determined to be harmed then it's "back to work". If no objective change could be measured, then it's down to cost / benefit discussion among the team.

Beetle -  the measurement that would likely reflect on what you are hearing is a radical zoom-in on the onset transient impetus. The shoulder will be cleaner / sharper. Similarly the top shoulder before decay will be cleaner and simpler. Deep knowledge of the interactions between capacitance, inductance and impedance reactenances and reflectances might predict what measurements might reveal - thereby guiding measurement and evaluation strategies. But when a company is developing products on less than a one per year cycle, such measurement luxuries must be budgeted. Over the course of tens of thousands of hearing / measuring cycles, we learned what to trust and what required further verification. Note that such engineering verification was directed toward analysis for future projects and products more so than optimizing the present product. Note the systematic thematic progress from 1976 to 2009. Lots of learning, standing on one's own shoulders, so to speak.

Holco - since you performed multiple changes at the same time, I wish to ask a favor. Can you compare SQ with / without the drain wire?

Thanks, Tom

Frank - your description suggests the EQ is not working. I suggest getting a test recording. In your room, the deep bass should be big and boomy, and the woofer and midrange should visibly pump much harder than without the EQ.

Holco - I appreciate you sharing your work with us. I am getting Jantzen wax coils for direct comparison with ERSE foil. Both use the same 5-9+alloy and both are wound on sophisticated tensioning equipment. Both are bonded, but ERSE is resin and Jantzen is paper & wax. Sweet development. Thanks. 

PW - Your listening position is far enough away to allow wider spacing, but anything over 10' gets dicey for center-fill. I generally consider distance to walls as more critical than distance between speakers. Your 5' behind is ideal, at less than 3.5' (±) the auditory brain tries to conflate the bounce with the primary signal causing smear. Great at 5'.Small side distance stimulates room modes as well as exascerbating the reflection bounce. 2.5' is very close to the wall. As an experiment, can you bring the speakers together?

I avoid same dimension behind as beside. If 5' behind, then perhaps 4' beside, leaving approximately 9' between.
I apologize in advance if you have tried and ruled out these ideas.I am quite interested in how you are using the PowerPoints. Please share.

PW - what is the ceiling height of your tunnel?

mrpostfire - Your room is large, which is great, but its dimensions present some potential standing wave issues. The average ceiling height is the same as the width and the length is double that dimension. That setup is less than ideal.

If there is any way to get your speakers farther from the boundaries, that might help. Also, opening doors is good, especially if near corners and/or center of long walls. As a learning experiment I suggest the following:Your width is fine, although I might try 8' between speakers to give 3.5 to side walls.Length in difficult rooms sometimes works well with golden ratios.
Listening position at 18.5' from wall behind the speakers.Speakers at 10' from wall behind them.That gives 8.5' from speaker line to listening chair.
Or reverse the room end-for-end if better for traffic. 
You are then closer to the near-field with generous room all around to delay the room modes relative to the initial direct sound.

If you can try this arrangement, please let us know how it works.

PW - in the absence of other responses, I offer similar advice as to Mr Fire. The room is a huge part of the playback experience and can often be directly addressed very cost-effectively. Your geometry would benefit from effectively shortening the room as well as convex diffraction / diffusing on the corners, especially where end-wall corners meet the floor. Pleated wool draperies are amazingly effective as is wool / horsehair (old fashioned) carpet under-pad. Let us know your progress.

Prof - here's a sketch of what I know. Please redirect if I have not addressed what you want to know. 

The fellow who claims to have designed the 2.7 as Thiel being "very demanding", had little to no input on it. He was interviewed for the project, but was not chosen.

Those directly involved included Kathy Gornik - as listener-evaluator and executive from the beginning and quite conversant in Jim's approach, values and methods. Likewise Dawn (now Cloyd) was part of the mix since she was 6 years old. Rob had been on hand and paying attention since the mid 80s. I re-hired Walter Kling, who had been part of the start-up but left for an architecture career, as part of my escape team. Walter was a manufacturing genius and had good, analytical ears, but did not intrude into Jim's space. Walter left when Jim died.

After I left in the mid 90s, Gary Dayton became Jim's assistant. Reports are that Gary had a good ear and that he co-evaluated sonic considerations with Jim, but he was not an engineer. Gary was part, along with those mentioned above, of the "Thiel team" that evaluated the outside work of new product development.

The 2.7 was developed as a joint venture between Team Thiel, creating the topology and cabinet particulars and evaluations, with the engineering performed by lead engineer Tim Gladwyn at Warkwyn Enterprises in Canada with complete design capability and the resources of the Canadian National Research Lab, among the very best in North America if not the world. The 2.7 was complete before the New Thiel owners brought in their own design team for the post-Jim products. The process I saw was that Warkwyn would develop a prototype built on Jim's research and prior components and topologies and coaching from Team Thiel. That prototype would be sent to Lexington for evaluation and be judged as some form of inadequate, and Warkwyn would go back to work. The expense was enormous and unaffordable. I was visiting when the final 2.7 prototype was delivered and was able to compare it with the 3.7 in Thiel's listening room. I would judge it a completed design, albeit more similar in components and price to the 3.7 than prior model 2 generations to their model 3 siblings. All previous 2s would have used a smaller diameter midrange to cross more efficaciously with the smaller 8" woofer. The 2.7 has a huge electrolytic cap bank to roll off the bottom of the midrange where the 8" woofer wants to cross in. 
You know how the two models stack up from your extended comparison.

So I suppose you could say that there was a core team at Thiel who were quite versed in what a Thiel speaker had to measure and sound like, but no one who could actually perform the engineering to get it there. 

Your final question gets to the nit of it. All the work was on Jim's shoulders. Jim did not have proteges studying under him. Although lip-service and disagreements surrounded that MO, in fact Jim was introverted, solitary and a solo inventing designer. Not only did he develop the products, he developed the research tools to develop them. And he was effectively the CEO, a very hands-on manager. Thiel Audio was an extension of Jim's interests, talents and considerable abilities.

Great effort was expended trying to identify someone to take on Jim's legacy. Kathy tried to bring in engineering talent. That failed precipitating the company sale. The New Thiel owners likewise pursued that path vigorously. I wasn't part of that search, but I trust that such a candidate was simply not available.

Prof - I heard what you mean. I don't know whether the difference is in the frequency domain - the mid / tweeter coax is identical. The XO is different enough that I can't make any direct comparisons, except that the 2.7 has a big bank of electrolytic caps in its feed, which would soften the attack transient. 

The MDF baffle of the 2.7 would also soften the transients, taking the edge off the attack compared to the 3.7 aluminum baffle and nacelle - cap.

I have not seen lab measurements of the 2.7, since Stereophile didn't review it. I'll know more when I get a pair of each for potential upgrade; that's at the end of a long list of older classics.

Guys - I should add something to the 2.7 vs 3.7 comparison. I have many unknowns in my knowledge equation - piecing together facts of history for better understanding. Beyond yesterday's thoughts, there are some manufacturing matters that may flavor the stew.
As you suspect, I have been comparing and contrasting various components. My workhorse is the PowerPoint 1.2. I know it well, it is simple (6"x2way), its room interaction is simple and predictable with its 45° launch which eliminates floor bounce, and its crossover has very little compensation due to the maturity of its drivers: styrene fillet in the woofer and total Thiel motors in both drivers. I have been auditioning 3 versions so far:
A: Original made in Lex with classic Thiel components = best of form developed over the years and used in higher-model traditional Thiel products.
B: FST1, made in China with a mix of Lex (caps) and FST (coils) parts.C: FST2, made in China will all FST parts and assembly - tested in Lex.Note: the FST components were co-developed with Jim and are intended as clones of Thiel's traditional parts. But later FST coils are known to be physically inferior, and the FST caps have migrated from propylene to polyester via Beetlemania's 2.4 experience and my PowerPoints. I would be most interested in examining XOs of Prof's and Brayeagle's rejected 3.7s. But, alas, such rigor is ethereal.
However, our listening tests are controlled, double blind and results are consistent among listeners. Results may apply to the 2.7 vs 3.7 discussion, since similar manufacturing history applies.
First, there is no clear winner of Lex vs FST. There are differences which stack up as FST being more incisive, cleaner and detailed - leaning toward cool and analytical. Lex possesses more cohesive solidity, naturalness and ease. Listener's preference varied. These results were surprising, but true and withstood swapping drivers and multiple tests. For discussion let's let the results stand and apply them to the 2.7 vs 3.7.
Flash back September 2012 when I heard final 2.7 samples compared to 3.7 at Thiel. Both of those were made with Thiel parts; both were tweaked engineering prototypes, the reference standard of each model. I, along with all others present, heard differences which I summarized previously, which I have been projecting on Prof, Brayeagle and others here. My projection contains flaws. Namely, the 2.7 manufacture never left Lex with traditional parts and the 3.7 gradually went to China much like the PowerPoint described above. So, I surmise that the 2.7s among you are potentially more similar to the 2.7s I heard than the 3.7s among you to the 3.7s I heard. I don't know your 3.7 serial numbers, but I project that the lower the numbers, the more likely they are to perform like the reference prototypes. One of these days I hope to compare an FST 3.7 to the original Lex 3.7 to evaluate differences. Until then, who knows what to believe.

Good morning, gents.

Beetle - I say that I don't yet know enough to make any calls of good or bad news. As time goes by I will compare notes with Rob and others and get more solid footing. I did not mean to imply that Lex was inferior. In fact I hear an overall rightness and musicality in Lex compared with an overall analytic precision in FST. Shades of allure. 

I am learning about specific alloys, materials and methods in the various parts and over time will be able to make educated choices among the extant parts. For example: remember our cap noise test which revealed no noise? Such tests, including listening, will identify re-usable vs non-re-usable parts. As I introduce upgrade parts, my knowledge base will grow. 

Such is the collective institutional knowledge of a company or designer. That knowledge is quite weak in Thiel-land today. But, progress is being made, albeit slowly. Getting started is the hardest part.

Your thoughts on 2.4 upgrade vs all-FST might be of interest to the group.

Beetle - thank you for consolidating your experience for us. It all sounds quite straight-forward as written, but considerable study and thought went into this trial and, as you know, considerable time elapsed while researching and considering this first trial build. This build is fairly high-end; although more money can always be spent for greater but diminishing returns. The upgrade parts cost alone is somewhere in the $1K range. (We haven't actually counted beans.) I want to commend and thank you for your enthusiasm, patience and careful listening tests with detailed notes and summaries, along with your good-natured willingness to try new solutions when something failed. (For the record, we are experimenting with heat management and mounting the resistors on aluminum heat rails created a subtle sonic anomaly.)

Among the points made above, I wish to comment on one in particular. At the end of the second-last paragraph, you address the "flaws in the recording" issue. This issue has been a constant companion during the entire lifespan of the company. Nearly every reviewer picks it out as a problem, often called out as a shortcoming of the speaker. Indeed there are very many very good speakers which do not exhibit this trait. At the very beginning, when co-developing the 03 as a phase coherent point source vs "normal" topology, we sweated blood over whether to take this plunge, knowing it would remain the core of our struggle forever going forward. We did and it did. Our central reason is philosophical: "the best speaker reproduces its input signal with the greatest fidelity in all its aspects." Phase coherence has been both a burden and a torch that led our way.

This "warts and all" approach is, as you know, not the normal approach, which seeks to present the most involving musical suspension of disbelief with the broadest range of recordings to the largest body of listeners. Everyone at Thiel learned to live with the "bare, naked facts" dilemmas, as did many critics, dealers and users around the world. But we still hear today a drum-beat of criticism around how revealing Thiel speakers are.

My central misgiving in this upgrade project is that it seeks to intensify that "music under a lens" approach. Every day I hear technical, mixing and mastering errors or shortcomings on recordings. In fact, in my life after Thiel Audio I have provided evaluation services to recording artists by identifying such flaws in their production mixes and trial masters (and that is with stock CS2.2s.) So the results and report of your upgrade journey gives me encouragement. I believe and you concur that by removing more of the subtle artifacts introduced by the speaker, we are more closely approaching the original musical event, making a firmer connection between the music and the listener. I believe that connection brings joy, even if it reveals some flaws in the music making process. 

So, thank you for your participation and thank you all for your interest, which has helped me take on this rather complex and challenging project.

Dhoff - I'll add some context to your location question. Mostly, the spot you choose is up to you. You keep track of locating the speaker in your space and dimensions are handy for doing that.

In the background, is that the issues in play are rather larger than which spot you pick. The driver integration, lobing, room reflections, etc. are more dependent on the room variables than on the measurements. The spot I use is the tweeter plate, which is the same distance from the listener as the sonic centers of the midrange and woofer, which are both hard to determine, but closer to their voice coils than the baffle.

Relatedly, the angle of launch into the room is of great importance, but lacks an easy numeric measurement. That angle is influenced by how far the listener is off-axis, and also how the mix of the various frequencies of the various drivers is delivered into the room and the listener. I tilt the speaker back vertically to account for my 42" ear position (against the 3' design position). That tilting also affects the sonic envelope due to floor interactions which are different than the design assumptions (seated on a couch at 3' ear height).
The fun never stops.

Jack - I'll jump right in since I have set up many rooms with CS5s. I suspect the root of your problem is having them too close to the wall behind them. The 5 goes down below 20Hz as a point source, so you are getting lots of back-bounce reflection for your aural brain to try to conflate into the direct wave-front. See what happens when you pull them out to 5' behind them.
Also, you probably know that they present a potentially challenging sub 2-ohm load at their lowest reach. If your amp is running out of current, that wreaks havoc on imaging.

Hang in there, when you get them right, they can be marvelous.
The folks here have good ideas on amplification and placement.

Jack, my first thought was what Beetle suggested, but turning the room around is a very big deal, and listening from the dining room is probably a goal . . . 5' is a lot, and somewhat more than necessary, but more is better to separate the direct from reflected wavefront. But 2' is not enough. The auditory brain tries to integrate sources separated by less than a few milliseconds. We are testing.

Your equipment is great and as you say, should not be the problem. However, bridged amps are less comfortable driving low impedance loads. The 7b has huge power, so matching impedance is probably more critical than power. I don't know the circuit of that amp to make a recommendation. Experiment and let us know.

I do know that my first approach to upgrading the CS5 would be to separate the bass drivers (3x 8" on 2 bass circuits) from the uppers with 4 identical amp channels running the pair of speakers. At the 400 Hz lower midrange crossing, the impedance is 4 ohms and rising slowly, which presents a sweet resistive load for the upper amp. The bass current draw would be sequestered in its own channel.

Your problem is the core problem that limited the life and acceptance of the CS5. Both at the time and more so in hindsight, I believe a 4 channel setup for the CS5 might have made it a success.

Andy - same coax but higher crosspoint to the 8" 2.7 woofer requiring 416uF series blocks. 400uF electro + 15 PP + 1uF styrene / tin foil. The lower 3.7 crosspoint needs 226uF as 3x75uF PPs plus the 1uF S/T. ( no Es in the 3.7 signal path.)

As I've mentioned before, Jim wasn't alive to apply his typical methodology, which would have been using the coax he was developing for the 7.3 to feed the 2.5 just as the 2.4 had been fed by the 7.2 and the 2.3 and CS7 were co-developed.
That jujitsu of dropping the 3.7 coax into an 8"model 2 and calling it the 2.7 is not Jim Thiel methodology, but rather a way to use an extant driver in a fall-down product, creating a model 2 more similar to the 3 in performance and cost.

You are right, big film caps are expensive. Note that the 2.4 upgrade requires only 42uF series feed caps.

Andy -  the circuits control the drivers for a net 6dB / octave slope all the way down, including the interaction with the top end of the port where another pole is added to the top end of the woofer xo. Various conjugate circuits are applied to counteract bare driver anomalies. Since the usable driver range extends more than 3 octaves beyond the crosspoint, drastic measures can be called for.

As I said, I doubt that Jim would have tried to cross that midrange with that woofer, requring so much attenuation. He would have developed a smaller midrange with a higher natural bass roll out, to better match the 8" woofer with less brute force required.

Andy- I sent my response before reading yours. Agreed, lots of competing priorities to juggle.
Tom

The impact of phase coherence is hard to communicate and harder to prove. In fact, the industry at large has proven to its satisfaction that it doesn't matter. I have spoken here about the personal experience and psychoacoustic study surrounding why I judge it to matter deeply.

The odd jujitsu is that the human auditory brain is so good at assembling-synthesizing known tonal sounds from the transient impulse data stream . . . that in an intellectual way, we enjoy the decoding process which we call hearing.
For myself, and the small minority of those who 'get it', there is a direct path to the core being when that sonic analysis-reassembly is not necessary, when the unadulterated musical signal arrives as natural sound. As Andy says, and as the physics supports, and psychoacoustics agrees, there is only one way to accomplish that direct stream in a multiple driver system: first order crossover slopes.

Now, the cat's meow would be to have drivers with broad enough range of resonance-free operation that the phase coherent wave-front could be coupled with a ripple-free frequency domain response. By the way, a larger budget would permit additional driver resonance control and more fine-tuned frequency response to get the best of both worlds. I believe that Jim's 0.7 drivers have that potential - the midrange goes beyond 20K without breakup and so forth and so on. It would take a visionary designer with youth on his side to continue Jim's work to get those results.
Magico and others at the expensive high end get various aspects very right. But as you say, for those who have tuned in to the joys of phase coherence, there is always still something missing in non-coherent designs.

Ekohn - They're not bookshelf, but you might consider PowerPoint1.2s. They are wall or ceiling mount, so they take no floor-space. Same driver as the SCS4 in an aluminum wall-mount cabinet. Their configuration minimizes boundary effects for extreme transparency.

Hi Kent - Dana's music comes from that place we all want to go . . . it goes by names such as life and love, and sweet inspiration.

Her inaugural album "Dancing at the Gate" is in many ways my favorite. One small piano in one small room with one pair of mics recorded to DAT and mastered to CD with no compression, equalization or effects - with the greatest of care. Close your eyes.

Thank you.

Kent - Yes, precisely.

Jitter is timing errors which were believed by many to exist beyond human hearing, because when translated into the frequency / tonal domain, the artifacts are beyond the audible range. The auditory brain must work hard to weed out the temporal artifacts of jitter, causing fatigue or other distancing mechanisms. Dither spreads them toward randomness and therefore more easily ignored.

In higher order slopes, the onset transient arrives in distinctly parsed segments, one for each driver. The ABrain does indeed combine them, using memory, later-arrived reflections and decay parameters to make sense of them - all occurring later in time after brain work. Live music comes all intact. Well recorded and made music played back through single speaker / headphones can preserve that all-intactness. Multi driver / higher order slope speakers cannot preserve it; although they minimize the deleterious effects by creating timing with few or no sharp discontinuities, saying the ear can't tell the difference between single or multiple arrival times. In a manner of thinking, they are of course correct - the ABrain is very skilled - unless you are a listener who appreciates the coherent rightness of unified arrival transients such as heard in real (non-reproduced) music and sound.

That particular component of "realness" appreciation is shared among many Thiel owners.

Reference Recordings are completely phase correct, as is your playback chain unless there is a wiring error. So the signal at your speakers is phase coherent and if the speakers keep it straight you can hear that proverbial pin drop. Enjoy.

Phase is used to mean different things. The signal reversal that ronkent mentions is actually polarity, where the entire signal is switched plus for minus. That is straightforward to troubleshoot. Put a 4 volt DC signal way up front in your signal chain and your speakers should push forward into the room. (except that some crossover types push some and pull the others. So, use the woofer as your guide.) The "button test" identifies whether the recording is right or backwards. This is also polarity (called absolute phase). Proper polarity will sound focused and reversed polarity will sound diffused. Negative polarity is now generally considered an error, but it happens.
The stuff I am talking about references the phase relationships among the harmonic structure of the signal. That's a real rat's nest which the user can't unscramble. Those ratty problems are generally ameliorated by speakers which scramble phase - higher order networks are already asking your brain to unscramble and there simply isn't enough processing power to further decode what's going on in the actual signal. Higher order networks are in that way forgiving - more recordings with more problems will sound better, or at least their evils will go unnoticed.

Please note that I recognize my opinions as being marginal. On a panel of audio engineering experts, I would be crucified or perhaps kindly tolerated. However, on a panel of aural neuroscientists, I would get at least polite inquiries regarding my observations.

Todd - I mis-spoke to a degree. Much of the best modern equipment is phase coherent, but there is plenty of gear that employs global feedback, steep filters, etc. which introduce phase anomalies at super-sonic frequencies, which are known by some to be audible. Also cable can introduce such anomalies. Big discussion with contentious disagreement.  So, I retract my global statement. Note that much modern state of the art gear is paying attention up to perhaps 200kHz. Reading the performance of square waves, impulses and step responses tells the story. Beware when a review makes excuses that include "beyond human hearing". To match the ABrain's discernment in the time domain (a couple milliseconds), you need a couple hundred kHz in the frequency domain. So, old-fashioned brick wall CD upper limit 22kHz filtering will introduce phase distortion that is audible.  The gear that you guys consistently choose is generally very good in its phase performance.

ronkent - that live album at Stony Mountain is a normally produced remote recording, pretty good as they go. Her studio albums have higher audiophile values. Above I described the first Dancing at the Gate, later Will Ackerman produced Color of Light which has his (as in usual Windham Hill) high production values, but uses 18 mics and lots of gizmos and signal manipulation to get "his sound". That is the album where Dana and I "caught" (on CS2.2 / Classe) numerous, ongoing plugs and other artifacts which Will with his $multimillion studio insisted couldn't be heard. Then Dana produced locally "Above the Fields", which is very carefully and well done and can take you there.

If you like solo piano and have a contemplative streak, I highly recommend Dana's work.

RonKent -
Re: "hi Tom,   thanks for the clarification on phase, absolute phase, and all those other phase things."To clarify, you did not err in your vocabulary. The term "phase" is applied widely to both matter of phase and matters of polarity. I was trying to clarify that there are various related happenings, which would be better served by using two different terms: Polarity to refer to the direction of the signal and Phase to refer to the inner relationships of various elements of the signal. The term Absolute Phase is often substituted for Polarity, which would be the more precise term.

Bighempin - if you contact Rob re the cap / nacelle on the 3.7, you might also ask the following: Does he know the serial number breaks for the 3 versions of the 3.7?  I have developed XO schematics for the original, revision 1 and revision 2, but do not know where they occurred.
If you find out, please share it with us all.
Thanks. Tom

Beetle - a 3.7 upgrade is quite straightforward since the drivers are so well behaved, there is not much circuitry in there. The tweeter caps could be upgraded to CSAs with perhaps an ultra bypass around the 1uF styrene/tin. Electrolytics to film for permanence. Resistors to Mills MRAs. Film feed coil on the woofer. More upgrade for less cost than the more complex 2.4.

You'll see more from me after the Thiel Audio bankruptcy settles.

Tom

Andy - I second your motion. As car guys say, the least expensive option is the one you have. So if you have 2.4s, then upgrading resistors and caps is a huge bang for the buck.

I concur with the 2 sub solution. The CS2 - 8" woofer just can't move tons of air. Two smaller subs definitely trumps one larger one. I use Thiel SmartSubs . . . just because. But they are noisy and not presently repairable, so I can't recommend them. If you can cross a pair of subs properly, everything gets big.

The 2.2 bass was our first passive radiator and as such it coupled to rooms better than anticipated and came out slightly under-damped and somewhat (1.5dB) higher in level below 200 Hz. It has the fullest bass of any Thiel product, which was warmly (ha ha) received by the public, but considered by Jim to be in error. Note the 1/3 octave Stereophile graph showing some excess bass. Notice also in the Stereophile review that the cabinet is possibly the quietest Thiel ever, including the new x.7s with curved panels. ( I have a fix for that 300 Hz and ringing, plus a hardening agent for the MDF baffle to increase rigidity. The 2.2 is the first (1990) cabinet designed from the ground up with in-house 5-axis CNC capability. I went wild with braces because they were so downright feasible! Note also the quality of the custom caps - those yellow styrene bypasses were from world-class German film / tin foil. The tweeter was our own from the ground up design for the CS5 - it is a powerhouse, even by today's standards. The woofer is the first iteration of the double cone with curved front and straight back. It is polypropylene with air core and works extremely well. That design became the basis for the present double aluminum with styrene fillet midranges. The most ordinary element is the midrange, but even that "paper" cone has polypropylene fiber reinforcement.

As you might guess, I don't feel the need to make many excuses for the 2.2. As Beetle mentioned, they are my workhorse which I use to critique recordings in the making.

Pops and others have expressed fondness for the pre-coax format. I agree that there is something simpler and cleaner in the wave launch. The coincident coax addresses a fundamental problem with first-order networks: vertical integration of the lobed radiation patterns. The coax solves it. But, IF you get your ear at the correct 35" up, the problem is solved at that listening position, obviating the need for the coax, which does introduce low-level anomalies of its own. The x.7 coax is better because the wavy surface spreads the tweeter-edge wave nicely.

Todd, you are not nuts. There are some significant strengths of the 2.2, and resolution is less in the the bass and midrange. The new caps and treatments will upgrade the overall performance considerably.

Beetle, the 3.7 XO pic you attached is for all 3 drivers. The mid and tweeter have separate motors (unlike the 2.4, etc.) with the midrange XO having the greatest part count (16 compared to 17 for woofer and tweeter combined).

Holco - Thiel used Jantzen and ERSE wire coils as equivalents.I have ERSE foils on hand and some Jantzen Wax on order.I will compare them directly, but not yet.

Rosami - there are hypothetical failures which could damage a driver. Other DIYs here might know specifics. A feed resistor on the tweeter, or even midrange, could short rather than going open (highly unlikely). Such failures would usually be visible, as in burned parts from current overload. Look at your crossovers and if they look OK they probably are OK.

Driver failure is generally caused one of two ways:A: lead wire fatigue from long, bold useB: voice coil burn-melt from over-load, almost always associated with clipping amp.

Anyone else have thoughts?