Thiel Owners

Roxy - I bought the Birdseye pair on Long Island for future use. Thanks for the lead.

duramax - thank you for the base offer. One of these days I hope to get that far; for now there is so much work to update geriatric products.

It doesn't get much if any better than Pass in my book. Presently beyond my reach. Bill Thalmann of Music Technology, and formerly Conrad Johnson, is souping up my pair of Class´ DR9s to a performance level far beyond their stock form. That's still months coming, but considering his previous work, reputation and what I heard last week that he did with my Class´ DR6 preamp - I have high hopes and expectations.

What Bill is doing with the DR9s is to take out the 'balancing' circuitry at the inputs which is op amp based. Its only practical function is to allow XLR to reduce noise in long interconnect runs. Otherwise, that circuitry is of lower quality than the rest of the amp. We  gain room for better (very good) caps. CSA / CMR / PURITY. He's upgrading some storage caps and some critical resistors too. He can optimize bias, etc, by committing to bridged-only operation. When bridged the DR9 delivers 400 / 800 / 1100 watts into 8/4/2 ohms. Bill can get the noise better than new. And so forth and so on. I ordered the relevant ClarityCaps with the recent sample order. The sample order so long in the making has finally been confirmed. 

roxy54

Since the mid 90s, I have used a pair of Classe DR9s in mono for my professional studio and consulting needs. They were more than adequate, especially via RCA inputs rather than the OpAmp quasi balanced mode. After I bought Thiel' Audio's ashes in 2018, I migrated to a Benchmark stack (DAC3B, HPA4 headphone/preamp, and a pair of AHB2 power amps.) They meet my present playback and research needs very well.

Along this recent path I wanted ways to power Thiel's difficult loads at prices musicians could afford. To that end I've tested a lot of gear that comes across my radar - much of it via this forum. Adcom was such an exploration. Here's a snatch of what I learned. Some of the Adcom gear was designed by Nelson Pass, whom I adore. Much other of their gear is more ordinary. In particular the GFA555 (straight) was Nelson's shot at simple, clean, innovative design that checked his boxes of voltage/current stability, short signal path, stability into low impedance, etc.

I bought a low serial number GFA555 and later a higher number 555 and then (on recommendation) a 555 MkII. My evaluation was the low# 555 was clean and direct, the higher# 555 was somewhat shouty and jittery and the MkII was smoother, less engaging, more ordinary.

Jim Williams of Audio Upgrades in Carlsbad rebuilds/ upgrades studio recording consuls. He hotrods the GFA555 (not MkII) and informed me that Nelson's 555 depends on high quality parts to shine. The early 555 renditions featured best of form Japanese components assembled in Japan. Later they went to Taiwanese parts and board assembly. The MkII was a French outsourced design that shared only the chassis from the original 555 plus the name.

I sent one of my 555s to Jim. The upgrade was wonderful. Increased bandwidth on both ends, better transient speed and settling time, noise floor, etc. all substantially improved. For a few hundred bucks. That amp went to Bill Thalmann of Music Technology for audiophile tweak recommendations. Then back to Jim for implementation including best of form film resistors, etc. Point of exercise is to create a no-excuses solution with under $1K investment for folks in my world.

The Adcoms don't perform at the Benchmark level. And the Benchmark might be  outclassed by high end gear, usually at prices beyond the reach of most.

I'm thrilled with the Benchmark's performance and price; plus the Adcom GFA555 (modified) and Classé (modified) ground me to my personal history with performance that I judge as really fine, especially considering price.

thoft - let's consider possibilities. Your 18'x 25' room with a non-flat ceiling sounds pretty ideal to me for a pair of 3.6s.  The Thiel room where duramax heard the 7.2s probably had the Krell FPB 600. That room is 14'x 22'x 35' for 10,780 cubic feet. Your 18'x 25' room has significantly less volume, depending on the particulars of the vaulted ceiling. As stated, a major design attribute of bigger speakers is to provide more bass into larger rooms. However, the 3.6 bass fills the Thiel listening room very well. I would never call it bass-shy, much less 'extremely weak'. Your room is significantly smaller.

I imagine you explored speaker and listener placement to minimize room modes.

I don't know whether your Classe 25 amp delivers current into a 2ohm load. I read reports of it delivering to 2 ohms, but its specifications rate to 4 ohms (minimum). Your amp running out of current at Thiel's 3 ohm load is under suspicion.

The 7.2 and 3.6 are in the same league for impedance brutality, rarely climbing to 4 ohms.

Another thought is something being wrong with the speakers. Your symptoms could be caused by one woofer being disabled in some way. Do you have any test equipment to compare the 2 speakers? Tone generator, REQ, etc.  If not a momentary connection to a 6 volt (9 volt with greater care) battery can show the polarity and vigor of the woofer response. Listen as well as watch what happens. 

Troubleshooting demands that we reduce variables. Battery test is cheap and effective and simplifies the decision tree.

vm1 - Rob is the expert, but he sometimes goes dark due to personal circumstances. So, let's go. The CS2.3 came out in 1998 and the coax was revised early 2001. The revision was adding a bucking magnet to the coax to eliminate video interference. That also quiets stray EMF field interactions with the crossover network. All good. Possibility of other changes unknown to me - I wasn't there.

I have a factory XO drawing dated as 11-17-08 and 2-2-01 which times up nicely. Rob has said the coax change was made at #4567 which is pretty late in the product life. I can share this schematic with you, and would appreciate a photo of your XO and serial number. If interested, send me a PM.

vm1 - Thank you for your kind words. We did indeed hope to create speakers to bring musical enjoyment for a long, long time.

Regarding the CS6, there is lots of good information in this thread, some quite recent. In broad terms, Thiel was a cost-plus operation, what you paid reflected what went into the product with old-fashioned (tight) margins. If the CS6 sold for double the CS2.3, it had twice the manufacturing costs (simplified model.) The model includes having twice the budget for amplification to support the larger speakers. And a larger room . . .  Supporting a CS6 purchase is much costlier than the 2.3.

Here’s a little catch-up report. During the summer I got snatches of time to evaluate outboard crossovers and internal wiring. I have added some stunning KimberKable solid silver 3-braid to the mix. Some may remember awhile back when some of us compared single-polarity runs vs +/- paired-polarity runs. We were comparing amp to speaker runs and the main problem with singles was variable wire to wire distance causing transmission anomalies. But we were all enamored by the quick, uncomplicated sound of the parallel configuration.

I spoke with Ray Kimber, who wholeheartedly recommended the single polarity solution, which is now included in my trials. In the cabinet, each run can be stuck to the wall with butyl rubber goo. What a sonic treat - from the point of view of ease and lack of dynamic congestion. Kimber’s samples of AGSS 3-Braid (stranded) and AGSC(MST) 3-Braid (solid) employ 3 conductors in a braid for each separated polarity run. MST stands for Magnetic Stealth Trio. Insulation is teflon. I’ll be getting copper samples soon.

The CS1.5 is my current workhorse, having 4 relatively short runs from the outboard XO to the drivers.

In the realm of disc-spinners, I’ve broken into the twenty-first century. I picked up a Sony S9000ES player. Alas, it came with the wrong remote and I haven’t yet played SACDs with it, but I’m searching for a RMT-D122 required for this player.

I have finally fired up the SCD-1 bohemoth that I snagged from the Thiel Audio auction. Its SACD section is kaput via Bill Thalmann’s diagnosis of a motherboard issue. But the CD section is quite a treat. Wow. I have retired my erstwhile Philips CD80 which has served me well for 32 years. Time comes for everything.

Anyone here know anything about the SCD-1’s selectable filters? The (70 page!) manual gives marketing speak descriptions of the 4 alternative ’slow’ filters. I’d love to learn the actual performance parameters of each.

The operational and service manuals have lots of information, including a caricature description of the filters. However, I located Sony's white paper on the SCD-1 which has more detail including frequency responses of the 5 filters. However, I want to look at how the reconstructions affect the time domain. I'm developing a mental framework, but that task would be more effective and productive with more information.

Do you know where that filter discussion / information might be found?

The control match is confusing. The match for the S9000ES is a RMT-D122. Ebay, etc. says the D111 will work, but it doesn't work with an SACD player. Bill Thalmann is sending me his spare. Now, to find the remote for the SCD-1. Always something.

Great review - great reviewer.

I get the audio preference bit . I got a contact at Sony technical service whom I’ll query for specs. 
Thanks.

Robert - As I recall your parallel internal wires were MonsterCable or equivalent. Twisted pairs (as in stock Thiel) address the relevant problems, plus they are what Jim designed with, considering their inherent capacitance, inductance and resistance. If I land on Kimber MST or the Iconoclast equivalent, there may be crossover adjustments for the differences in C/L/R from stock wire. These braided runs resolve propagation effects differently than twisted pair; fans believe they do so more elegantly. It seems these effects are too subtle and/or complex to be modeled in obvious ways. I had previously dismissed them due to my lack of understanding of the interacting variables. Ray has encouraged me to include them in my trials.

Previously I had auditioned the Iconoclast braided conductor speaker wire. It is excellent albeit very expensive. I have approached them about virtues and problems with their wire as single runs rather than the +/- runs that I auditioned. And the beat goes on.  

In my considerable messing around with posts, augmented by some industry opinions that I trust - -  - I suspect that geometry may be more important than materials. The original Thiel posts with plastic caps were compared to many 'superior' contenders and always came out on top or on top considering big price jumps for 'better'. I suspect Thiel succumbed to market pressure to have something 'better'. The later Thiel posts are gold on brass (I think), plus pretty beefy. My experiments lead me to believe that bigger is not better, and can indeed be worse due to eddy current generation due to abrupt changes in conductive mass in the presence of rapidly changing signal level and direction. I'm not a fan of beefy connectors regardless of their materials.

WBT makes a low mass, small contact x high pressure post that I am evaluating. They developed a way to deposit 3 molecule thick gold onto good copper. There are also tweaky posts which take that idea further via non-conductive carbon dshells with only a single conducting wire on the tangent. Danny Ritchie of GR research produces a thin shell tube connector. I don't know anything about Mundorf's solution.

I have no final determination, rather a developing opinion that low conductive mass is better.

I am looking forward to your Mundorf vs stock report.

Nice - looks like what I was describing, except for sticker shock.

Tom

Some remembrances of Amberwood.

From its beginning, Thiel offered four primary veneers: Black Walnut, White Oak, Teak and Brazilian Rosewood (extra charge). Over time we stocked virtually anything that anyone requested. That was possible because we worked from raw veneers rather than laid-up faces or panels. We matched, jointed and laminated in-house which is highly unusual (some say crazy!) As Brazilian Rosewood became more scarce > illegal to import by 1991, and Teak was becoming lower quality and more expensive, I sought a beautiful, sustainable, affordable addition to our stable of offerings. That materialized in 1990’s CS2.2. The wood goes by Santos Rosewood, Bolivian Rosewood, Pau Ferro (ironwood), Caviuna and other names. To verify ethical / sustainable sources, I went to Brazil and Bolivia to sort the liars from the thieves. I landed on a Japanese mill in Santa Cruz, Bolivia as our primary source - impeccable in every way. I toured the mills, log concentration yards and some harvest operations in the company of the head of the Bolivian Forestry Department, the dean of the University School of Forestry, and Jim Martin, our importer and president of ’The Forestry Fund’ which underwrites tree planting, education and sustainability programs around the world. He had the checkbook that underwrote a serious sustainability program for Pau Ferro with the native Chiquitano tribe as our long-term partner. Think curare blow darts protecting the managed forest plantation. That’s another story for another time.

Back to Amberwood. This Pau Ferro timber is what has been called ’Rosewood’ in European (Danish, etc.) furniture since the 1960s. It’s not a true (Dalbergia) Rosewood, but has similar properties. Among its many international names is Santos Rosewood for the southeastern Brazilian port of Santos where it is exported to Europe. But none grows within thousands of miles of Santos and most of what is sliced in Brazil is stolen from its native range in Bolivia. I wanted a clean name for our clean sourcing of this wood with a very dirty history. The plot thickens in that two primary species of Machaerium are combined for market (a common practice). The lowland (Machaerium acutifolium) is fairly dark purple and goes locally by Morado (purple). We adopted that name for that type. The more rare upland (Machaerium scleroxylon) is lighter and more variable in color, often with more exotic dark striping. Lacking a separate commercial name we felt free to assign one. Kathy Gornik engaged in dialogue with some dealers and customers and coined the name ’Amberwood’.

We were able to segregate the two types due to our direct Bolivian source. But things get more complex. The two species hybridize at mid elevations and conditions. And we learned the hard way that some of that middle-type wood was more photosensitive - bleaching to amber, while other trees stayed darker - mellowing to Morado. Side-step to the veneer inspection / buying process where typically three slices are removed from the sliced flitch (at 1/4, 1/2, 3/4 through the stack). These samples became my inventory-planning samples. I stored part of one of those samples from every ’Amberwood’ log in a sunny location to determine how it would age, forming the basis for naming that flitch either Morado or Amberwood.

This long story addresses a small corner of the complexity of offering real wood finishes at the level of attention that we did. I’ve been told that my ’ageing for identity’ process didn’t continue after my departure, which leads back around to why some Morado isn’t very dark and some Amberwood isn’t very amber. It seems that Thiel Morado became a red-stained finish and Amberwood became more varied as luck would have it.

Enjoy your Amberwood / Morado speakers - Tom

I based my dis-recommendation on multiple grounds. Rob knew the work and disrecommended it. I also know that Jim's approach is to balance the effects in the frequency and phase/time domains. We don't know how this XO tweak approaches that work.

It may have merit. If anyone tries it out, a report here would do us all a service.

About the wire thing - I’m living in that rabbit maze and it’s fascinating, complex and difficult. ll post more as I find time. A brief response to the ’solid or litz’ remark is YES. There are so many interacting variables of geometry, conductor purity, and insulation dielectrics that can go wrong in so many ways. George Cardas famously said (something to the effect of) ’solid conductors stay out of trouble easier’.

Fifty years ago (Thiel beginnings!), little credence was given to wire as a performance element. We gained access to (at that time) obscure and esoteric considerations about wire via our cousin Ted Lyon who was a senior physicist on the Jupiter Space Probe project. Ted introduced us to wire considerations and solutions which resulted (among other things) in our adoption of ultra pure, long crystal, polished solid conductors in teflon. That solution persisted through Jim’s career. It can be improved upon, but with peril and expense. I’ve presently been massaging those considerations and variables for a long time.

As time has passed, the term ’litz’ has faded from use due to foggy definitions. Originally it meant each individual conductor separately varnished (insulated), but came to apply to sub-bundles being varnished, which doesn’t address the inter-strand conduction problems. (That phenomenon is itself controversial.) That sense of no strands or conductors touching any others is the sense that Jim meant by 'solid or litz' being best. In today’s jargon, those individually litzed conductors would be called ’solid wire’, even if their gauge were smaller than a human hair.

Besides Thiel’s 18 gauge solid, and many others’ fine-gauge (individually insulated) solids, the foil category fits the isolation requirements. Jim used Goertz foil speaker wire in his later years. The thinness of foil conductors solves the skin effect problems of round wire as well as meets the individually insulated requirements for non-signal migration. But its inherently high capacitance makes it inappropriate for use with some amplifiers, and likely to perform differently among many amplifier choices. So I am avoiding that avenue.

I’ll come back with more comments and reports about my work over these last months. It’s quite a trip.

duramax747 - I wonder if Rob would be interested in partnering with you. Perhaps he could be one of the customers and individuals could group-buy from you, and he could get additional inventory to sell to others. I would be interested to investigate non-metallic solutions that you have mentioned. We might also experiment with isolation inserts to decouple the spikes, etc. from the outrigger. There is experience on this forum to guide potential solutions. Those experiments could be engineered to be reversible for further learning.

unsound - the 1.5 figure is from John Siau, the amp's designer. Their feed-forward topology maintains stability until it doesn't - when it interrupts the circuit. BTW: I'm all ears for your 'better recommenddations'.

The 6' speaker cables came from my feedback to John re the looser bass of the bridged mono configuration. Clearly audible improvement chopping my Straightwire Octave 12' cables in half. I use my 6' cables from my centered amp stack.

I suppose the old amp thing is my personal hobby. There are so many GFA555s out there for a couple hundred bucks and with a few more hundred bucks, they greatly outperform whatever these 'normal', non-audiophile people are using. That solution appeals to me.

I wish two things. Larger BM amps, which John Siau rejects the need for, and higher impedance loads from Thiel speakers, which Jim Thiel rejected the need for. That's a long talk in itself.

Some thoughts regarding speaker height -

Although the distance from the woofer to the floor has some relevance to the frequency content, its contribution is minor compared to the relative distance of each driver to the listener's ear. Those relative distances are critical to preserve the integrated sound composed of individual onset transients from each driver. The distances are established by the relative placement of the drivers along their baffle and the ear's relationship to those launch points. Higher frequency geometry is more critical due to their small wavelengths.

Thiel design assumptions are an ear at 36" from the floor and a distance of 3M (10') with a minimum of 2.5M (100") more or less. Moving the speaker higher or lower is quite similar to moving the ear higher or lower. The relative driver distances are distorted.

A fairly safe way to keep the geometry straight with very little compromise to the integrity of the composite wave is to vary the vertical angle of the speakers. I suggest marking a spot 36" from the speaker bottom, and making a target at your actual ear height. Hold a builders' square at the speaker height mark and tilt the speaker until your sight-line hits the ear target. If you raise the speaker by any means, the compensation is to tilt forward to bring the triangulated geometry back to focus at the ear. (And vice-versa for high seating.)

If you use REQ, FuzzMeasure, etc. the proper tilt, and a mic at ear position will produce the best-looking step response. Other tilts/heights will produce a hole or lump or offset between the early-arrival tweeter onset transient and later-arrival woofer onset transient. Later Thiel products with coax / coincident upper range drivers solve the tweeter x midrange issues, but the high vs bass timing issues still remain. 

Thiel's design target for bass Q is .7 in an anechoic environment. That is on the articulate / lean side of the tracks. Room gain will augment and time-extend the bass to the extent that the room is large enough to support bass frequencies.

Here is a link to a relevant article by Robert Harley published in The Absolute Sound. 

FWIW, if anyone has experience and opinions about the Gaia vs standard Thiel spikes, I'm interested in hearing about it.

The CS1.5 spent its whole life on Stereophile’s Class A (limited bass) list. But its design constraints are real - a seated listener at least 100" away. Also, the only baked-in timbral balance assumption is that it is on a floor. The anechoic (no floor) frequency balance is shelved down 2dB below 200 Hz such that floor placement makes it flat.) Placement relative to walls, including the wall behind it, greatly affects amplitude and quality of bass. Your stage height observation is due to spectral interference from lobing of the individual drivers. Those propagation patterns integrate beyond about 8’ listening distance. I suspect you can get it close to right by careful attention to ear height and cabinet tilt. You gain some positional flexibility by listening up to 25° off axis (which would be straight-ahead at 10'.) If you have placed your speakers along the room’s diagonal axis, then the side wall reflection will be greatly reduced. If not, the Thiel wide-dispersion characteristic works against you in that small room.

duramax747 - congratulations! The room is such a very big deal. Treat your AC mains like signal wires with star quad twisted runs, etc. and high grade receptacles, etc. Run off 240 volts where you can. Very worth the trouble.

No, I haven't heard the 7.2s. No suitable space or amplification yet. My amplification is going the be my good ole Classé DR6 and pair of DR9s. Bill Thalmann / Music Technology did a fabulous job upgrading the 6. The 9 project is waiting for Purity Caps (to ship in early March!) It has good reason to be equally stunning. We're dedicating each amp to mono and taking out the (funky) balancing circuitry to make room for goodies of Bill's dreams. Sometime this calendar year I hope to have the 7.2s set up somewhere that really shows them off. Not yet, maybe soon.

Max - what a fabulous setup!

The DR79s are only 100 watts (stereo) into 8 ohms. We are dedicating them to bridged mono operation. Their stock power is rated at 1100 watts into 2 ohms. Hotrod will increase output. I realize that more is better, but these amps are what I have and know, and Bill T. has encouraged this particular upgrade.

My thoughts on house wire include the effects of Electrical and EMF radiation/corruption on the transmission properties of the power cable. The loads are not steady-state due to less than perfect isolation of the audio devices. Power cable commonly comes in flat and round (with a twist) configurations. I recommend the round variety with hot leads opposite. (Not star quad, but field propagation is better than flat wire when loads fluctuate.) Also consider the radiation characteristics of the outlets themselves. All 'hospital' or 'laboratory' grade devices are not created equal.

Keep us posted as you build out this room.

Regarding amps: In addditon to my hotrodded learning labs of Classé DR9s and Adcom GFA555s, my reference amp is the Benchmark AHB2. A single stereo amp may not have the oomph you need/want, but in bridged mono a pair might. Power aside, the distortion and noise characteristics of this amp is extraordinary and its unique feed-forward topology pegs the output signal profile as nearly identical whether stereo or mono. The only difference is that the damping factor goes to half in mono, but chopping your speaker cables to half length corrects that. They maintain their operating characteristics to below 1.5 ohms.

It took me awhile to get comfortable with their operation and differences from 'normal' amps. I'm hooked. You can read all about it at

Lloyd - what is the height dimension of your room?

jon - it could be anything in the entire chain from source to speakers. Time for classic, systematic troubleshooting.

The likelihood of XO problems is very small, and XO problems tend to exhibit frequency-related / balance problems. A rubbing voice coil is possible. Feel free to PM me for troubleshooting advice.

lloydviii - thanks for the room clarification. Whereas 12’x12’ is really bad, your room lies on the edge of the ’Bolt Area’ of ideal proportions. Your additional adjoining spaces almost always make things even better. Exposed glass is problematic. The record-making process assumes symmetry in the playback field. Work toward that however you can.

loydvii - not long ago in this  forum there was a list of likely cable candidates for Thiel speakers. I would like to add Straightwire to that list. Specifically their Octave II (now III) series incorporates all of Straightwire's core technologies and is quite affordably priced.

Lloyd - I've had good results from pleated-cell blinds - the heavier audio-grade variety. Draperies work best when pleated to provide air spaces and an undulating surface. Beware that too much absorption can kill the dynamics. Start with a little and add if you need more. I would start with the blinds.

Roxy54 - it's hard to know. I had some clues from an acquaintance who had been personal secretary to Adcom's president in the day. She advised  me to find an 'original' and helped puzzle it out.

It's not uncommon for a manufacturer to create a worthy product, get good reviews and dealer buy-in and then erode that foundation via lower cost components, sources and so forth. I was personally chagrined to see some late Thiel CS2.4s that didn't hold a candle to the original Lexington variety. So the world goes.

I suggest keeping an open mind to the potential of the GFA555. Nelson still loves it as designed; and we have certainly taken it up a few notches.

masi61 - Greetings. There’s been no response most likely because there’s no safe answer. Everything depends on everything, and your room dimensions, absorption, damping and losses interact with the floor coupling. I, too, await with interest any observations from those experimenting with GAIA feet.

I can tell you how we approached the matter at Thiel (early days / first 20 years), which may have changed, but not too likely to diverge radically from the following.

Speaker development occurred both in the lab and under multiple test conditions. In other words, we did not tweak for or against any particular coupling conditions, since those are so complex (as stated above) and unknown, to be counter-productive to project or assume. Bass target was .707. Baffle step (progression from omni-directional to directional propagation behavior) was first calculated and then verified multiple ways: speaker hanging in free air, speaker on pins on a solid, non resonant surface, on a solid elevated platform, buried in a sandbox, and in at least two listening situations - all under measurement and critical listening.

Over time we learned to correlate these conditions to shape toward our target .707 x -2dB bass shelf in anechoic free-space. Our surfaces were these: Outdoors was A) flat roof with rubber membrane roof over hard foam insulation on deck = highly damped and non-resonant. B) parking lot with asphalt over packed clay (heavy truck surface) which is also non-resonant. Mic positions were overhead, ear-height (floor bounce) or ground-plane (no floor bounce). Note none of these has any reflective / reactive coupling component. The sandbox puts the speaker on its back, firing up with baffle edges flush with the ground plane (no diffraction). Alternately the speaker is in the lab wall in a quasi infinite baffle. Comparison of sandbox to wall-mount lets us see cabinet vibrational effects and leakage losses as well as edge diffraction effects.

Indoors had two major environments: A) the farmhouse had stiff hardwood on joist construction with very little bounce, but significant bleed which differed with and without spikes. (Oak floor on hard southern yellow pine subfloor over SYP joists) 1903 Victorian Farmhouse. We treated spiking as a way to effectively eliminate enclosure recoil / sway, but ignored tonal additions / subtractions as arbitrary. After 1980 the lab was moved from the farmhouse (which had incorporated a cross section of the previous elements, to industrial space in a second floor, wooden-floor space, then two subsequent concrete-floor spaces. The concrete floor was industrial freight warehouse spec. The factory purpose-built music room had that same slab floor topped with glued-down industrial hair underpad and tight wool / jute-backed carpet for a very quiet noise profile. (A 10# iron ball drop did not ring).

We also took every speaker in development back to the original farmhouse living room which produced a warmer, fuller room sound than any of the lab or factory listening environments.

Of course there were show environments, which we took as they came. And we received criticism for rarely tweaking the room or floor coupling for ’best’ performance. We used shows more for dealer and reviewer engagement rather than showing off the highest performance of the product.

All this is by way of describing that there are many, many interacting variables with little to no way of predicting how your variables will stack up against neutrality or your tastes. A particularly tricky business is separating room modes / placement issues from floor coupling / leakage issues. Have you used any of the dimension calculators to assess your room issues? You are welcome to post your dimensions here if you wish and I’ll respond with first-pass performance comments.

I have no experience with any of the commercial isolation / coupling products. I have learned a little from show setups and local installations. With the stock Thiel pins you can effect their floor interface by using blu-tac or mortite in the sockets and/or under the round end at the floor termination. On a carpet you can tune the interface by how hard you force the points into the carpet or whether or not you put anything under the point (rivet, washer, penny, checker, matchbook, etc.) Historical note, the later, wider stock points with 60° included angle do not readily pierce the carpet,but they do concentrate the load well enough to minimize cabinet recoil. Using these simple, free approaches, you can get a pretty wide range of isolation / coupling with various floors.

I realize there’s very little practical guidance in all these words, but practical advice lies beyond my ability to comment. Hearing some GAIA comments might shed some light.

prof - sorry if I missed that.

That could very well be the Canadian price of the CS2.4. In the USA, Thiel acted as its own distributor, managing marketing, promotion, retailers, reviewers, service, etc. We attributed about 25% of our sell price to those functions. In foreign markets (Canada is actually foreign, eh) a third party distributor supplies those functions. Even though we discounted export sales, distributors generally added more cost than we did, partly because their warranty services cost more than our in-house support. And the Canadian dollar sometimes fell considerably below the US dollar which was reflected in variable sell prices at different times. Canada was a good market for Thiel.

kheine - Clearance is tight enough under the cabinet that some banana plugs are too long to fit. There are 45° locking banana plugs that solve that problem. Send me a PM if you wish.

JA - Thanks for asking, and I suspect an update is in order. I would characterize the present state of the HotRod Garage as ’distributed’. I have found suitable workspaces around the village with some being more suitable in warm than cold weather, and for different purposes. These developments represent real progress.

My work is expanding my knowledge / solutions beyond classic Thiel implementations. There were many areas of subtlety that were simply beyond consideration in a company whose vision was to bring outstanding performance at the most affordable price. I am developing and evaluating advances that would not have fit that classic Thiel paradigm. But cost-effectiveness is still of central concern. One could spend as much on a few caps as we spent on an entire speaker system.

As you know, I began experimentation on my CS2.2s because I had them. I have expanded that stable to 13 models (including subwoofers) with more lined up. As I’ve mentioned, I’ve been focusing on the 02, which has been quite a trip because nearly nothing in that product is identifiably ’Thielesque’. Any change I make, such as stranded to solid wire, makes it sound more identifiably ’Thiel’. The workhorses or any model have served as sandboxes for trial and error. The survivor outcomes of the trials are applicable to all models. And there’s a good, solid list of techniques and technologies.

This winter we’ve entered the realm of actual, presentable product development. The backbone platform is the SCS4 - fifth generation of the 02 portable 6.5" two-way. The SCS4 embodies virtually all signature developments from 1976 up its 2007 release. However, it still contains much room for improvement, some at very little cost. The driver and its control circuitry stay; much else gets attention, some of which might later require slight XO re-tweaking. Interested parties might put a pair of SCS4s on their radar.

As a blast from the past, I am expecting a pair of original 01s - first iteration. As an example of Jim’s instincts it shines a light on his innate values, much like a recording artist’s first album illuminates their inherent nature. Virtually nothing of the 01 survived the arc of his career, but the instincts did motivate all further work. As one example, Jim held high sensitivity / efficiency in high regard and employed minimal crossover circuitry in service to that goal. As he sought further refinement his crossovers became famously (infamously ?) complex, and system efficiency suffered. In his later work, as he was able to engineer his proprietary drivers toward greater linearity, as well as break ground with lighter moving mass and stronger magnetics, the speakers became considerably more efficient while achieving outstanding linearity. Anyhow, I will relish examining, measuring and listening to a pair of the first generation 01s, designed in the farmhouse with slide-rule and rented oscilloscope, and built in the garage with finish sprayed on the back porch. The product was unique enough to garner (first year CES) distribution in Germany, then other European success, which raised some eyebrows in the USA.

I would like to thank you and everyone on this forum and especially those of you who have joined as collaborators in these explorations. A final note is that after a year (more or less, I’ve lost track) the ClarityCaps along with the new Purity samples are scheduled to ship March 2 from Wales. Patience is a virtue. Satisfaction is more fun.

And the MDF baffle, no matter how thick, is worth upgrading. I've settled on Torrified (heat-hardened) bamboo floor stock as an excellent, cost-effective brace material. Retrofitting isn't easy, but it's worth a good look. And the list goes on. I'm addressing older speakers first due to my learning curve as well as the need to re-specify the non-replacable drivers. The beat goes on.

JA - I really appreciate all the backgrounding you have done to point me to these Thiel speakers. Thank you.

Prof - we have demonstrated various improvement avenues for any / all Thiel models. Beetlemania posted his 2.4 results on this forum. There have been others. The upgrades all leave the basic stock platform unaltered, but upgrade various aspects that can benefit from improvement. I haven't done anything with the 2.7, since they are late-model and excellent. But, we know that we can improve them with our bag of tricks, if and when the time comes.

Regarding your stock 02s. I can recommend here a tweak that does no harm. Get inside and rig the fiberglass insulation away from the crossover network for better thermal management. Get some butyl rubber sticky-tape and route the driver wires away from each other (woofer / tweeter) and tape it to anything solid that you can, or just use the tape itself to keep it still. And, solidify the coils however you can. I add ordinary shellac or varnish and zip-tie them as tightly as possible.

Those coils were hand-wound, and their looseness robs transients.

These mild tweaks will tighten up the sound without changing the sonic character. Or, find another pair if you want to do more. First pass would be to move the crossover away from right behind and in line with the woofer EMF field. And then there's cabinet wall bracing and passive parts quality, etc. etc, etc.

Hello All - regarding MQA. I’ve been paying attention to MQA and its claims since the beginning, and I have been somewhat bewildered by the fairly intense controversy surrounding it since the beginning. I know some mastering engineers who swear by it. I also know some who reject it and some very astute listeners, some with significant technical chops who consider it (various degrees of) less than helpful. This past weekend I had a personal experience which I want to report.

I recently procured a pair of SCS4s which I’m documenting for actual upgrades (for the real world.) I took them to an audiophile friend with a good system and amazingly analytical ears. Our source was various reference pop titles via Tidal. His turf, his call. I was underwhelmed and disappointed by the SCS4s in his setup. On working through the maze of possibilities Saturday night it occurred to me to explore MQA, since John had mentioned it as being offered by Tidal. I read Peter Moncrieff’s MQA section from his ’Digital Done Wrong’ series - coming away with a cogent set of strong arguments against MQA.

Let’s digress to Peter. For those of you unfamiliar with him, Peter has published the International Audio Review since the early 1980s. He is brilliant, self-assured and not deterred by the norms of polite discourse. I have always found him to be right. Peter is who assigned the title CS to our 03b at the 1983 CES. He got it and reported it very succinctly.

Back to MQA. Peter itemizes how the ultra short sampling schema of MQA misses the ’unlucky samples’ that fail to describe sharp transients. What I heard at John’s house was music without upper transients as well as a global haze or ambient cloud, noticeable mostly in the upper registers. I messaged John and he reported that his Tidal MQA filter was indeed engaged. We have not yet had a chance for direct comparisons with and without MQA. I should add that a pair of Klipsch RP600Ms did not exhibit ’the pleasant vagueness’ that the SCS4s did. I felt this note was in order to ask your experience and opinions about the subject. I’ve studied enough about the process and heard enough examples to at least question the process.

JA - John’s system is decidedly low-key, and interesting for that. He had sold his system before early retirement to Sandwich. His primary source is streamed music via Tidal, generally high resolution with MQA option. He has a turntable, but I haven’t seen him use it. No disc player at all. The server is (I’m blanking the brand) mid-priced, good performance. His previous server was PC NAS based which he had hot-rodded for verified bit-perfect delivery to the USB output. (John’s background includes MIT electrical engineering and computer science. He knows stuff. The new dedicated server outperforms his PC rig. Amplification is a Prima Luna integrated at about 35 watts, in which he has rolled tubes to firm up the bass, de-glare the mid-highs, etc. Speakers have included Thiel CS1.5, 1.6 and now 2.4s that he bought from audiojan (through this forum). Cables are Belden via BlueJeans, although we have run some of my experiments through his system. Presently we’re looking for a time to compare my StraightWire Rhapsody III ICs and Octave II SCs, as well as a borrowed pair of Iconoclast II speaker cables to his stock stuff. He has some sort of good AC filtration.

His room is well-enough proportioned, but has glass walls behind the speakers and the listeners. Ceiling is high enough, but reinforced with flat trusses. We have swept it and it’s pretty clean, but could definitely use more work. He is a very busy man, and we find time for audio fun less than once a month.

He was a founding operative of JazzBoston (which didn’t make it). He knows musical performance and production and has an amazing ear which I have come to trust more than my own. He is younger with hearing all the way up.

I would say John’s setup is good enough to show intrinsic behavior of whatever we’ve put in (such as Benchmark stack vs his PL). I know this MQA bit-truncation (or whatever we’ll call it) is real, but not yet analyzed enough to make much sense. What we both heard points to the kinds of misgivings I’ve been stirring around.

Robert - I’m sorry that at the present time my comments are more theoretical than practical. My experiments based on fundamentals is that I dislike any iron / steel and even anything electrically conductive in the high-intensity inductive field directly in front of the drivers. I’m working toward a stiff fiber mesh to give physical and visual protection while eliminating electrical / emf coupling. Even though these late model grilles are very acoustically transparent, I suggest removing them for critical listening. I don’t know whether you’ll hear a difference, but I would love to hear anyone’s report on the subject.

Yes, grilles are a perpetual problem. But fragile drivers are a perpetual risk. I'm working on a more sheer grille that is not metal or foam.

I haven't gotten re-conformation, but the CCs were scheduled to ship Feb 2, now March 2.

Grilles - The issue is multi-faceted and full of trade-offs. Early 3s had wooden 3-D frames stretched with fabric. Fabric absorbs some high frequencies; Jim voiced assuming fabric. 1983’s CS3 used the baffle-level of the frame to complete the baffle's rounded edge profile, and the aerial parts to support fabric away from the baffle. 1989’s CS5 had a 3-D steel frame with aerial fabric (away from the baffle.) The low (1/4") profile of the frame reduced its diffractive contribution to low enough to ignore. But we knew it was there. 1990 CS2.2 and 1992 CS3.6 used the baffle perimeter edge profile of the CS3-3.5, but eliminated the aerial structure. The all-MDF 3.6 frame proved too fragile and a side-bar reinforcement was added. We considered aluminum and steel and steel won for cost. We knew side-effects (more on that later.) 1995’s CS7 (my last industrial engineering contribution) again addressed the same issues and again landed on steel. Note, that baffle was cast concrete which required re-bar in the narrow areas beside the woofers. Those eddy currents could be heard and inferred from measurements. Note that 1999’s CS7.2 upgrade changed the frame to aluminum and eliminated the rebar via fiber reinforced mineral casting medium replacing concrete.

Directly to your question: 2002’s CS1.6 seems to be the first product with the thin steel plate supporting stretched fabric. The shallow baffle indentation was probably deemed not too problematic. An aluminum plate could not have been fastened with magnets. (BTW magnets near a driver are detrimental.) This ’fabric on plate with magnets’ solution persisted. The SCS4 has a formed steel cage in a shallow perimeter groove, adhered with magnets. Nice solution, with downsides. I’m replacing every element of it.

Regarding ’more on that later’. Jim was a disciplined pragmatist. Any and all design elements had to carry their cost burden. Incorporating ’better ways’ could increase costs which were doubled by the retailers’ 50 point margin. We were constantly juggling those equations. Although not sexy or particularly marketable, value engineering was a major aspect of Thiel’s approach, both in product outcome and in manufacturing technology.

So, concerning audibility, Thiel (which included Kathy as marketing director) played tug of war quite a bit. How much can be heard by whom (% of target market), etc. Every brand has a niche. As an aside, in the early 90s it was clear that something had to be done about home theater to stay viable. Thiel dove in and the rest is history. But there were other options considered. Among them was my idea of creating a brand offshoot (much like Lexus is to Toyota.) The name was to be ’Perigee - the closest approach’. It would follow a different paradigm that allowed and developed its own layers of sophistication to stretch the cost/performance paradigm. Jim’s principle objection was that it would cheapen the image of standard Thiel branded products. Fair enough. That route would have concentrated our resources in high resolution stereo playback. And it would have required product development collaboration beyond Jim's comfort zone and vision. 

More complete solutions for the subtleties of ultra-fidelity music reproduction exist and can be developed. Many were beyond our knowledge or scope, and remain so today. Some have come into focus or feasibility as time goes by. New materials science is a robust field. And there are old ideas that fell away and may be re-imagined today.

Rob - the magnets are installed into blank-bottom holes drilled from the back of the baffle. I don't know how many are in the CS2.4. In the SCS4 there are 8. Two pairs about an inch from the top and bottom corners plus another 4 flanking the coax driver. The fields are strong enough to need a tool to remove the perforated steel grille. I'm not a fan due primarily to theoretical considerations of interacting magnetic fields with an electrically conductive aluminum baffle between them. My objection notwithstanding, I'm confident that Jim vetted the implementation and did not object to it. The magnets are probably far enough away to not do any real damage.

However I would rather have a mechanical solution and lose the hypothetical downgrade mechanism.

Interesting woofer indeed. I wonder if that might have been a development prototype. Its acoustic center would be farther back than the star-plane unit. Smileypete.com is a local Lexington business rag and that article was published before Kathy sold the company. So it's not necessarily from out in the wild.

The mechanical issues are real and worth improving upon.

However, I believe the far larger issue is the interaction of the crossover components with the fields emanating from the driver(s), especially the woofer. These interactions and their sonic effects are not subtle. I had previously noted here that the CS2.2 qualities of 'hoodedness' and bass-transient overload (which sounded like a hard, mechanical splat) went away when taking the XO out of the cabinet and spreading out its components.

The present experiments on the SCS4 move the XO network from a densely packed panel on the back, behind the coax (in a high EMF bath) to either A: outboard with same layout, for testing, or B: split into separate woofer and tweeter boards, mounted on edge with shock-mounts on the cabinet bottom. EMF is more than an order of magnitude lower in the new location, plus mechanical vibration is reduced to near nil - and throw in qualitatively better thermal management for kix.