Thiel Owners

Tom - Thiel had gone to non-magnetic (stainless steel) screws by 2007 SCS4. I don't know when / which other products migrated to non-magnetic screws. I had verified (and taken to heart) your observation regarding non-ferrous connectors.

Today's biggest (repeat) lesson is that XO networks don't belong behind a driver, whether shielded or not.

I've been looking at / listening to grille effects on the SCS4. It is somewhat eccentric in that it has no fabric and the full-baffle is covered by a perforated steel mesh that is quite transparent. The electrical / magnetic fields meter shows that the intense fore-aft fields through the center of the coax are modulated by the metal grille. The EMF changes shape (becomes wider), but most notably, the electrical field is reduced by an order of magnitude in front of the grille. Back fields are unaffected. With my head a foot in front of the coax, I can hear a slight lifting of a veil by removing the grille. I would call this change so slight as to be barely noticeable. Other products may differ. My high-frequency hearing is far from what it was. So others may hear more difference. With all this pesky learning I would remove the grilles while listening.

Over the years that seems to be common among reviewers and listeners. I always attributed that to the fabric, but it seems there are other elements involved.

I've been informed that the 'teardrop cone' woofer was indeed a development prototype. It seems that a calendar year elapsed between that prototype shown at the January 2006 CES press conference, and the actual shipment of the finished 'star-plane' product in early 2007. It seems reasonable that much of that time was consumed by iterating the flat star-plane version of the concept along with its necessary crossover implementation, which would include midrange and possibly tweeter XO changes because the woofer acoustic hole went away. I bet the final XO was simpler.

I presume that the acoustic launch plane of the star plane would be farther forward than the teardrop cone prototype. Time alignment would require a different baffle angle - but the cabinet seems unchanged. There is no analog time delay in the XO (as there was in the CS5.) Much remains mystery to me.  

I don’t think there are any teardrop cone speakers out in the wild. The pair that I saw at Thiel / Lexington in 2012 was #1 & 2 which went to shows, lived in the showroom, etc. It had two identical-looking star-plane diaphragms.

Yes, that's what I am told by David Cubine who was our long-term in-house publicity and customer service man. Dave says that photo was taken at that press conference. Note the passive radiator has the later 'star plane' diaphragm (names are mine.) Seems that Jim found a way to marry the voice coil to the star plane and eliminate the undesirable pocket made by the cone.

Hello folks - here’s a side-trip down memory lane with some destinations in mind.

I recently got a pair of original model 01s, with two more pair of later versions coming this summer. What a trip! The only reference I’ve come across here on Audiogon is someone asking about the Conceptions Electronics source - never much if anything about their bones or performance. But nevertheless, the first work of a designer reveals his soul much like the first work of any artist. There is a lot to love in the 01, even though not much of anything about it survived the ordeals of time. The 01 preceded much of audiophiledom or our awareness of various subtleties, and certainly of time / phase alignment. Mid 1970s didn’t know that stuff. Jim’s native impulse was to design a portable, flexible, honest speaker for myriad uses and users. That was realized in a high efficiency, highly dynamic, accurate transducer with extraordinary bass and linearity. Imagine a bookshelf-sized (02) cabinet that is flat ±2db, 30Hz to 17kHz at 95dB/w-m with linear 6 ohm impedance. Its use of a purpose-built equalizer raised market red flags, but also differentiated its performance and our fledgling company from the masses. (There were masses of new companies entering the fray in the day.)

Despite the shortcomings, I am thrilled with their performance. And they provide a foundation for some further explorations - specifically their two drivers which are no longer available. The 10" Eminence woofer was used in the 01, 01a, 01b, 03, 03a (and nearly phantom 03b). The same requirements were filled by Vifa when our needs exceeded Eminence’s directions (PA and Musical Instruments.) The CS3 and 3.5 share the same needs for sealed bass with linear extension to 40, 30 or 20Hz depending on model & EQ setting. Combined there were 11,000 pair produced using the ’same’ woofer, and a major cause of obsolescence is driver unavailability.

Note also that the never-produced CS4 (4-way) was upstaged by the CS5 (5-way) via market demand, not internal preference. That CS4 would have two of the same 10" woofers, a 6.5" sealed lower mid (think CS7), a ?new? upper mid, plus what became the UltraTweeter used in the CS5, 3.6 and 2.2. The CS4 would have used a bass equalizer. There is a lot to love about having such a woofer and tweeter for service and upgrade for all those models.

Back to the 01. In my small venue - live music and recording work, I use various Thiel models as monitors or sound reinforcement. Most recently the SCS4 is drafted, but, I am enamored with the 01 for that duty, especially as souped up with modern implementations. The 01 provides real 30Hz bass, 4+ x the efficiency, less demanding impedance, and a neat, retro mystery vibe. BTW, I’m recessing the tweeter into a wave-guide to get what has become known as phase and time coherence, where the rebound of the tweeter meshes smoothly with the woofer onset, one tweeter-cycle late. Thiel’s ’phase / time coherence’ now seems to be called ’phase / time alignment, or coincidence’. Fair enough, language morphs. And springtime is just around the corner which brings access to my fair-weather studio.

JA - the SCD-1 makes good music, especially feeding the Benchmark DAC3. I just received the OPPO 105 from Bill Thalmann. Long-term, iterated design upgrade. Very mature. Interestingly to me, his development references included no digital counterparts, but rather his Oracle Mark IV into his own phone stage and best CJ souped with teflon caps, etc. All new analog circuitry. Sound is clean and neutral with no discernable digital glare and solid bass. Soon I hope to make direct comparison with the ES9000 and SCD-1.

lloydviii - I have not yet used the OPPO as a transport, but I plan to run a comparison when I collect the components in a worthy playback system, later in the spring.

So far, so good in the quality of playback from the Oppo’s analog outputs. Best I’ve had; an obvious upgrade from the Sony S9000ES, which is my only prior reference.

My intent was to have Bill Thalmann of Music Technology to provide this signature piece. He believes in it and I believe in him. I recognize that perspective as out of the ordinary, but it’s how I roll. This deck is his final iteration and I am honored to have it.

He and I chose this platform among others. The 105 has more connectivity and a few internals that Bill preferred to the 205. And the Darbee stuff was expendable because his work removed all the existing analog circuitry. This particular unit came from The Music Room which re-sells the units that PS Audio takes on upgrade trade for their own units. BTW: this Oppo deck was used in PS’s decks until they stopped production.

Following is an excerpt from Bill’s correspondence describing what he has done to the unit. He has applied his life skills including his prior career and close involvement with Conrad Johnson toward making his version of digital done right.

I will report when I’ve done some comparisons between his DAC and the Benchmark DAC3.

Bill Thalmann says:

The Oppo 105 uses an ESS SABRE DAC chip, which was SOTA at the time. ESS has come out with a few newer DACs since then, but none are really different in performance. The sound of a disc player still depends primarily on the design of the analog output circuitry and the type and amount of digital jitter. My upgrade squarely addresses both issues.

I’ve never compared it directly to a Benchmark DAC. My primary reference when I did this was my Oracle Mk VI turntable with an SME Series 300 arm and a Soundsmith Aida cartridge through a CJ Premier 15 (all teflon upgraded) phono preamp. While I still think the all analog setup had a slight edge with regard to sound quality, the difference was, IMHO, very, very slight. I am very interested in getting feedback from you even though, at this point, it’s somewhat academic to me as I have no plans to build any more of these.

I don’t have specs on the upgrade beyond frequency response (flat to 50k – intentionally rolled off above that) and subjective testing for the most part with regards to noise and rise time.

and more:

Hope you get lots of good listening from the Oppo.  Very interested in hearing what you think of it once it’s had a chance to burn in. 

Basically, I built two discrete power supplies (each with their own power transformer). One provides regulated 35V to run the discrete JFET analog amp/filter and the other an EXTREMELY low noise 3.3v power supply to power an 80 MHz very low jitter clock module for the DAC section. The clock frequency is higher than Oppo used, but is optimized for audio.  Oppo evidently optimized for video performance.

The 3.3V supply is as clean as a battery, and may be cleaner, and its output filter is a 1 Farad cap. It is highly regulated using IC’s from Maxxim designed specifically for clock supplies.

I take the signal directly from the DAC chip and feed it into my amp/filter board where the DAC output current is converted to voltage, amplified and filtered (at around 50 MHz) and fed directly to the 2-channel output jacks of the player

That’s it in a nutshell.

Bill

The 9000 is a keeper for myself or a friend. I have compared the internal DACs for CD conversion of the Benchmark DAC3, Sony SCD-1 and Sony 9000. The Benchmark betters the Sonys, but the comparison is not embarrassing. In all cases the better player has deeper and firmer bass and is quieter / less jittery overall.

Thanks for these equipment leads and for pointing me to Bill T. I had lost track of him after CJ where he and I had the 'same' job - manufacturing system developer, design for manufacturability, and all those 'do it' aspects of product developmemt - each of us for our company's first two decades.

Lloydviii - consider that 105s are younger than 103s. These things do wear out. And compare the connectivity options. The 105 may be the swiss army knife of players. You won't need the Darbee upgrade for use as a transport and plain non-D 105s are a bargain. The Music Room is where I would start asking and looking. TT

anzen - welcome. There is quite a bit to unpack. I’ll make a brief overview here as a bit of an update of work in process. All of our work is well beyond the SE version. The SE substituted only 2 feed caps, changing nothing else. That ClarityCapSA was their state of the art about 20 years ago. Major improvements have ensued. Extensive comparisons have landed on CC’s CSA with copper end caps - Purity is even better, plus a super bypass can be paralleled for greater definition. RTX or Duelund .01 to .1uF where it matters.

Your coils look good. Keep them. Foil coils would change other parameters and would require redesign. Resistors are worth replacing. Mills in series feeds, others can be Dayton precision 10 watt or their 20 watt (hollow tube) vairety if values available. (from Parts Express).

Regarding boards, it’s not so simple. Your coil values are tweaked for PCB mounting. If you revert to P-to-Point, slightly different coil values are required. Also we’ve learned that one-sided wiring of masonite P-t-P is sub-optimal. So, an entirely new layout scheme is required with edge mounting rather than stock flat mounting of the boards.

I am pleased to be working with duramax747 on CS2.4s as a live developmet project. But whatever comes of that, your largest consideration is your budget. You could responsibly spend $thousands or $ hundreds depending on the scope of your choices.

Send me a PM if you wish.

And on a bright note to all: remember those custom Clarity and Purity caps from last year? They’re scheduled for delivery this week after an odyssey of delays.

I've got some GR Electra Tube Connectors coming. Seems like the best I've seen.

Here's what I've pieced together about the 2.4SE. We didn't know that Jim was dying, but he did, and wanted to memorialize this model as close to his heart. It was originally intended as all-cosmetic. He loved that bright red birdseye maple, and added the outriggers, some other touches and his signature. The crossover upgrade was possibly an after-thought, or nearly so. The whole gang used that idea as an opportunity to compare and contrast XO components and they settled on the ClarityCap SAs (CC's then state of the art.) I suspect those gold bolts are non-magnetic. Anybody know?

I believe the signature run was 150 pair. Thiel's MO would be to have made all those cabinets as a batch, but the crossovers would be made as needed. They were assigned their own set of serial numbers which would go from 1 to 300. I am speculating these details, but haven't found direct competing evidence.

It seems from our direct experience, that the crossovers migrated from early numbers being point to point on masonite and later units being various generations of Chinese components on printed circuit boards. Varying levels of quality. Also, Rob has supplied SE upgrade kits for non-SE cabinets which would have retained their original serial numbers. I have seen an early pair of SEs with classic Thiel parts made very well, and later SEs with shoddy coils and parts quality I found to sound harsher. So, they're not all created equal.

I'll add that my further work with some of you has gone further. That work is not yet ready for prime time, but progress is being made.

It's been a long time since Beetlemania reported in these pages his complete, systematic upgrade where every XO component was eventually changed including hookup wire. JA might remember the pages of his reports.

Duramax - indeed the connector is paramount to performance. I've taken a very deep dive into signal propagation, and know there is much confusion and lack of insight in the arena. The wire guys have paid good attention to metal purity and dielectric characteristics, but still haven't settled on a 'best' geometry. I continue to investigate that aspect (with some tangible success). What seems to be nearly ignored is the termination which plays a huge part in proper propagation through the signal chain.

Consider the signal path as like a water-slide. As an example, consider the end of the slide to meet a spade connector. Smooth flow turns to whitewater. Independent of metal purity, gauge/ampacity/ etc.the turbulence has significant negative effects.  I am not overstating, and you are correct that listeners would be astonished.

For practical considerations I want to disconnect the amp and the speaker. Presuming an outboard crossover, we can attach the input and output pigtails to the crossover and have good user interface. At the amp end I strongly recommend locking banana plugs due to their inline arrangement of field propagation. Of course your idea of direct solder connections could be better. But I posit that the geometry of that soldered connection is critical. Axial, left-lay twist is the correct way.

At the outboard XO we can manage inputs from one or multiple amps with inline soldered connections. Same on the output. Note that I had some custom bolts made in 6-9s copper which I have now abandoned due to the propagation disturbances we are discussing here. I replaced those with thin-wall copper tubing with wires fed into each end, touching in the middle and soldered at the entries. That out-performs any post I have tried. GR's Electra takes that idea up a few notches with purer copper, thinner walls, and the convenience of unplugging.

Note that solder per se is not an ultimate solution. Heat disrupts the molecular structure and metals dis-similarity is less than perfect. In summary, I consider the roughly $12 / connection to be funds well spent. 

Note that the 2.4 is electrically a two-way; the mid to tweeter crossover is visco-elastic / mechanical. So we'll need two pairs per channel which will cost around $50 / channel. That's a lot more than nothing, but less than one typical upgraded capacitor. The proof will be in the listening, but there is a lot to like here.

For your straight-through solder proposal, you might consider the copper tubing idea at the speaker inputs. I plan to re-fit my amps with Electra outputs.

Regarding CS2.4 resistors. As I've mentioned, I have identified thermal dissipation and stability as worthy of attention. Doubling in parallel is a good way to do that.

However that 16 ohm resistor does not need a higher value, being in series with a capacitor that blocks low frequency current flow.

Resistors that are worthy of attention include: Woofer R1 - 4 ohm, and R2 - 2 ohm.

Coax R5 - 30 ohm, and R6- 3 ohm. Those high-current resistors are most worthy of Mills (etc.) upgrade since current flow induces noise and heat induces instability.

audio 1326 - I can comment on some of your possibilities.

I have used the CS2.2s since 1990 in all kinds of applications; I know them better than anything else. They are a target for my upgrade activities, so significantly improved performance will be gotten from them in the future.

At Thiel we had an original Bryston 3B (as well as its successors). I find it somewhat brittle despite its big bottom end. The STs are an improvement, and the 'cubed' series is first-rate.

The Denon is current-limited by definition - it does not double its output into half the load impedance. But as JA mentioned, Japanese made Denons have good bones. And theCS2.2 only drops below 4 ohms for limited time and has a very flat impedance curve, much gentler than later designs. So, it will work.

I like your Adcom GFA555 much better.  The earlier Japanese boards use a higher quality transistor shared with the GFA535. Nelson Pass's designs have simple circuitry dependent on high parts quality. I have two of these amps and the earlier (Japanese) has a distinctly cleaner sound. The GFA555 (straight, series I) can be upgraded by Jim Williams of Audio Upgrades in Carlsbad CA at two levels-the higher using film resistors. The upgraded units punch way beyond their price class. I use a pair in bridged mono configuration as part of my listening and test setup. 

Progress is being made. The ClarityCaps came this week and are being distributed to the various collaborators here.

jafant - Yes, I have been working with Dave since I began this venture in 2018. We have had 3 orders in total to compare critical elements using CSA, CMR (plain and encapsulated) and now Purity - all in the range of voltages. I've also compared other brands and borrowed from Dave some exotic flat stack and series caps. I'm quite confident regarding where the path has led. This present delivery took a year due in part to fallout from Covid, in part from start-up rigors of the Purity line and, in part due to two ownership changes at ClarityCap. Things now seem to be on solid footing.

A couple of developments: one is the Purity line which multiplies the advantages of the CSA copper end caps which are 1mm thick, and outperform every other end cap I've studied or heard. Purity increases those caps to 10mm thick which allows some 3D geometric utilization. Purity were projected to add 30% to the CSA cost at the same voltage, but in fact have approximately tripled the price of the CSA. The Van den Hull pure silver wire is now only available in Purity line whereas it had also been an option (very worthwhile) in CSA.

Dave had also guided me to testing Reliable's RTX (best) ultra-bypass caps which use a multiple parallel x series configuration for astonishingly good performance at a very affordable price. Wilson bought Reliable and has tripled the prices, which may still be worthwhile. Meanwhile Duelund has some magical stacked foil caps that are oh so good at considerably higher cost.

My challenge is choosing combinations of caps that retain Thiel's traditional value orientation, albeit on a higher plateau. Thiel is not Wilson or Magico. We won't be incorporating $1K caps.

Dave has been an enormous help, and generous with sharing his extensive knowledge. Again, thanks for the introduction.

Richard - despite the detail I put into my own responses, I am not very familiar with with very much equipment on the market. So I can't provide any meaningful comments on particular equipment and setups. I can say to avoid any amp that doesn't specifically state their 4-ohm load performance because if it could, they would say so - so it can't. A straining amp is the most common cause of tweeter failure - by far.

You trouble-shooting should identify whether or not your high-frequency problem is indeed a tweeter problem. If it is, you hit 'our' problem on the head. That Thiel UltraTweeter moving system is obsolete, and the soft-dome replacement available from Madisound has different sonic characteristics. It will work, and we will eventually have a certified replacement. There is also a drop-in midrange replacement from Madisound. My (eventual) replacements will be performance upgrades over the originals.

I suggest pulling the woofers and examining the crossover boards. Your electrolytic caps could be at end of life. Look for bulged cases and evidence of high heat.

We have proven considerable sonic / performance upgrades via crossover upgrades via parts, layout and thermal management. Circumstances have prevented timely progress, but patience will be rewarded. I suggest you keep the critters. 

Keep us posted about your progress.

rc1985 - I can offer some general thoughts. The 1988 CS1.2 is a worthy speaker, having our first copper shunting rings on the woofer and our first aluminum tweeter (with ferro-fluid). Its development paralleled the 1989 CS5 and shared technology development between those co-products. I consider the CS1.2 as definitely worthy of up-grade.

Your woofers are from Vifa and have rubber surrounds. Foam went out with the original 'O' series. The surrounds should be renewable with rubber rejuvenator. I suggest you contact Rob Gillum at CoherentSource Service for recommendations.

I have done work on the successor CS1.5, but not yet on the 1.2. I am willing to get XO information and make recommendations. However, I will caution that replacing XO components is a rather expensive route to higher performance. We might rather look at other low-hanging fruit of port aerodynamics, grille-board attachment, wire, terminals, XO placement and signal routing. We can wade in via PM if you wish.  

Thiel placed focused attention on product longevity. The caps are high-temperature / long life, and resistors and coil are (practically) immortal. Driver tinsel leads might eventually fatigue and fail, but in the absence of extreme use or upstream problems, they rarely fail. I have 01s from 1980 still intact.

For peace of mind you might pull the woofers and look around.

JA - Let’s call off the hunt for additional 02s. I have 4 pair which are serving my purposes well. As an update, I have used them as a baseline to directly compare ’normal’ vs ’phase coherent’ as well as additional approaches and solutions. The 02 has both positive-polarity drivers with the woofer following the tweeter output by a nicely timed single cycle. This is where Wilson and most others landed after experimenting with British style reversed polarities. The 02 had hand-wound coils and entry level xo components from which I mapped the upgrade territory. For the ongoing laminar wavelaunch development with Doug Pauly, we settled on fairly radically modded 02s with the woofers on a standoff and tweeters behind a wave-guide, and first order slopes. The drivers and XOs were changed to CS.5s which are better in every way than the original 02s. The cabinets have been braced via various methods, teaching effective solutions going forward.

In short, the 02s have served as a learning lab and launch pad from which we have graduated to the SCS4s which incorporate most of Jim’s late career advancements, and are now carrying the development load and teaching additional lessons in an economical, compact and more sophisticated package.

From my personal perspective, coherence lifts veils that allow my work to move forward. So, let’s consider the 02 as a lovely historical starting-point for Thiel’s journey. A pair provides daily music in my living room.

Amberwood they are. And what a nice-looking pair. The two woofers look different. A prospective buyer would do well to ask questions and get Rob Gillum involved. FWIW Rob has deep backup for the 2.2 woofer.

audio1326 - I’ll chew a slice of your query. I investigated your question for affordable amps for the CS2.2 and landed on Adcom, and specifically the GFA555. In that process I sorted out some considerations, summarized here.

The GFA555 was designed by (the legandary one and only) Nelson Pass, as were the 535 and 545. Simple, low parts count, high current circuits which depend on high-quality parts. First generation boards were built in Japan with ’best’ available parts. Later runs moved to Taiwan with lower quality parts. I don’t know the serial numbers. There is even better Japanese transistor sets and components available now. (See Audio Upgrades below.)

The geriatric capacitor problem is a real but manageable problem. A competent electronics shop can measure performance of the caps. If they’re good, they’re good.

I bought a GFA555 and a GFA555 mkII which is generally considered ’more refined’. My journey landed on the original (Nelson Pass) GFA555 being superior and upgradable (by reason of its intrinsic superiority). I would consider the added cost of a 555 over the 545 or 535 to be a no-brainer. Transients require power and more is better if circuit finesse is not sacrificed. The 555 gives up nothing and gets more power.

The winning aspect is that Jim Williams of Audio Upgrades in Carlsbad CA upgrades this amp to two performance levels. The upper level adds world-class metal film resistors to the basic job. Cost is in the $hundreds, not $thousands. Very highly recommended.

Furthermore, Bill Thalmann of Music Technology is reviewing my Jim Williams GFA555 upgrade to recommend and test further audiophile enhancements. I intend to recommend this amp with upgrades (wherever the particulars fall) to Thiel Renaissance clients as a high value, high performance amplification solution.

I will not comment on other contenders, having no personal experience - with the exception that Prima Luna is an outstanding performer for its price. (As long as you enjoy tube rolling.)

JA - I'm having a blast. I'm newly in a suitable space that was built in the 19th century as a dowel-making workshop. Good proportions and bones. Focus is on finalizing the (myriad) wire trials using (this round) the SCS4 with outboard crossover. I am again surprised how much improvement is made via simply moving the XO outboard free from microphonics and away from the driver EMF zone. There is notably less difference between wires when EMF is reduced. The wires act as antennae and quieter is decidedly better.

This weekend a young, female subject from a musical family, but first exposure to Thiel, listened to an Adam Cohen recording (Boats, from We Go Home). We listened through A: stock SCS4 and then B: outboard XO with original XO including wire). B brought some tears and comments about surprising involvement & connection.

More wire comparisons on today's docket.

JA - are you suggesting that I'm approaching middle age?

Bass is hard to get right, especially as the task is to translate the producers' projections onto our playback reality with a break in psycho-acoustic continuity. In other words, our 'ear-brain' knows our playback environment, but not the performance environment nor the producers' second guesses about our playback space. Over time, bass balance in production has gravitated toward a standard - in loudspeakers less so. There exist many products with strategically underdamped bass creating a big, loose hump in the upper bass. Think British monitors. Thiel's bass balance target was flat assuming only a floor under the speaker which neither added nor subtracted bass content. Any other assumptions are bound to be wrong because playback environments vary drastically and unpredictably.

Jim's design tools utilized free-air (hanging in a tree or later from a tight-rope) with the mic either 3 meters out or on the ground at various speaker heights. These free-field measurements were integrated with ground-plane measurements where the speaker was placed in the middle of a large, heavy truck grade asphalt parking lot (empty, after hours). Ground plane mic placed at 2, 3 and 4 meters out to average boundary conditions. Anechoic (free-air) measurements exhibit a -2dB shelf below 200Hz which comes up to flat in half-space - on a floor. The further room gain added by room reflections and resonances are matters of set-up and preference. 

The SmartSubwoofers addressed those boundary effects with corrective circuitry.

There is a family of considerations for high frequency balance that I'll save for another time. Over the decades, criticism of Thiel speakers has been toward too-lean bass (rather than vice-versa), and popular opinion sometimes favors speakers with objectively heavy bass. Bass balance is a hard question because so much depends on the installation particulars and user preferences. My intent here is to say that Jim assumed nothing about structural reinforcement or subtraction of the bass in the playback space, leaving that to the end user. I haven't found a floor interface product that honors all the factors of that interaction. Jim's working assumption was to leave that set of interactions null. His speakers state his interpretation of correct on an imaginary, neutral floor.

The Viewpoints use the same sealed bass and Power Driver as the PowerPoints. I also fell in love with those - practically unbelievable aren't they?

JA - I can share what little I know, and perhaps others can fill in some blanks. I’ve not seen a ViewPoint system except in reviews and photos.

There is one PowerDriver, Thiel’s 6.5" x 1" concentric/coincident driver. The woofer and tweeter systems are both shared by the CS7.2. This woofer has different motor parameters than the SCS, CS1.5, etc which are optimized for reflex bass. This PowerWoofer is optimized for sealed bass. It uses neodymimum magnets and the shallow, exponentially curved front cone optimized for its wave-guide functions for the coaxial tweeter. The backing brace is a straight-sided, deep cone of cast styrofoam without the back skin of the 7.2 upper woofer. Early versions of these drivers were made in the Lexington factory and featured a removable tweeter module to facilitate repair.

The tweeter is shared by the 7.2, and (I think) has the silk surround and catenary dome geometry. That opinion is part speculation and part observation. My (scores of) swept measurements show the oil-can breakup to be very controlled and above 30kHz. I’ve never seen that (great) behavior from a spherical dome or rubber surround.

Before the SCS4s, my experience with this PowerDriver was in the PowerPoints. I have 4 pair at various levels of upgrade, plus two pairs of SCS4s presently in the lab. Every individual driver measures like the same driver. Clones.

Both the PowerPoint and the ViewPoint share the 45° launch geometry where one plane of the driver’s wavelaunch is supported by a 45° wall plane at the driver rim to eliminate wall-bounce as an installation problem. The propagation into the room is orderly and organized and the bass is supported without a suck-out / bounce (which must always be managed in a floor-standing or stand-mount speaker).

The PP is flat to 80Hz with a sealed box 12dB / octave roll-off below that. The viewpoint (which I have never seen or heard) claims -3dB bass to 60Hz. That extra bass probably comes from tangential 45° mounting to both the wall and the monitor screen itself for a larger (quasi infinite) baffle. The in-ceiling (HigherPlane & PowerPlane) have differing tunings, but all use the same PowerDriver and all are sealed, so all mate very well with subwoofers. Thiel supplied either an external crossover tuned to each model or the Integrator which could control multiple model mixes. That low bandpass crossover introduced two more poles (12dB/octave) to create a now-conventional 4th order in-polarity low crossover. (Some of you know that I take issue with 4th order slopes, but it does create a powerfully practical solution.) All the enclosures are aluminum and all the crossovers use Thiel’s styrene/tin film & foil ’yellow’ 1uF bypass cap.

I have seen several pairs of SCS4 and PowerPoint speakers which all exhibit the pains of transferring manufacture from Lexington with long-established Western components and point-to-point in-house construction - to later FST / Chinese executions with diminished x progressively improving components and execution (beginning with fairly shoddy and progressing to fairly fine.)

This migration to China was done to down-size the Lexington plant as Jim’s attempt for a simplified down-sized manufacturing operation to have a chance of survival after he was gone.

I am presently researching how the FST non-removable tweeter might (or might not) be adjustable in launch-plane. Its transient arrival is a few micro-seconds too soon and I hope to ’fix’ that as a point of honor to Thiel’s goal of best-of-form phase/time wavefront integrity. I know the problem's magnitude is in the dust, but still hope to clean it up if I can.

Back to the ViewPoint. All units were custom made by trimming the extrusion to match the height of the mating monitor. I suspect (but haven’t confirmed) that the actual driver enclosure volume is fixed and the extrusion length merely acts as baffle extensions. The written and verbal reviews have been consistently stellar. I look forward to learning the particulars of crossover position and feed-wire routing. I do know that the PowerPoint XO is positioned tangent to the driver magnet(s), which is the least harmful place - whereas the SCS4 XO is positioned directly behind the magnet(s), which has the highest level EMF radiation / interaction. I am once-again surprised and pleased in my SCS4 project by the improvement gained by better physical implementation of those physical aspects. (which of course are always more cumbersome and expensive to implement.)

I welcome and encourage any of you with experience of the PowerDriver, especially the ViewPoint, to chime in with your experience, further thoughts and corrections of any of my conjectures.

Cheers, Tom

Correction to my previous post. The moving system of the PowerDriver tweeter is that of the CS7.2 with a spherical dome and rubber surround. The newer SCS4 tweeter moving system was co developed with the CS3.7 / CS2.7. It has the catenary dome and silk surround. Also, I had mis-remembered the SCS4 tweeter breakup frequency. It is about 27kHz with very nicely controlled behavior.

Richard - can you please tell us (or repeat) the dimensions of your room? That matters a lot for power requirements. 

Barnett - that room has tons of potential. Can you tell us the dimensions?

barnett - nice system. What is the ceiling height to the actual, hard ceiling?

barnett - a couple of thoughts about CS2.2s in your room.

The room is wonderfully designed. Congratulations on however you got there. The dimensions are right in the middle of the Bolt Pattern which allows for leeway regarding the contribution of your diffusion panels on the side walls. Also the Bonello Modes stack up very nicely with increasing density per third-octave frequency rise. All good for a musical-sounding space from most anywhere in it.

The picture gets murkier when considering bass, and especially adding a subwoofer. Many of your bass modes fall on whole-notes and intervals which will be accentuated and long-lasting when stimulated. Specifically you have reinforcement modes at 24, 33, 41, 49 Hz (rounded) which fall on notes at A=440Hz. By the way, they fall off-note as the room width acts smaller than 17’. It’s hard to predict exactly how your diffusers affect the functional width of the room.

These modes will definitely be made more troublesome by adding subwoofer(s) bass extension. Those will not be very responsive to equalization. So consider tunable absorption devices to address those issues.

A word about the CS2.2. Bass runs out of steam with a hard ’splat’ when driven hard with bass content. John Atkinson proclaimed in his initial review that the passive radiator and/or woofer was bottoming. He bought and used CS2.2s personally for a few years until the bass problem got the better of him. He is a bass guitarist after all. That speculation turns out to be false. The problem is real, but it fixed itself when I took the crossovers outboard for redevelopment. That ’splat’ is part of an overall veil-shimmer-overload that is part of the CS2.2. It goes away (quite gloriously) with EMF management. If outboarding crossovers is too much, I’m also working on a set of inboard solutions centering on moving the crossover(s) from behind the woofer to the bottom of the cabinet (where they should have been all along.)

Plenty to chew on here. Have fun.

Tom

I can't comment on which amp would best fit your circumstances and orientation. And I don't know the Parasound. I would always go for the larger Adcom. I will have two 555s, they bridge nicely. You do want a dedicated Audio circuit, and upsize the feed wire while you're at it.

These dimensions are actually better. Your Bonello modes / third ascendency is now full (rather than missing one step) and now most of your bass modes fall squarely between notes with fewer needing attention.

Outboard crossovers will be available as plans or kits or finished products - in due time. Not yet. We'll also be offering an internal (significant) upgrade where the XO remains inboard but moves from behind the woofer to the cabinet bottom. Not yet ready for prime time.

Yes, tunable traps will address your bass mode issues. But I have a bigger problem with subwoofers, which was shared by Jim, but practically unavoidable in the Home Theater milieu. Short rant:

The ear-brain does a fine job of providing phantom fundamentals. When the harmonic structure suggests missing lower partials, we just make them up. Of course it's better to get them actually heard. However, when a subwoofer supplies the fundamentals, they are typically a full cycle behind the upper partials of the sound package. Depending on the cross-point, let's say it's like the lowest portion of the sound emanates from 20 to 50' behind the sonic image. I surmise that a larger than average segment of the Thiel population is tuned into the time element of music. As such, we, and I speak for myself, can find this delayed bass less than satisfying. The 2.2 is reflex bass, which does that bass delay crossover at 45Hz. Subwoofer crosspoint might be tried there for no further time-domain harm.  A sealed sub (such as Thiel's own) reproduces all the way down with no additional phase rotation in the deep bass. 

barnett - the sound in your recording is not the overload 'splat'. Your noise sounds like wire buzz or similar mechanical vibration. Something may be hitting the back of the woofer or passive radiator, or alternately, the surround of the passive radiator may have come loose. There is an interior surround identical to the external one. You can feel it by taking out the woofer or see it with a mirror.

Cheers.

I am suggesting that you pull the woofer and snoop around.

Regarding the ’splat’ in general. The splat is the audible effect of a larger problem. Interaction of fields in the speaker system cause negative sonic effects at all volume levels - becoming worse as volume increases. Some have described the 2.2 as ’veiled’ and/or ’squishy’. Sorting out the internal field and propagation geometric issues lifts that veil and tightens up the squish, and the bass.

I observe two ’types’ of Thiel speakers: those with crossovers distant from the drivers (mainly woofers) and those with XOs in proximity. Those with distant XOs (think CS2.4 or 1.5 & 1.6 or 3.7 & 2.7) share a clean effortlessness that the close XO models don’t manage to attain.

I am developing various solutions short of outboarding the crossovers. Later passive radiators such as the 1.5 were removable, allowing the XO to be moved from behind the woofer to behind the PR. A new CS2.2 removable PR could address this issue. A removable cabinet bottom for repositioning an inboard XO may be feasible. Half the battle is identifying the problem. No end to the fun.

unsound - we’re actively working on the 3.5. It is perhaps my favorite Thiel speaker in that it captures Jim’s primary insights. It’s the only place I’ve ever heard honest 20Hz, in-time bass; and with a footprint of a square foot and waist high.

Jim Williams of Audio Upgrades in Carlsbad has re-worked the EQ for substantially better performance. Bill Thalmann of Music Technology is mapping a fully balanced unit with a few more audiophile touches. Balanced will require an EQ unit per channel.

Low-level crossovers incorporated with the EQ is an excellent idea, but beyond my present scope. With the new (Purifi, etc.) switching amps, multi-amp drive could be heart-stopping.

As a historical note - Thiel’s first product that never made it to market was just such a configuration - small 3-way with built-in driver-dedicated equalized amplification. In the mythical revival department I think such a speaker could rock given today’s technical resources. Not for me, though.

Another note, the CS5 (1989) followed the 3.5 and could have been equalized, but Jim had become discouraged by the reactionary response, especially from sophisticated reviewers, and chose to get sub 20Hz sealed bass without EQ at the cost of 2ohm impedance at the bottom end. The subsequent CS3.6 in the early 90s became the watershed. Rather than developing an upgraded EQ to continue the model 3 tradition, Jim opted for reflex bass. It was properly implemented, etc. etc. but nonetheless diluted the commitment to phase / time coherence that had built the company’s market niche and reputation. I considered that a market-capitulation at the expense of root principles.

Imagine a CS3.5a using the 3.5 woofer and upgraded EQ, an updated version of the double-cone 3.6 midrange and an updated (CS5, 3.6, 2.2) UltraTweeter using the 3.7 / SCS4 moving system. I bet you’ll like it.

audiiofilio123 - two thoughts come to mind. A: you like more bass than the CS.5 delivers or B: something is wrong.

Your room is appropriately sized and I'll presume your equipment is up to the task, with the caveat that the CS.5 has a low impedance requiring an amp that is comfortable driving 4 ohms. Insufficient current delivery can manifest as bass-starved. 

Thiel speakers are often considered bass-lean. Their tuning has a Quality Factor of .707 which we think provides the 'best' compromise between bass quantity and quality. However, the CS.5's target market was Home Theater and its Q target was '1' - providing a 3dB hump before roll-off. So, it has more apparent bass than Thiel's music models.

You might play the same-channel or combined mono into each speaker separately to ascertain they both have the same bass output. Let us know what you learn. That speaker punches way over its weight!

Anthony - please keep lurking. Sorry the going has been so slow. We're close to an upgraded equalizer and working in other areas. First we need a tweeter based on the UltraTweeter used in the CS5,3.6, and 2.2, now obsolete and unsupportable. Next we need a midrange built on the CS3.6 dual-cone. Upgraded XO is taking shape with considerable research on this and other products. Currently using the SCS4 as my workhorse. It will be worth the wait.

-Tom

anzen - it’s a complex, multi-dimensional puzzle. In terms of pure signal propagation, PCBs outdo most PtoP layouts, especially if one-sided. PCBs shorter leads create smaller antenna effects. A new insight (to me) is how signal chaos is created when P to P has inputs and outputs entering the same solder-lug from the same direction. Two-sided layouts allow that distortion mechanism to be eliminated - with care and extra hassle. A requirement in this Renaissance project is the variety of quality levels supported in the same package. I want my layouts to accommodate various sizes of components. And on and on.

Placing the boards vertically facilitates thermal management via convection air over all component surfaces. Heat dissipation is a big deal for circuit stability under load. Edge mounting is more difficult and more optimal - but worth the effort since I’m aiming for a higher performance plateau(s) than original Thiel products. I’m pleasantly surprised how much room for improvement exists in the same platform with the same drivers, circuitry and cabinet.

Best material for XO panel is Panzerholz, German compressed and heat treated wood product that is extremely strong and internally very well damped - and very expensive. Next best is 1/4" masonite which comes pretty close at 5% cost. Outboarding significantly reduces need for vibration control. For inboard panels I’m suspending the panel in rubber grommets to decouple from cabinet panel resonance modes. Again, I’m pleasantly surprised how these various aspects accumulate into cleaner performance.

To answer more directly: I will be using wood-based point-to-point panels exclusively for their strength, superior electro-magnetic performance and layout flexibility.

Tom D. Indeed we avoid aligning the geometric plane of the XO panel normal to the velocity propagation wave. The vertical boards are aligned to present minimal area to the velocity wave. Adjustable resonance control is an attractive mechanism, both for the XO and the drivers. My approach tries to control the forces without the complication of user adjustability. I have prototyped tunable control and the future could lead there. The combination of bronze and Panzerholz works very well to dissipate energy while providing considerable strength. Progress is being made.

I agree. Outboard has lots of advantages. But I also want a significant upgrade that doesn't require the outboard box. I'm able to get a considerable inboard advance over stock. Outboard is better.

The floor interface is a complex problem, especially considering that each floor is unique and sometimes substantially different from other floors.

Unsound - that's a complication I hadn't anticipated. Wire is highly reactive and susceptible to noise from multiple causes. Having long runs between the XO and speaker would require specific engineering to maintain target performance.

I've become a big fan of shortest practical speaker cable runs via a pair of amps part-way between the centered source rack and speakers, with XOs part-way again. I've been presuming 1/2 meter umbilicals. Could be a meter.

unsound - thanks for the input. What may not be self-evident is that changes such as these wire lengths require system tweaking. I had presumed that the system would be optimized around some fairly short umbilical length, but am open to reconsidering: perhaps a short and long version such as 1/2 and 1-1/2 meters.

Of general interest may be how Thiel approached this tweaking problem. We assumed user preference for amps and cables. Voicing and final adjustments consumed months of iterative listening tests. The (sometimes maddeningly) long delays between projected new product introductions and actual roll-out included this micro-voicing to levels considerably beyond measurement. Add to that the unannounced XO revisions during a product's life for further refinements. These micro-voicings take internal wire lengths into consideration along with the myriad other variables and interactions.Outboard XOs extend the complex cable contributions from a few feet to many feet. An additional system variable.

I'm glad you brought up this issue - I'll add it to the list of considerations.

I'd rather you not as well. I want to address the relevant aspects of sophisticated performance and physical vibration control is a real as well as manageable issue.

Rob - no. My observation was general across the few speakers I have seen. NO gasket on the 2.7 fits Thiel’s strategy perfectly. It’s likely the paint is enough to seal the air. BTW are your driver screws magnetic?