Thiel Owners

Silvanik - don’t take it personally, I likewise am getting dead air. But, I have a report that Rob is alive and well and will soon be back in the saddle.

You might like the back story. Lexington is a medium sized city on a limestone savanah. Going east the Bluegrass Savanah gets rough and hilly toward the Cumberland Plateau, the western foothills of the Appalachian Mountains. Several years ago Rob bought an old hill farm in the knobs with a hunting cabin where he spends his private. It's now deer hunding season in Kentucky, and Jeremy reports that Rob is off hunting quite a bit lately. Jeremy is the son of Walter Kling, who co-founded Thiel Audio, was part of the communal undertakings in the beginning as well as my replacement in the later years. Jeremy grew up in the company and worked there on and off including juggling alligators at the bitter end. Jeremy says that Rob is alive and well and he’ll let him know that we’re looking for him.

Take care.

silvanik - surrounds are very critical on tweeters. I don't remember whether those are rubber or a synthetic elastomer, but stiffening could be an issue. Your REW measurements shouuld tell you quite a bit, especially via FFTransforms to give you time domain information. If you can get higher than 20K, do it, since that tweeter goes to 28K and then some and that ultrasonic range might yield hints.

That UltraTweeter was used in the 2.2, 3.6 and 5. Madisound has the companion fiber moving kit, which is said to drop in without XO changes. Birger Jorgensen of Vifa, who codesigned that with Jim, took great care for the fiber and metal domes to behave as similarly as possible. One of these days there will be a replacement, but that motor will house the fiber dome if you need one sooner.

Tom

Awhile back someone asked about Thiel binding posts, for which I don’t remember seeing an answer. From my limited perspective: In the beginning we used a post with minimal metal (rhodium over brass) and plastic caps. We got a lot of flack for "cheap binding posts", which in fact they were not. In the CS5 development in the late 80s we revisited those posts and chose them again rather than larger beefier, gold-plated ones. That was before we realized the hazard of eddy currents caused by greater cross-sectional area relative to the input and out wires. As we worked with Finite Element Analysis, eddies in voice coils came to light and we changed from aluminum to nomex. But it seems those lessons were not applied to binding posts, but I wasn’t there and can’t comment on whether attention was paid.

I see a progression to beefier posts in later models. My CS1.6s look like gold on brass via the file test.

My present work identifies the binding post as significantly upgradable. I developed a CDA101 (best available) 5mm bolt connector, but am getting some GR-research tube connectors for test, which should be better yet. They place the wire ends almost touching while adding very little bulk with a thin tube of CDA-101 copper. They look like what I would have designed if Danny Ritchie hadn’t beat me to it.

I have learned that signal path integrity matters quite a bit.

JA - you asked for more about Walter Kling. As I’ve mentioned here, Thiel Audio grew out of an intentional community and its wish to employ ourselves in a mutual undertaking. Besides myself with my wife and toddler daughter, there were my brother Jim and his girlfriend as well as Jim’s school-friend Walter Kling and my fellow Marianist Brother friend Fred Collopy. I had established some viability as a designing craftsman and included Walter in our Conceptions Design Studio. Kathy took on Marketing and relationship-development. Fred was a very early computer pioneer - everything about computers was pretty new in the late 70s. Fred signed up for business development. He later went on for a masters from Wharton in Decision Science and a PhD in Entrepreneurial Development. Walter was a multi-talented artist-craftsman with proficiency as a painter, potter and sculptor. He later went on for a degree in Architecture and practiced that until I wooed him back to Thiel Audio.

Of note is that none of us would have signed up to attempt Thiel Audio without the group commitment to combine all talents and risks for the common goal. We were in our 20s, and along with enthusiasm, change is what happens at that phase. Fred gravitated toward his advanced education and we all absorbed his roles, more or less. But his talents and inputs would have been more than welcome. Walter’s wife had left him with two young children, Jeremy and Jennifer. About 5 years in, he ran afoul of Jim’s alpha-male requirements. His departure, around 1980, was devastating, especially for me. We were moving out of our farmhouse beginnings to our "real" industrial space - a time when Walt’s forté of fixtures, tools and methods were needed. That all fell to me in addition to my teaching-training-coaching - process development responsibilities. It’s hard to overstate how much and how varied is the work required to push the limits, to try the untried and grow a company without outside capital or resources. Jim was an extraordinary designer, but those designs must be implemented in effective and reliable ways, every day, all day. Walter was part of that equation and his absence nearly cost us our company. One hundred hour weeks became the norm.

Around 1987 we expanded from around 5000 square feet to 15,000 square feet including moving everything while adding new capabilities and doubling capacity, all without missing a day’s work. I was able to hire Walter as construction manager for the physical plant part where I took on the logistics, project management and production design part. It worked. (Rob Gillum came on in that expansion as a young rookie and spent his working life in most of the aspects of manufacturing.) The next few years included cementing Walter back into the company, where he remained until Jim’s death in 2009.

Companies prune their PR around the story they want to tell. You’ve rightly heard a lot about Jim, but there was a strong team around him making the whole thing work. Walter was a central player on that team.

big_greg -I agree with Cascadesphil that Thiel’s can be more sensitive to input signal than most brands. In my opinion, that has more to do with their coherence allowing the ear-brain more scrutiny. But that discussion is beyond the scope of this response. Let’s assume that your signal is fine.

You stated your listening distance as 9’, which is fine. Design distance is 10’, and the closer you get, the more critical your ear height becomes. The propagation triangles have to resolve at your ear. Design ear height is 3’. If you sit high, you will get a treble-heavy and non-time-aligned wavefront. That’s a design constraint of phase coherence with multiple drivers and why Thiel gravitated toward coax treble sources in later models - with their own issues and challenges.

Another critical factor is early reflections, which likewise become more critical in a coherent design. The speakers are designed as point sources with very broad, even dispersion characteristics. As such they require at least 3’ between the tweeter and any reflective surface. Side-wall reflection is most often a culprit; absorption at the reflection point helps - a lot. Similarly, a low ceiling and/or hard floor can be problematic. In a small room, consider a long-wall layout.

Note that Thiel’s tonal balance is tuned for listening off-axis. At design setup the ear is 20° off-axis which is where the power (in-room) response matches the direct response. That straight-ahead position requires at least 3’ to a side wall, more is better, and absorption helps. In my experience most people aim them slightly inward to mitigate side-wall reflections. My experience is that straight-ahead with narrower speaker to speaker placement solves the issues better. Such particulars of setup are far more germane to performance than are particulars of equipment (unless grossly inappropriate.)

Thiels are articulate and precise, and for those who appreciate that, they can be very musically engaging. Many speakers are designed to be forgiving of problems - Thiels are not. Our goal is to faithfully reproduce their input signal - unvarnished. Much like a recording environment must be carefully optimized to capture a proper record of the recorded event, the playback environment must also be optimized. In the hi-fi hobby I believe we could create far more satisfying musical immersion by working on our environment and setup rather than looking to gear changes.

Keep the faith - the results can be wonderful.

@petaluman - the 03a was a sealed box design which produces a 12dB/octave bass rolloff. I think the tuning was .707 Q which Jim considered critically (properly) damped and some call overdamped. I don’t remember the unequalized -3dB point, but a fit curve could be overlaid knowing that the EQ curve was 12db/octave with the peak centered at 30Hz and then reducing symmetrically to zero boost. I would guess the upper blend point would be around 100Hz. So, without the EQ, your bass will roll off at 12dB/octave below that blend frequency. With the EQ, you get critically damped flat response to 30Hz, then rolling off at around second order. Excellent bass. The EQ did not add cut, it merely ceased adding boost.

The Thiel eq was discrete circuitry and neither the EQ or any Crossover had op amps. However there was an aftermarket product called "Golden Flute" built in a brass tube and powered by a wall wart which used op amps to achieve the same goal. I have heard that they were well liked and successful; but I know nothing about their particulars.

I know of no hot-rodded versions of the 03a EQ. I do know that the CS3 EQ, which performs the identical function, was more sophisticated in its circuitry and execution. And that the subsequent CS3.5 was the first generation to be direct-coupled with no capacitors in the signal path. The 3.5 had variable cutoff points at 20Hz and 40Hz to side-step troublesome deep bass room modes if necessary. The 03a and CS3 families are conceptually the same product - with the addition of the Coherent Source nameplate. Even though the woofers are different, their Thiele/Small parameters and enclosure size are very similar. An EQ for the 03a, CS3 or CS3.5 would work for your 03a as would Golden Flutes for any of those same models. The 01/ 01a/ 01b also has the same 30Hz x 12dB boost and would work. Avoid the model 03 EQ, since that was both ported and boosted, so its parameters are different. I am working on a 3.5 EQ upgrade for significantly better performance than stock.

The EQ transforms the speaker from significantly bass shy to excellent in every way.

tmsrdg - yes sir!

sdl4 - remember, I’m an electronics layman

1. Yes, an AHB-2 should drive your 2.2s. Medium-size and real-world are vague and subjective terms, but here’s why I say yes. The (clipping) protection circuitry is so fast and sophisticated that I can see the lights flickering without hearing any dropouts. That takes 100dB in my functionally rather large room. In case you don’t know, my room was weird. I built a resistive-wall (leaky) 4000 cubic foot room within an 11,500 cubic foot floor of a 38,500 building with various openings between the floors. Virtually no modal interactions, but hard to characterize its "size" since the direct space was moderately sized and the larger spaces were acoustically secondary. All that said, my preference and regular use mode was each speaker driven by a bridged AHB-2. But, a single stereo amp was adequate for moderate (85dB peaks) listening.

2. The AHB distortion profile is extremely clean, but I lack technical fluency regarding your specific question. I point you to the Stereophile and AudioScience reviews, among others, to see their qualitatively (stunningly) low distortion. It took me some time to realize that the bass balance was not thin or dry. Other amps (of my experience) add harmonics, especially noticeable in the bass. On a related note, the Benchmark DAC3 and Preamps-4 take additional measures to reduce second and third harmonic distortion. John Sieu’s essays on Benchmark’s site are quite informative.

I don’t know the Atma-Sphere amps, but I do know that the Benchmark class H topology acts differently than do ’normal’ amps, even full class A. JS sent me some distortion graphs illustrating that the 1 ohm profile is virtually identical to the 8 ohm profile. No, that’s not normal, and no, I can’t lay my hands on those graphs or explain the technicalities. But, yes, the amps do sound that way. With JS’s coaching, I rigged an experiment to account for my initial perception of ’better’ sound in stereo than bridged. The output impedance drops to half, lowering the damping factor. I shortened the speaker cables from 12’ (stereo) to 6’ (bridged) and the SQ difference seemed to vanish. This THX amp topology is amazing as tmsrdg said - all that weirdness goes away leaving only the music. Sorry to gush, it’s just hard for me to believe.

BM has a generous return policy in case their power doesn’t meet your needs. Or, if your budget can be stretched, two of them make for serious audio nirvana.

usound - Thank you for your thoughts. I share your concern about the power ratings, especially into low impedance loads like the (maddening) Thiels. To learn more about the discrepancy between the AHB-2 non double-down performance vs. my intensive auditioning experience with it, I engaged John Siau in conversation. Among the lessons Iearned was Benchmark’s eccentric power measurement protocol which goes like this:

They rate the amp at 0.0003% THD+N into all loads. In other words if it doesn’t shut down, that running spec is met with no allowance for additional distortion. The amp exceeds that 3 zeros spec at the power limits you cite (which obviously decrease into decreasing impedances.) But, if the traditional 1% THD+N spec were used, it would double-down as we want and as I (among others) experience it doing. It sounds and acts like it is doing what John says it is, and BM is very conscientious about its claims. Their internal tests go to 1 ohm continuous and their customer service tells me they are viable to 1/2 ohm resistive loads. All Thiel models are extremely resistive via Zobel networks on all drivers - which push the amp limitation from distortion-limited toward heat-limited. I have shut down the AHB-2 via overheating, but it required louder levels than my comfort zone. Admittedly, my installation maximizes radiation and convection cooling. When John evaluated the load graphs of Thiel models considering phase and impedance, he recommended stereo-only except for the CS7/7.2 = stereo or mono. Due to the BM feed-forward distortion reduction the only advantage of bridged mono is 4dB greater headroom before shut-down, no increased distortion as in all other amps.

I recognize that these claims run counter to everything we know about amp behavior. Have you seen any second-party lab tests using the traditional 1% distortion limits? I would love to see those results. Note that I (among others) have requested a higher-output AHB-2, but John is firm that ’it isn’t necessary’. Perhaps from BM’s perspective of primarily pro market and inability to meet demand, he has a point.

Note I am not arguing that better and more suitable amps aren’t out there. I imagine they are. My personal situation is needing an amp that drives the loads while telling the truth. The AHB-2 does that for me, at a price I can afford.

Here's the AudioScience Review

CS5 electrolytic crossover caps. Whoever spotted them was correct, despite the promo stating 'all film caps'. There a few, always shunts to common and always bypassed by the yellow 1uF styrene/ tin foil cap. There is also an iron core inductor in the subwoofer feed. I recall that layout space was the determining factor for that huge board. Outboarding that board would probably be required to allow upgrade of those electrolytics with film caps. Those speakers are 30 years old, nearing their hypothetical service life. None are in sensitive / dangerous positions. 

jafant - first a disclaimer. I have no real experience with the WATT or any other Wilson products, beyond interested observation. My connection is primarily one of curiosity. The WATT has been considered an audiophile reference from its beginning, and called such by JA in Stereophile. It turns up in very discriminating systems and spaces. My curiosity revolves primarily around what all of us at Thiel would hear from it and see in its measurements, which is a flawed design, a broken speaker (in its early iterations), from our point of view. The WATT/PUPPY illuminates by contrast how Thiel never really fit into the high end mindset. We required even-handed solutions to all identifiable aspects of performance and cost, whereas much of high end tolerates and even extols excellence in some areas regardless or at the expense of other aspects. 

The Wilson approach has certainly won the day. Most high end speakers would flunk Jim Thiel's first-pass analysis. They are plain wrong in many ways while being outstandingly good in some ways or other. But Thiel didn't engage in competitive analysis - at all. We had our hands full doing our own thing, independent of whatever 'the market' or 'the times' seemed to relate to.

David Wilson was a star. His audiophile recordings were extremely good. He with his soprano wife Cheryl moved in high circles. He had a ready audience and nearly demanded respect. On the contrary, at our first 1977 CES we came out with a lovely corner suite, and garnered a positive, encouraging response from many attendees; but there was no safety net. I remember an establishment industry person asking to general amusement why we were 'here'. 'Shouldn't you be barefoot and pregnant back home?' Kentucky wasn't seen as legitimate compared with the genesis of most aspiring companies. I'm not complaining, Thiel received solid, constant support and encouragement from the audio press along with pioneering dealers who wanted something different, what we wanted to provide. My point I guess is that if we had come from Kentucky with products that contained the design shortcomings of early Wilson products, we would have been summarily dismissed.

My comments are more general sociological observations than any particular analysis of any particular products. The marketplace is an ecosystem unto itself and companies make their ways however they can from wherever they begin. Wilson was a trail-blazer into a different sphere than where Thiel lived. We lived where we were comfortable, creating products we hoped would bring value to people with whom we identified.

And here we are. What a pleasure to be here.

@sdl4 - circling back to your inquiry about driving 2.2s with a single stereo AHB-2. I answered yes you could. But that answer is close to dangerous advice. Granting that it all depends on room size and loudness levels, one amp could easily run out of steam. The 2.2 has a very flat impedance curve that only drops below 4 ohms from 100-200Hz. Its sensitivity is stated as 86dB/2.83volts. Thiel recommended 50-250 watts into 8 ohms capable of doubling to 4). I have often driven them with one AHB-2 without issue. But my room (which I have described) is quite difficult to assess due to its unconventional construction. I also rarely listen above 85dB (VU peaks).

What gives me pause is the assumptions of a normal room and levels. This week I am running trials on CS3.5 equalizer upgrades. My amps are the PSA Stellar 300 @ 150/channel and the Benchmark AHB-2 at 100, which has been discussed here.

The 3.5 with EQ goes flat to 20Hz; its sensitivity is 88dB with a flat impedance curve that drops to 5ohms between 100 and 1K. It should be an easier load than the 2.2. In a room I will describe below, I can run out of amp about 100dB (VU) peaks.

This room is an uninsulated (leaky) camp with openings that somewhat consolidate the place into one acoustic space measuring 19’x 30’x about 9’ average ceiling under a shallow gable roof - let’s call it 5000 cubic feet. To fill this space at satisfying levels without worrying about stressing drivers via distortion, I would want a pair of bridged AHB2s.

Underpowering is what kills drivers. I appreciate unsound’s push-back against my marginal recommendation that one would be ’just fine’ for you. Perhaps you can find an AHB-2 to borrow for trials. Let us know what you get.

sdl4 - unsound has a good handle on the factors. But there are even more complexities. Jim did a lot of work on that balancing act including the unusually low reactance of his early bass tunings. The phase angles are quite low and at higher impedance where the amp can handle it better. But to your question: not quite even in today's world.

This afternoon I drove the system up to flickering clipping lights on the stereo AHB with Chris Thile - Edgar Meyers 'Bass & Mandolin' album. The calibrated Benchmark preamp clipped the AHB @ 2dB lower with the 20Hz equalized signal compared with straight-wire / no equalizer. I got similar results with Patty Larkin 'Strangers' World' cuts which are mixed for 'impressive' bass impact.

Jim's intent was to get that extreme and excellent bass extension without significant penalty. And it worked in the 1980s. Remember that vinyl was still the standard which is bass limited by cartridge tracking ability. But the landscape changed not only with digital, but with showcase albums from David Wilson and others which were engineered for 'unbelievable' bass. The Watt/Puppy could handle the real cannons in his recording of the 1812 Overture, but Thiel's model 3 would bottom out because they actually reproduced those deep frequencies. Similarly digital bass could go deeper with more power, and was therefore no longer predictable. So, even though Jim took great care, and his parameters balanced out to require little to no more power in the bass than that required by midrange transients, when the signal landscape changed, the elegance of his solution became practically less safe.

It's not that Jim didn't know about CDs dynamic peaks, it's more like he was trying to cope to keep the equalized bass viable. That woofer was the first driver with his magnetic shorting rings for a more stable magnetic field and it had a huge overhung coil to handle the required excursion. The overhung motor requirement was the largest cause of the end of eq. He figured out the drastic distortion of reduction of underhung motors and powering long gaps was virtually impossible because rare-earth magnets weren't yet in the mix. Long excursions and underhung motors contradict each other.

Phase angle was not a thing in reviews at that time. In fact Jim goaded JA / Stereophile into measuring phase at all. JA and Larry A came to our factory in 1988 to spend a day in Jim's lab learning why he thought it mattered, how he measured it and so forth - and they gradually entered the arena of phase and time.

A promising solution to the dynamic limit would be to match a powered subwoofer with probably a second order crossover at 80Hz to limit the woofer excursion while getting true integrated bass to 20Hz.

bobschliff - email Rob Gillum at Coherent Source Service. He knows.

unsound - that self-surround does break down over time. At the time we were trying to minimize the standing wave reflections and non-linearities between the cone and surround as well as ultraviolet deterioration of soft plastics. It worked well when new, and held up pretty well. Natural rubber turned out better.

Those woofers are bomb-proof, rarely fail in use.

big_greg - I’d like to address your brightness problem. It is odd that the Thiels alone are bright among good company. It seems reasonable to assume a speaker problem. I don’t know the 2.3s, but hear that the original tweeter could be heard as ’strident’ and the vented replacement 'fixes' it. I do know 3.6s and the only brightness, stridency, etc. I have found is from a baffle surface propagation phenomenon which is greatly reduced with grilles in place, and is being addressed in my present work.

There is the possibility that something has gone wrong in the XO which is disabling a shaping network in a tweeter or midrange. A physical XO inspection might be in order. But such an instance would be rare, and becomes statistically next to impossible in more than one speaker.

So here's another thought. I’ve had an interesting (disorienting) experience since getting my Benchmark front end yesterday. First of all: it seems phenomenally good compared with the PS Audio Stellar gain cell / DAC and S300 amp. BM is better in every way. The ’interesting’ part is how much more obvious are the cabling differences. I have some Straightwire, Audioquest, Morrow, Mogami, and the standard Benchmark / Canare that came with the stack. Hard for me to believe that the ’normal’ $35 BM interconnects seemed so much more ’right’. Same goes for the speaker cable (with not as much comparison.) I’m in cognitive dissonance since I’ve been doing wire for nearly a half century. I am presently evaluating execution levels in the 3.5 equalizer within two driving systems. The big tip-off was the Bass & Mandolin cuts. I have played the mandolin and was involved in the development of the Collings line. The nuances of chop and gulp, box and ring are in my ear and under my skin. I stopped blaming the equalizer when I added ’straight wire’ to my protocol. When I got to the all-BM / Canare cables things settled into that remembered musical reality. It sounds right and good and I’m not missing those audiophile traits that might be artifacts.

My point is that before you give up on the Thiels, you might make some comparisons with BM cables. The phase / time coherence of all Thiels puts the ear-brain in a different hearing space than do non-coherent speakers. I’ve addressed this idea before, and I admit it can sound hoaky, but my consistent experience over these many decades is that when the system gets right, the speakers disappear. Something smells wrong about your Thiel brightness problem. I hope you can fix it.

Good luck, and keep up posted.

Cables are indeed fascinating, and their interactions with sources and load are beyond my understanding. And a lot of them don't work well with Thiel speakers. For the record - much of my approach to this playback thing holds cost and cost effectiveness as central values. I have auditioned some wonderful-sounding cables (including Cardas) that could simply not stay on my radar for reasons of cost alone. Independent of my own constraints, I live in the world of Thiel's classic orientation: creating musical tools and solutions that give access to sophisticated, satisfying musical playback experiences within the budgets of ordinary people. 

I have some samples of Iconoclast to get into my works. They use similar principles to those Kimber from the 80s (don't remember the name) that were woven, flat, and cost $1K/pair-foot. They were mind blowing. I'm looking forward to the Iconoclast because of their particular pedigree and geometry. They may serve as a reference of reference-level internal hookup wire. But . . .

roxy - I actually don’t know the name. It was his prototype he brought to CES, possibly around 1980. We used it in our show system and took some flack because the cable cost more than our speakers and power amp combined. Seems like it was a woven mesh, perhaps 3/4"- 1" wide x very thin. I remember Ray saying the insulation (perhaps Silicone) was applied with a mouse during layup. It may not have yet been finalized or had a name. Whatever it was, his cable lifted the veils, tightened the image and the bass and pulled out the stops in a jaw-dropping way. There were industry folks in the room (this was Thursday, setup day, Chicago, June) who were all impressed. Remember this is when ’the industry’ had not yet admitted that wire could matter, therefore it was snake oil. I do remember that Ray wanted us to use his wire for internal hookup wire, which seemed to cost more than the drivers. Out of the picture for us. Sorry I don’t remember or ever knew any more about a model or name.

Oh yes, I told Ray about Guarana, the Amazonian stimulant, and he eventually imported it and made a product called ’buzz gum’. Guarana is a left-handed caffein-like molecule that makes work in the Brazilian jungle possible (and pleasant.) To this day I keep some buzz-gum as a conversation piece to spread around our woods’ work crew when I was producing tonewood. . . . But still no cable name.

I have learned that the speaker cable in Thiel’s listening room was Kimber Type 88.

Let’s grant that Jim, et al, judged that to be his best choice along with his Krell FPB600 and as I remember a Sonic Frontiers preamp. Whatever the gear was, it certainly out-priced the relatively modest speakers. Jim’s logic was that the better the speaker, the better it translated its input signal. (Period). His job was to make that transducer and the job of others’ was to make their parts of the chain to do their jobs. Clean logic, dirty implications, especially when espousing a speaker (and the implied system) for everyman.

The problem gets hard for real reasons. Phase coherence allows the ear-brain to scrutinize everything - I have itemized some details in these pages, but basically everything goes under a sonic microscope. Getting it right becomes a far more complex puzzle, but so nice when pieces fall into place.

I won’t be buying any Type 88, but I will be testing some Iconoclast along with my poor man’s double helix contender.

sdl4 - TPC was normally called Electrolytic Tough Pitch (copper), which is motor / magnet wire. The others are more refined in their impurity, crystal and surface parameters. There is a case to be made that ETP is good enough as long as the other parameters are executed with care. I would have to be convinced due to my personal experience - but it is practically impossible to sort out all the various components of what makes a wire 'better'. My present Iconoclast sample is ETP/TPC.

mrfizziks - I would have to do a web search. Rob would know. BTW, Rob is working toward getting help to keep up with business. If necessary, we could plug in a CS1 woofer, which is conceptually and practically the same driver with some Thiel refinements. I could compare the XO circuits for compatibility. But, first try to run down the original Credence. I know we had rebuild kits at one time . . .

@improvedsound - PM sent.

Hello all - in response to some questions from some of you who are upgrading your crossovers, I’d like to post some of my reflections. I have uploaded a CAD drawing of my modular outboard crossover cabinet to my virtual system. That drawing is empty of components, which can be added as appropriate. The cabinet system holds individual boards for each driver and can be size-scaled as needed. It is simply form follows function as were the speaker cabinet designs.

First some general background.

Upgrading a crossover is fraught with uncertainty. All components interact with all other components, and all components have some amount of other electrical elements. In other words, a cap doesn’t just have capacitance, but also has series resistance, which varies with load, etc. Foil coils present differing resistance and capacitance than wire coils, etc. Re-balancing those elements after any change is highly demanding, and beyond my knowledge. I try to stay out of trouble by being cautious and self-critical of sonic changes. These considerations are why Rob is ’unenthusiastic’ about cap changes, or any changes. He has made his own determinations whereby he wants to stay with Jim’s results. I am cautiously venturing further and value all feedback from you who brave the territory.

As a general comment, I will be repeating some things I’ve said over the past few years here, on the chance that repetition might be better than omission for some.

Why consider outboard? Every performance parameter is improved by getting the crossover away from the electromagnetic, vibratory, hot stew inside the cabinet. The crossover is built-in for cost and marketing considerations. I have developed three solution styles to address the problems.

1: Hang the crossover on the outside/back of the cabinet (with a covering grille)

2: Locate it in a ventilated and isolated sub-base

3: Locate it in a separate enclosure a couple feet behind the cabinet.

Let’s look at some problems.

A: Temperature fluctuations are destructive. Components’ electrical values change with temperature, more than a little - enough to make a notch filter not match its target driver resonance, etc. A hot resister can influence or even melt a cap. Driver coils can burn out or come loose. Note that all elements generate heat, not just resistors. Jim used small-gauge coils where appropriate because their dissipation factor is better than a small resistor. Heat dissipates by convection into a cooler ambient environment, but about 80% of the heat is dissipated by direct radiation. So, keep components away from other components and away from elements that reduce direct radiation. Note, we can increase dissipation by raising components off the board for 360° radiation and convection. This improvement is only possible when we isolate the xo from the vibratory environment in the cabinet where all components must be locked down.

A special case is temperature inside the sealed cabinet, which can reach about 200°F. I am venting the enclosure via small diameter air inlet on the bottom and outlet on the back near the top for convection air-flow. Long-term sessions remain more sonically consistent via more consistent ambient temperature.

A difficult case is cooling the voice coil. An aluminum VC former connected to aluminum diaphragm works great - the cone can get hot enough to sizzle spit. But resultant electrical eddy currents are sonically degrading. The bullet phase plug (such as the 3.6 midrange) allows airflow to the coil. I have developed a heat sink on the back of the cabinet for each feed wire to draw heat directly from the wire while electrically isolating them to avoid eddy currents. The motor structure gets surprisingly warm in heavy use. I’m hooking a drain wire from the back plate to a heat sink on the cabinet back to dissipate heat as well as drain the electrical charge which accumulates from the relative motion of the parts.

Thiel speakers have always been considered best as small signal systems most suitable for simple music rather than rock or Wagner. Taken together these heat management methods move Thiels toward more robust use.

That’s all for now. Next session I’ll speak to EMF and other fields.

@improvedsound - what a lovely playback system you have! I am honored to have Thiel speakers in such sophisticated company. I'm not sure how much knowledge there is about Antelope and Merging in the American hi-fi marketplace. I love them from the world of high-end recording. I especially appreciate the Merging converter due to its roots in the Nagra DAT technology, which I used heavily around Y2K when high resolution audio was in its infancy. We chose 24bit/ 88.2kHz because Redbook 44.1kHz rendered more cleanly with the simple 2X math. Anyhow, I love your gear and wish that you might talk about how you approached your 2.4 upgrade, what you learned and what you achieved. Thank you.

beetlemania - these tweaks have accumulated over the few years we've been working on these upgrades together. The 2.4 is in a relatively good position among its peers, with its 2 crossovers at distnace from each other and from the woofer. Also, those boards I made for you placed the resistors centrally, away from caps. You might consider a through-vent in your cabinets - after I work out the particulars - to draw cool air from the bottom to exit near the top-back. More learning required.

JA - I don’t know what to say! That is quite a kludge. The WAMM became a pinnacle of audiophile extravaganza. It (Wilson Audio Modular Monitor) launched an enterprise that audio history will be chewing on for a long time.

My personal perspective is the degree of difference between Wilson and Thiel, along with the huge success that Wilson achieved / and continues to achieve. Whereas Thiel spent considerable energy containing costs, shaving margins, internalizing capability, and working toward balance of all aspects of sound reproduction, Wilson embraced filling an affluent market niche where higher price was a fundamental advantage and some performance aspects could be ignored. Jim was especially flabbergasted how ’the market’ could forgive the WATT’s 0.33 ohm highly reactive impedance at 2kHz.

I have a vivid memory from the mid 80s when I attended Wilson’s introduction of their first generation WATT (Wilson Audio Tiny Tot), before Puppy came out. That opening demonstration was illuminating in so many ways. The Corian cabinets with lead damping were impressively inert. The sound was nowhere near flat - and the assembled audience was so avidly enthusiastic. David Wilson’s presentation included justification of the $5000+/pair price ($12K today) in terms of cost, including "more than $250 / cabinet for just the machining of the mineral loaded polymer baffle." Since I knew the material, I would have been embarrassed to claim $5 machining cost. And so on and so forth.

I’m not chewing sour grapes, merely expressing my personal astonishment of how the brand was embraced from the outset and over the ensuing decades. A NYC dealer once told me that "someone with a half million $ to spend needs to find a half million $ product." Another regaled me regarding how Wilson had done everything right.

Thanks for the question. I am reminded that we ordinary folks live in a very different world than some others. And there may be more of them than us judging by the direction the market has taken toward extremely expensive offerings. I might add that, with the exception of happenstance convention and dealer showroom sightings, I have never actually experienced Wilson music playback. Perhaps it’s wonderful beyond words, and I’m merely expressing my lack of sophistication.

For the love of music.

petaluman - I can't be of much practical help other than referring you to Rob Gillum at Coherent Source Service. If the 03a schematic exists, he would have it. There is an aftermarket op-amp EQ called the 'golden flute' in a brass tube. Here's a little context.

The 01 series had a 30Hz EQ for which I don't remember the topology, but I doubt we used op amps. The 03 was 'improved' and also different in that it was a combination ported box with optional EQ. The 03a was sealed and its 'improved further' EQ came with the package. I do know that the CS3 is discrete with all film caps, etc. but may have caps in the signal path. (I have a unit, but no schematic.) The CS3.5 EQ is all discrete, all polystyrene caps (except for the unregulated power supply), metal film resistors and direct coupled - no caps in the signal path. That unit is presently being upgraded. Now upgrades for all obsolete transistors are available and soon there be will a significant upgrade. Still working on it.

So, in general, the EQ keeps getting better, but the only schematics I have are the original 3.5 and 3.5 Renaissance in process.

As I've mentioned, the woofer / enclosure parameters are conceptually similar for the 03a, CS3 and 3.5, but not for the 03 due to its ported bass. Let us know what RG says.

For the record, JAFANT provides this forum for which I am grateful for this way to share insider information among us.

JA - I figured out the WAMM part, but jumped to the WATT because I had a little experience with it, but not the WAMM. Have you ever heard a setup?

beetle - I think their hall ambience retrieval is real rather than an artifact. Consider that a large portion of speakers have the 100Hz bump, but don't 'hear' the hall. I suspect that Wilson's heroically quiet enclosures along with their ultra high quality crossover components reduce low-level spuriae enough to allow perception of deep musical subtleties. 

Those kinds of parameters are some of what I'm addressing, while keeping an eye on classic Thiel value/cost constraints. Our collective outcomes are promising and encouraging to me.

mrfizziks - Rob at CSS is swamped. Best approach is thru his website. Keep trying.

Your tweeters are Polydax HD130x105D34E and your woofers are Credence 6C 16PP1.

JA - budgets are fascinating things. At Thiel, we never entertained trying to satisfy someone who had $six figures to spend. We tried for a performance plateau that fit ourselves and our projections of our imagined customers.

There are real improvements to be had. A great cap can cost $1000+ which could add $10K retail to a pair with normal manufacture and distribution markups. Wade through that labyrinth of focusing the next performance plateau and each product might go for something like triple the original sell price. I’m presently working with those equations.

As an example: among the many possible thought-problems - imagine a tricked-out CS5 with its same or slightly updated drivers but with 2 major upgrade foci.

1: Replace the bucket brigade delays in the two midrange drivers with physical displacement.

2: Upgrade the signal caps to CSA and resistors to MRA-12s.

The XO part count would drop to half and the sound quality would soar. That’s on my to-do list. The present CS5 owner already has the platform and upgrades don’t need the 2X retail markup.

Note that the physical baffle part of that CS5 upgrade might have happened except for the tumult caused by the rampant conversion from hi-fi to home theater in the 1990s. In an alternate universe, I imagine Thiel having concentrated on hi fi product refinement rather than the tremendous investments and demands of addressing home theater. Not only would the state of the art have been stretched, but the internal workings of the company would have been more manageable. That’s second-guessing reality. In fact, I don’t know if Thiel would have survived those market changes without jumping on the HT bandwagon. Anyhow, we are pushing the ball up the hill, a small step at a time.

audio1326 - we live with what rooms we have.

Complex geometry with openings into other rooms are challenging. My previous studio (shown on my virtual system photos) was such a room. L-shaped in the corner of a 30x48' second floor with bleed to both the 54x48' area below and attic space above. I don't have Finite Element Analysis in my toolbox, so I worked it throughout a year (or more) to excellent results. I can answer specific questions from that experience among others. You will be the best judge going forward - it would be daunting for anyone else to offer opinions on such a complex space.

Point of caution (information) regarding the GFA amps. The original 535, 545, 555 are Nelson Pass, high current, etc. etc. The next generation (series II) are not. They only share the case and name. They are well-considered and sometimes preferred, but note they are similar in name only. One reason I chose the original 555 is because Jim Williams / Audio Upgrades does such great upgrade(s) for very reasonable cost.

jafant - "There is something to be said about staying focused on One’s strong suit in Audio."

There was a turning point in the early 90s where we could have concentrated more on our home turf of high performance / accessibly priced stereo . . . or addressed the recording industry. Those options felt more right to me. Kathy managed the dealers and she was lobbied hard to address the emerging home theater market. We had introduced the SCS in 1990 which was a HT natural and Jim had always wanted to solve some of the inherent problems of subwoofers. Jim related strongly to bass as music's foundation.

He co-developed a class D bass amp with Birger Jorgensen at Vifa before Class D was much of a thing. That project consumed extensive time and resources that might have been spent on hi-fi projects. That custom amp powered Thiel’s first SW series subwoofers, which required him to troubleshoot and repair them due to lack of general knowledge in the field. The slippery slope.

Thiel’s further HT products worked OK in the marketplace, but they fundamentally changed how we approached our business. There was CEDIA and a host of new market considerations and different dealers. There was Chinese sourcing to meet market price expectations. There was simultaneous co-development of multiple products with differing specifications / tolerance requirements.

At the heart of the matter is that Thiel Audio existed primarily as a platform to enable Jim to design products; and more products provided more design challenges. So it worked, in that respect; but, on the other hand, there was less focused attention available for deep diving.

In hind-sight it’s interesting to me that various ’observers’ counseled us to ’stick to the knitting’. I agreed, and I withdrew. It comes down to what the lead man wants. Jim wanted to exercise his abilities producing innovative musical products that mattered. Diving deeper would have required a larger R&D team. And that was never in the cards.

Lenny - your LFE hookup may have caused the failure. It is also possible that you just blew a fuse. Look around and see what you find. 

We have not located those schematics after New Thiel moved to Nashville. As background, there are two functions, the amp, and all of Jim's boundary location correction circuitry. We can hope it's in the amp.

TT

abmeyer - the CS7 and 7.2 can be differentiated by their midrange driver. The original 7 has a flat polystyrene driver, whereas the 7.2 has a shallow anodized aluminum cone.

Tip - 2/3 of the weight is in the bottom half, so laying it down onto its back (onto pads so you can re-lift it) becomes a more

manageable 50# more or less. Just a thought.

dmac67 - regarding those 3.6s

As you probably know, Thiel products were engineered down to the gnat hair scale. The drivers were massaged by their crossover circuitry for physical roll-in and roll-out of 6dB/octave with the midrange controlled over 7 octaves. The drivers by the time we passed 1980 were custom designed with many innovations to reduce distortion and generally perform better than we could buy on the open market. The 3.6 midrange incorporated our patented double cone with a deep straight back cone and a shallow flared front cone for internal damping, higher stiffness to mass ratio, a top end approaching 20kHz, etc. etc. Any off the shelf driver will be at best a thin approximation. Any differences from the Thiele/Small parameters of the stock driver will miss the XO targets to some unknown degree.

Regarding his ’redesigned crossover’, we don’t know the level of change and/or skill involved matching the new driver to the woofer and tweeter.

Regarding the tweeter as ’better than aluminum’ Some folks prefer each; Jim chose the aluminum due to its greater consistency and higher frequency breakup mode which he could choose not to notch-filter (the high-Q filter can be heard.) Jim and Birger Jorgensen at Vifa together co-engineered that fiber dome in parallel with the aluminum one. It is a drop-in replacement (its low end rolls out same as the aluminum one.) The ’sound’ is somewhat mellower - some prefer it; but its ’sound’ is not as crisp and articulate as Jim’s products.

If I were looking at that Morado pair, I would gather some power-handling information on the new midrange. Our Vifa P11-MH01-04 midrange handled around 50 watts RMS (from memory), which is a high bar. First order x 7 octaves is a tall order.

Let us know how you navigate the waters.

duramax747 - good score!

My forensic observation reveals that crossover production for various models was taken to China/FST and that early FST crossovers had weaknesses and improved over time. However, it seems that FST XOs used MPT/polyester caps where Thiel had always used (superior) MPP/polypropylene.

Additionally, we know that Jim personally checked, measured and approved review samples. So, your pair is what a hot-rodder would call 'blue printed' - exactly as intended.

Hey prof - thanks for the Art Dudley reference. Time marches on. Even though we kept arm's length from our reviewers, I remember them all fondly. Critical review was a crucially important part of Thiel's very existence. Art Dudley, Wes Phillips and many others helped shape the industry and the companies in it via their thoughtful, informed appreciation and criticism.

Thiel's cabinets were a monumental undertaking - one that could not be justified except that we wanted to produce functional objects worthy of our intended customers' livng rooms. (And I got to travel extensively establishing our veneer supply network.) No complaints.

@vair68robert - what a mess. It sounds like Nashville Thiel may not have had final product testing, because your value drifts would have failed your finished speakers.

We took quality assurance very seriously. There are many approaches - our methods were always changing - with lots of thought and care. The 1980s saw the advent of ISO-9000 where manufacturers processes were certified rather than their final testing. We bought from suppliers with essentially the approaches that became codified in ISO9001. In that environment, incoming testing is not necessary. Every part can be assumed to be correct. The flip side is that if any single part flunks at any point in process, then the entire batch is quarantined, the problem is investigated, solved and re-documented. We co-developed our implementation with Vifa (and their group of co-producers) to good success.

Please permit me a story, an important, meaningful story that helped form Thiel Audio's identity. Consider the CS3 woofer. It was our first real foray into partnered product development. That woofer was a big leap ahead, especially with the copper shorting rings and field-focusing top plate geometry. It also pushed the envelope of what a woofer was expected to do: perform well and consistently to 10kHz. That long, even high-end roll-off depended on tight control of many factors including viscosity of the bobbin / cone glueline. We spent more than two years working on that woofer with Vifa. Perhaps because there had been so many iterations (?10+?), and perhaps because we had developed so much mutual respect, perhaps a dash of time / annual cycle pressure . . . the first production run of woofers had an anomaly that missed all of us. Its upper range, perhaps 4-8kHz had too much energy. 'Normally' a woofer's response in that range is irrelevant; but with Thiel's first order rollout, it mattered. Good lessons were learned that paid strong dividends over the following decades.

The root cause turned out to be a change of adhesive caused by the Danish government outlawing epoxies for health concerns. The replacement 'equivalent' glue acted differently at operating temperature than the prior well-understood epoxy. The new glue eventually also failed in the field for our first (possibly only) recall. That 'problem' clarified our MO for customer service. We replaced every one of the woofers at no charge and our dealers magnanimously swapped them out at no charge to us! To make lemonade from the lemons, we provided a record (yes, vinyl) to each customer as consolation for their trouble. The record was Michael Hedges 'Aerial Boundaries', which was both musically and technically masterful. That situation became legendary and performative - defining our image both internally and in the marketplace. Kathy Gornik gets much of that credit. 

When I first heard of your wildly out of spec components, I was dumbfounded. As the particulars came to light it made some sense. New Thiel was not the same company with the same values of product integrity and customer satisfaction. I wonder how many more 2.7s are out there from New Thiel.

In the eventual realm, I am working with ScanSpeak (Vifa) on a 'new' tweeter in that motor frame. My target diaphragm has the potential of greater specific stiffness than aluminum without the ringing and without the high expense of carbon. Work in progress. 

For now, that soft dome (I don't know how soft) was engineered to fit that motor and alleged to be plug-in compatible. Jon, see if Rob can make a frequency sweep for you on both driver types so we can compare and contrast.

It's no accident that the 1.5 hung out on Stereophile's Class A (limited bass) ranking for 5 or so years. I used it for the past 7 years at our small village live performance venue to startlingly good results and reviews. I used it with a Thiel SS1 and Thiel's Passive OX set for the 1.6 (very similar bass parameters.) I paired it with a Prima Luna, which is a nice sonic match. Plus the 1.5 is in my upgrade sites. It's quite a little honey.

jonandfamily - we look forward to your report. Here's some additional back-story about that soft dome.

We developed that UltraTweeter for the CS5 in 1988. It had all of Jim's advanced techniques, and exceeded our expectations. The CS5 had woven lower Kevlar drivers, and if an appropriate soft upper midrange driver had been found, we would not have gone to aluminum. However, the MB Quart 2" aluminum dome did the job better than anything else, and we didn't have time to develop our own driver for that use. Market pressure. The aluminum tweeter had more predictable response particulars and the metal domes dissipated heat better. That's where we went.

When trickling that tweeter down to the CS2.2 we reconsidered the fiber dome to mate with the pulp midrange. Jim went with the CS5 tweeter and gradually migrated to all aluminum diaphragms rather than all Kevlar / fiber / etc.

So, I consider that soft dome tweeter as native - it was co-developed with the hard dome and should be very good, with possibly some advantages. The only one I've ever heard was in comparative evaluation during 2.2 development around 1989. I'll be getting a pair for comparative measurements and listening tests.

As a general caution, heat dissipation is a big deal. The soft dome don't be a heat dissipator like the aluminum dome is. Don't over-drive it.

Keep us posted.

danomar and gman - I don't have direct knowledge of the PCS, it coming after my time at Thiel plus never having actually seen one. But I can throw some scattershot around it in case something may help.

The Y2K PCS was a trickledown product from the 1998 CS2.3, which was Jim's first generation passive-coupled mid-tweeter. That 2.3 driver also graced the 1998 MCS. The PCS - Personal Coherent Source - was billed as a desktop system. I didn't know about a stand. I've heard that an upgraded PCS was in the background haze, which would have utilized a third generation passive coax being co-developed for the 7.3 along with a 'normal' 2-way active coax in case the passive didn't cut the mustard for the pinnacle 7.3. Such a gen3 coax would also have gone into the CS2.5. None of that happened.

An interior woofer brace with drawbolt was designed for the CS7 / 7.2. I don't know about similarities / differences to the PCS.

Regarding comparisons - the most direct family resemblance would be with the 2.3. The 2004 CS2.4 coax was improved and the XO refined. I've never heard either the 2.3 or PCS nor seen schematics of the PCS. My speculation is that it would likely have true first-order crossover slopes because the driving problem necessitating the hybrid first-second XO is overtaxing the bottom end of the tweeter. The CS2.3 / PCS driver has that mechanically coupled midrange driver to carry the bass end of that compound voice coil. (Clever, no?)

To round out my comments, The SCS4 / PowerPoint (etc.) products all returned to true first order, having the advantage of Jim's further tricks for increased tweeter power-handling. Just for grins I'll add that my SCS4P (professional) in development places a thru-wall heatsink on both the woofer and tweeter for thermal stability when using the speaker hard for hours on end. 

Cheers, Tom

JA - thanks for the tip. I won't be pursuing it. Bill Thalmann tuned up the old Thiel SCD-1 which had been previously upgraded. It's quite a piece of gear. The SACD head was not salvageable, but I have very little SACD hard media. I'm beginning my investigation into software storage and playback of high resolution media, while avoiding 'normal' computer circuitry. I auditioned iFi's Zen streamer - pretty impressive for the price.

Hello Prof, good to see you.

The Original Series was indeed sequential, 01, 02, 03, 04 with advances marked in small letters. The 01(01a, 01b) was Thiel's first marketed product. There are some still out there and Rob says they have a loyal following. As a seminal product, the 01 illustrates Jim's fundamental values. They were small, inexpensive, very sensitive, flat with unbelievably effective bass. Deficiencies included refinement, delicacy, coherence - the audiophile virtues.

We sold about 1000 pairs between 1975 and 1983. The breakout success of the CS3 in 1983 (actually a 4th generation 03)  changed our dealer profile and overwhelmed our production capacity. The 01b was dropped for those reasons - it was still selling strongly.

The 01 was a 10" x 1.5" two way with equalized bass and 3rd order slopes. Its veneered 0.85 cubic foot enclosure had a foam grille, and the woofer a foam surround, both common at the time and prone to ultraviolet deterioration. The on-axis response was flat (+/- 1.5db) from 30 to 18kHz in an enclosure only 13% larger than the 02. Its woofer was Eminence's first custom driver (model 10101).

Within easy access 70 miles down the road, they supported our customization of their bomb-proof musical instrument workhorse which they built under their own name and for Peavey at a clip of thousands per day! The tweeter was by Long Engineering and sported a mylar dome mounted in a phenolic ring for minimal diffraction.

The equalizer was the 01's claim to fame. Unlike Bose, which pushed small drivers way past their range of efficiency, Jim's EQ boosted 10dB (as the 03 - CS3.5) to stay below the power required for midrange peaks. The equalizer was optional, accounting for $75 of the $350 retail price.

Being our inaugural start-up product, our quantities were too small for cost-effective purchase of anything except drivers which we picked up from the factories in our 1958 Edsel station wagon. We wound our coils on 'George', our shop-made spinner and etched and drilled the EQ printed circuit boards. Of course we made every aspect of our cabinets including our veneer faces - all in the original farmhouse. 

Our first CES was spring 1977 where we showed the 01 and 02 to an enthusiastic market including the German distributor of DCM Time Windows. Strong German reviews and market brought interest from high quality east-coast USA dealers. And the rest is history.

We were naive and unexposed to any audiophile considerations or market. The 01 was designed as an all-purpose speaker that did well what Jim considered the necessary virtues. That Eminence 10101 woofer persisted through the 03, 03a, 03b sequence and is still available through Rob. I bet you'd like them if you can find a pair. Thanks for asking.

fitter468 - I've never actually heard an Ayre piece of gear! But they get rave reviews and user comments.

Which BJs XLRS did you get for what position? I've spent the past month and a half on cable evaluations and would appreciate knowing what you have and where it is in the system.

hifi28 - what seems to work best is to ask a specific question for someone who is knowledgeable to answer. Matters of repair and maintenance are best addressed to Rob Gillum at Coherent Source Service.

The CS2 is of historic interest to me, being the first of its series and the highest-count sales of any Thiel product. Its woofer was a Vifa/Seas unit, probably repairable by Rob.