Thiel Owners

Wow, even the Cardas binding posts won’t set you back that much! I would buy them again - recommended.

Rob will only have the CC SAs which were CC's best at the time, for the series feeds.

Appears that @vair68robert has CS2.7, not CS2.4, so Rob's Clarity SAs are almost certainly incorrect capacitance

Yes, really tough to get a solid mechanical connection with the Cardas posts but, if you’re using Litz wire, the tinned ends necessitate a solder interface either way. I couldn’t compare other than the OEM wire and binding posts to the full Cardas but I can tell you the difference put a smile on my face. 

I’m looking forward to reading about your choices, progress and results.  

@vair68robert

I was doing my yoga this morning and found my eyes about a foot away from the outriggers on my 2.4s. Do you have outriggers? If any part of the outriggers is not tight, the Ayre tone glide can result in a rattle over portions of the sweep!

@vair68robert 

top nut on my outriggers occasionally gets loose but it’s possible one of your spikes is doing the same, especially if the weight is mostly on three spikes rather than all four as could happen with an uneven floor. Worth a quick check.

I’m very happy with the Cardas binding posts even if it’s tough to get a good mechanical connection. Tom had me strive for that standard wherever possible. A couple of other places I had trouble: 1) connections on XO boards where multiple coils, resistors, and caps come together (especially joining the Litz wire to the bundle), and 2) connection to driver posts, all the more so with the Litz (but even the OEM solid core was pretty much a solder connection there). 

@andy2 

The only remaining OEM are the drivers and cabinets. If I had Lexington rather than FST boards, I would have salvaged a few coils and, maybe, bypass caps.

@bighempin 

I would desolder. Note that Thiel used unleaded solder which has a higher melting point. Your iron/gun should be at least 140 watts. Speaker should be on its back so that the driver can’t fall out and prevents solder from dripping onto the driver. There is enough slack in the wire to access the terminal but not much more. A second pair of hands is very helpful. I did this solo on my 2.4s by setting the driver on a shim above its hole which leaves plenty of room to get at the terminal.

It’s not that difficult but you can gain confidence by practicing soldering and unsoldering bare wire.

I suggest adding the second pair of posts when modding the XO. Beetle may chime in - he looked in to that with his 2.4 upgrade.

My experience with bi-wireable Vandersteens led me to explore this option with my CS2.4s and Tom Thiel encouraged me to try it. Also, the OEM binding posts have brass whereas the Cardas CPBP are rhodium over silver. The OEM hook up wire in my “SE” version was sourced from FST and, based on the parts quality on the FST-sourced boards, I suspected was sub-optimal (those of you with Lexington boards have higher quality parts). I replaced the posts and wire with Cardas and added separate binding posts for the low and high pass boards.

I sonically compared single runs of speaker cable from the amp using Cardas copper jumper plates to double runs of identical cables. The biwire configuration consistently sounded more relaxed and liquid. Some songs also sounded a bit more dimensional and clear via biwire. The latter characters are especially subtle but I heard the effect on more than one song.

This is not an easy or inexpensive upgrade but certainly worth it for me. The low-hanging fruit, however, is to upgrade the caps and resistors, maybe the coils if you have boards made by FST (it appears that Thiel Audio started using FST shortly after Jim Thiel’s passing). So, start with the XO parts, especially if you’re on a budget. But if you’re insane (like me) and want to squeeze every iota of performance out of your Thiels then dual binding posts should be on the agenda!

I'm looking for a pair of 3.5s as well as 1.5s.

@tomthiel

Hifishark (https://www.hifishark.com/) might be the "best" one-stop online source to check. They appear to be an amalgam of eBay, US Audio Mart, TMR, and other worldwide sources. You can filter results by country. Here are a couple of relevant hits:

https://www.usaudiomart.com/details/649548973-thiel-cs-15-pair/https://www.ebay.com/itm/143396134741

Probably several hundred pairs of CS1.6 out there but, I would guess, not many 1.7. I had an early pair of 1.6 (2002) and I think the serial number was in the 200s. I could ask the man I sold them to. 

I posted a long version answer perhaps a year ago on this thread.

Almost *exactly* one year ago (link)

@cascadesphil

Have you tried the rack on a side wall (maybe with amp on the floor between the speakers) or pulling the speakers more into the room, away from the rack?
Could make a difference in imaging.

Soundstage (NRC lab) measures speakers at 2 m - 79”. Their CS2.4 frequency response looks *much* better than that measured by Stereophile. In fact, the 2.4 has the flattest measured response of any in their database (actually, there is a Magico model that was similarly flat). There was at least one test by JA of a Thiel (or was it a Vandy?) wherein he acknowledged the problems around measuring the speaker at only 50”. But he ignored that issue thereafter. 

@jafant 

Curious question! Firmware H, 2017.
why do you ask?

@jafant 

That’s a question for Ayre. My guess is the firmware makes zero difference to the sonics, probably more to do with the display and remote operation. Gary at Ayre will be able to answer your query.

@sdecker
FWIW, the CS2.4SE did away with the 1 uF bypasses on the coax feeds, ie, full 14 and 28 uF in single caps rather than 13+1 and 27+1. I am aware of a third-hand report that a 1 uF bypass can have a deleterious effect on SQ (supposedly better to go either higher or lower values for bypass caps). I don't have any information on whether this is related to Jim Thiel's decision to do away with the bypasses. But, clearly, he preferred the sound of 14 and 28 uF Clarity SAs over the original 13+1 and 27+1. CSS apparently has 14 and 28 uF SAs available although the ESA are supposed to be a step up from SA.

My modded boards have 14 and 28 uF values in single Clarity CSAs, the 14 uF bypassed with a 0.1 uF Multicap RTX and the 28 uF bypassed with a 0.22 Multicap RTX. The bypasses improved the "jump factor" and added, maybe, a smidge of resolution.

Buying these parts from a retailer, I would probably get 10+3.9 uF (and a 0.1 bypass) and 18+10 in 630 volt Clarity. Maybe CSA for the bigger value and CMR for the smaller value (for maximum SQ) or all CSA (to save money and space). Jantzen Alumen is said to mix well with Clarity, also.

Regardless, I highly recommend replacing the 2.4 resistors with Mills MRA-12s. You can replace the sandcasts on all boards for about $100. Money well spent. Sonic Craft carries all the correct 2.4 values although you have to mix and match from among the older Mills and newer Vishay-Mills.

Anyhow, thanks for posting about your ESA upgrade a few years back. That gave me courage to try my own mods - and I'm super happy with how it turned out.

@sdecker
Nope, I didn’t PM you. I’ve owned my 2.4s for only about two years. Tom Thiel coached me through most of my mods. I replaced every part from the binding posts to the driver hookup wire. Sounds glorious. 

upgrading from a stock CS 2.4 to the CS 2.4SE model.

sdecker doesn’t really have SE-equivalent. The advertised changes were swapping out 13 uF polypropylene + 1 uF polystyrene for 14 uf Clarity SA and 27 uF polypropylene + 1 uF polystyrene for 28 uf Clarity SA. The unadvertised changes were sourcing the boards from FST which reduced the parts quality of the other caps as well as coils (not sure about the resistors). FST were also printed circuit boards instead of point-to-point. It’s also *possible* that the hookup wire on the CS2.4SE was not as good as the original CS2.4.

sdecker kept the 1 uF polystyrenes and used Clarity ESA (instead of SA) to replace the 13 and 27 uF coax feed caps. Probably, sdecker has something *better* than the SE-version although I’m curious to know the sonic consequences of the 1 uF bypasses relative to full capacitance from single caps (ie, 14 and 28 uF). Perhaps Tom Thiel’s trials will include comparisons with and without the 1 uF bypasses.

Thanks for your post, Tom.
Interesting that Rob found the ESAs to sound worse. My experience was that nearly every change improved the sonics. Perhaps because I had a good coach ;^)

I am considering those electronic upgrades as end of project decisions. My focus is on re-bracing and re-baffling which is progressing well.

As you know, I'm a big fan of upgrading the passive parts. And I think a lot of the improvement I hear can be had for less money. That said, I'm looking forward to trying your baffle treatment.

the 630V caps and far-more-expensive Mundorfs et al were simply too big for the space on the XO board

The higher voltage versions, with thicker film, are reported to sound better. I didn't compare different voltages on the coax feeds but did directly compare 160 and 250V versions on the woofer shunts. Some say that shunt cap quality makes little, if any, difference in SQ but, to my ears, the higher voltage version sounded a bit more relaxed and with more bass heft and impact. A subtle difference but one I am willing to pay for.

In my case, the larger caps were not an issue because I started with completely new boards/layout. So, no need to fit a new cap into an existing space.

@twoch I've done tie-dye once but fail to see how that is relevant to this thread.

@tomthiel @sdecker  OK, you now have me very curious to try a baffle treatment. But my DAC is away getting an upgrade (I'm currently rediscovering my LPs), so I'll have to let that settle in before I can do a fair test of changes to the baffle.

@sdecker 

My analog section is decent: Kenwood 500 (direct drive) mounted with BenzMicro ACE, tonearm wiring replaced with Cardas feeding an Ayre P-5 . . . But I almost never use it. Until this week, I hadn’t used it at all since probably last winter. 24-bit files sound superb through my DAC, all the textures of vinyl combined with the dead silent digital background. Best of both worlds. 
I am planning to try Tom Thiel’s solution but probably not until after my “new” DAC settles in. Thanks for your observations about the baffle morphology on the 2.3 v 2.4. Very interesting! I’m motivated to experiment.

Pretty much all speaker designs are a balance of trade-offs. If there was a perfect formula then all speakers would implement that design. Sealed bass enclosures are generally considered to sound the most accurate and that is certainly what I hear. But sealed boxes are limited in extension and ability to play loudly. IMO, Jim Thiel’s passive radiators are a bit short of SOTA sealed boxes in terms of quality (think CS5, for example) but nevertheless are a highly satisfying solution, especially at their price point.

@last_lemming Do you have a way to measure SPLs, confirm your subjective impression? When I rebuilt my CS2.4 crossovers, I was able to compare the before and after frequency response using the SafeNoise app on my iPad and a tone glide.

BTW, not that I think this is your issue but have your tried a set-up with the speakers on the shorter wall? Either way, I suggest taking a look at the Cardas website for some placement ideas to try after you figure out your problem.

@last_lemming My suggestion to use an iPad was not good in your case. The iPad mic is good from about 200-3k cycles but really drops off below 200, coinciding with your problem area. A better mic is called for if you want to measure in the region you think is amiss. It worked fine in my case because I only needed to detect relative, not absolute, differences.

@cascadesphil Nice, thanks for the information. I don't necessarily need that now but might be fun to play with at the price!

@vair68robert 
Thanks for the update regarding wool batting.
My SEs had rectangular fiberglass batts (no paper backing) stuffed into the passive radiator chamber. My new boards have much more volume so I peeled off some of the batt volume in attempt to end up with similar residual fill.  

@jafant I made no comparison. Just assumed that the previous volume was “optimized”, a dubious assumption given that the volume can easily be modified by compression, ie, there is no one OEM volume that can be discerned. I merely made it as close as possible to how it was when I first removed the radiator to view to boards.

@andy2 Thiel Audio found a formula that sounded good to them and they pushed that vision as far as they could. No doubt they considered and listened to alternatives but first order filters became their cornerstone. Tip of the hat to them for finding something good and sticking to it. You can decide whether it sounds best to you. All designs have trade-offs. My ears quite like Thiel’s solution.


I have a lot of respect for that. I mean, look at Magico. Their models are all over the map in terms of design philosophy. I think their main skill is throwing gobs of money into the company, trying whatever design element they can think of (well, mostly borrowed from others). IMO, they’re mere hype from TAS.

From my perspective, Thiel Audio had two main problems: 1) Jim Thiel didn’t train an engineer to carry his vision forward; and 2) New Thiel owners had almost no experience in high end audio.
Do you really think Thiel Audio folded because of adherence to their design principles? I sure don’t. Quite the opposite.

I am not arguing about the technical merits of time-phase coherent. I am only arguing about Thiel business model as if it is financially viable. I would not criticize PSB or B&W as they are able to find a way to be viable even if using a different design strategy. To me it seems like Thiel had put themselves a bit into a corner with such a singular mindset - that is first order time-phase coherent or all else which may be correct technically, but financially, it did not have a way out.

Or perhaps Thiel had been successful *because* of the "singular mindset". Thiel speakers were (and are) unique, well-executed, and great sounding. And when Jim Thiel passed without any other engineer among the small business’ staff, subsequently bought by people with no experience in the industry (who steered the company completely away from Jim Thiel’s design principles), it was predictable that Thiel Audio would fail. We’re talking about *Thiel* Audio.

B&W is part of a much larger company. Can you name any of their engineers? I know Laurence Dickie designed the Nautilus but I can’t name any current engineer at B&W. Will Vivid stand after Dickie passes? What do you think will happen when Richard Vandersteen passes? Jeff Joseph? It looks like Wilson Audio is making a go of it with Dave Wilson’s son as the main engineer. ARC successfully transitioned from Bill Johnson and Ayre is making it work with Ariel Brown. But it’s not great for a small audio company to lose its founder and primary designer.

I suppose Mark Mason could have been more successfully plugged into Thiel Audio if Jim Thiel had designed me-too speakers. But, then, would Thiel Audio ever have been successful at all? Would we be on this thread talking about our Thiel speakers? Loving them? Modifying them to get even better performance?

@mr_bill 
The CS2.7 was developed after Jim T passed. It was a collaboration in house but with external engineering for the crossover. Search “CS 2.7” written by “tomthiel” to get the history. It has a first order network.

If you aren’t afraid of a soldering gun, get a pair of CS2.4 and either wait for Tom Thiel’s upgrade kit or pour through this thread for ideas of how to make a great speaker into a superb one.

I used an Ayre AX-7 to drive my CS1.6s and now using an AX-5 with my CS2.4s. Very happy with both pairings. Measurements can give a hint but listening is only way to know.

@8th-note Glad to read that your CS6s are working again. We are fortunate to have Rob/CSS! Thiel Audio went into bankruptcy about one month after I bought my 2.4s. Rob opened not long after and I asked him about replacement drivers. He said none are available for that model but that he can rebuild them. When I did my upgrade, I found his name on two of my drivers (one was dated well after the speakers were built, so he must have serviced that one).

if playing 100dB, mid asked to put out 250Hz percussives at 88dB (round 'electrical' numbers, not taking acoustic roll-off contributions to the end-result 6dB/octave slopes.)). Ask Rob how many 2.4 coaxes I've gone through, and not due to lack of clean watts and amps!

Interesting. If ability to play >100 dB was an important factor to me I would probably look at something like the JBL K2 S9800. Modestly efficient Thiels with their low impedance seem a, um, sub-optimal choice for those who like to listen at very high SPLs. And to still have your CS2.4s after burning multiple coaxes? You must *really* like their sound!

Indeed, this has been a fantastic thread. Thanks for kicking it off, @jafant

We are very fortunate to have Tom Thiel participating. When I bought my 2.4SEs, I had an inkling to upgrade the crossovers. But I had nowhere near in mind to take it as far as Tom coached me. I would have been nibbling around the edges and not making near the difference compared to the informed decisions from Tom's mind.

I remain *super* happy with my upgrade and haven't budged from my position that these are my last speakers. In fact, with my DAC upgraded I doubt I'll make any further changes to my system other than to try the baffle treatments Tom is experimenting with. 

@jafant 

With my particular combination of gear and assuming a typical CD, clipping will onset at an indicated “40” on the AX-5. I usually listen at about 30 for rock, maybe 35 if I’m extra rowdy. But most of my listening is 20-28 indicated, even less when I’m sharing the house.

The MX-R is far more powerful . . . and expensive. 

@jafant 
unsound is steering you away from Ayre because it doesn’t vomit 2 ohm watts like a moon rocket on the Stereophile test bench.

The CS2.4 presents a ~3 ohm load from 400-20k cycles. Two speakers need ~ 50 Watts to produce 90 dB at the listening position 9’ away (and that’s with a steady test signal, not music). Your DX-5 gain is a bit different than my QB-9 but not hugely so. In my system, clipping onsets around 97 dB with a typical CD. But I rarely ask for peaks north of 90 dB and most of my listening is 75-85 dB depending on genre and my mood. I’m probably only asking for 2-25 W for 90+% of my listening. The AX-5 is rated at 250 W into a 4 ohm load, Ayre doesn’t give a 2 ohm rating. That said, if you *do* listen at very high SPLs, the AX-5 probably won’t get you there. 
But before you believe me or unsound, listen for yourself, not via Stereophile measurements. 

@unsound

Um, one of us has the actual *experience* of listening to the CS2.4 driven by an Ayre AX-5.

Stereophile’s measurements help us to discern that the AX-5 is not a good match for the CS5 or Magnepan 20 and shouldn’t be mated with the CS2.4 *if* the listener demands 100 dB peaks. You might note that the CS2.4 never drops to 2 ohms (it was measured by 3 reviewers, Stereophile seems to have measured the lowest value). As for

But into 2 Ohms (green), not only is the THD higher, but the level was a *little* unstable at the lower frequencies

Not only does the CS2.4 never drop to 2 ohms but the AX-5 behavior JA is describing was observed only in the bass where the CS2.4 impedance curve greatly rises. Also, those were steady state test signals, not music!

Regardless, those measurements tell you *nothing* about how the amp sounds. You need ears for that part!

From the Thiel owner’s manual for my product:

Keep in mind that sound quality is usually much more important than sound quantity. There can be large differences in the sonic performance of two amplifiers of equal power, and this is more important than large differences in power. Most everyone will be happier with a 100 watt amplifier of high sonic quality than a 200 watt amplifier of mediocre sonic quality. For this reason, we feel there is no substitute for listening in making your amplifier decision.

Apparently, I’m in the "most everyone" camp. Good day, sir.

I think that while measurements might not tell all about how an amp sounds, some aspects of sound are quite predictable from measurements.

Perhaps in a vague way. But you either forget that I have actually *heard* this combination of gear or you seek to invalidate my experience. Piss off.

@prof

My thanks, also, for your comments re: CS2.7 v. Joseph Perspective. If I were buying new in that price category, the Perspective would be on my short list. And your experience adds weight to my opinion that you need to reach even higher up the price scale to significantly better a Thiel. My guess is that I would have to spend north of $30K to notably better the sound I'm getting with my modded 2.4s.

I’ve now twice mated an Ayre integrated with a Thiel. Been super happy both times. 
Interesting that being able to discern how an amp sounds from measurements is “objective”,  lol. 

@tomic601 
Did Stereophile review the Kento already? I didn’t renew my subscription last year. Not happy with JA’s promotion, even hype, of MQA. Jim Austin taking the reins was a deal breaker for me. 
Meanwhile, JA’s protocols really put Thiels and Vandersteens at a disadvantage. Comparing Stereophile’s and Soundstage’s measurements of the CS2.4, for example, it’s almost like they were not measuring the same speaker (Soundstage contracts with NRC which has a real anechoic chamber and they can measure at 2 m). The 2.4 has what appears to be the flattest frequency response in the Soundstage database!

Yes, in the 1990s there were a couple of Thiel reviews wherein JA acknowledged the "suckout" was probably more a result of his techniques than actual performance. From the CS2.3 measurements section:

The 50" mike distance I use is a compromise between the need for correct drive-unit integration and the opposed need for midrange resolution in the resulting graph. But it is possible that the lack of presence-region energy in fig.3 is actually due to the 50" mike distance and is not real, in that it disappears at the farther distances at which a speaker like the Thiel will be listened to.

There was at least one other Thiel review wherein he said something similar. But I think he didn't write a word about this in any Thiel review after ca. 2000. I am thankful for how JA built Stereophile up, particularly, becoming the only US audio mag with subjective reviews *and* measurements (even if the measurements have obvious faults, at least they're applied equally to all products). But present day Stereophile is not for me.

@sdecker nice post

@tomic601 are you referring to the Tekton? I’m very skeptical looking at the design and driver selections but they actually measured quite well on JAs test bench, FWIW. I’m curious to hear them. Not that not I’m inclined to let my 2.4s go!

@sdecker 

Indeed, thats, um, a bit over the top description of what is mostly cosmetic differences. All the more given that the SE boards were sourced from FST (at least mine was). Those cabinets do appear to be in better condition that mine.

Cables, real or placebo? IDK, but I do know that I’ve repeatedly experienced better sound, especially with speaker cables and ICs, not so much with power cords or USB cable. Oh, and with the hookup wire in my CS2.4 mod. I’m a fan of Cardas products but haven’t heard the latest “Clear” line. 
Sounds better to my ears and it bothers me not whether you agree of disagree. 

@jafant 

I ran Neutral Reference when I had the CS1.6. Now using Golden Reference. All purchased used - those things are $$$

My AX-5 is “J” production series. To my knowledge, any later production has no sonic improvements. 

I took delivery of an AYRE AX-5 integrated amp in black casing

Oh, fun! I suspect you'll enjoy your combination of gear.

@jafant I just had 16 ga speaker wire when my system was mid-fi (B&W, NAD, Rotel, Tecnics). Moving up to the high end, I found a great deal on used Cardas Neutral Reference. Ayre really likes Cardas, so that greatly influenced my decision. Now running Golden Reference (also previously owned). I’ve heard many other brands at dealers and RMAF but only Cardas in my own system. If Cardas is too rich for your wallet, maybe consider AQ?