Thiel Owners

I'm the one who won the eBay auction for the SI-1 integrator. Can't wait for it to arrive!

Prof.: Thiel crossovers can be used with non-Thiel subs. Setup may be somewhat less automatic than with Smartsubs, but no more and often less complicated than using non-Thiel crossovers. For example, the SI-1 makes it easy for non-experts to integrate any main speakers with Thiel Smartsubs, but it has essentially the same controls as the CR-1. Real differences: The SI-1 is fully balanced; the CR-1 is not (only its inputs/outputs are balanced). The SI-1 uses first-order crossovers; the CR-1 uses 4th-order crossovers.

It's also possible to use Thiel's passive PXO5 crossovers with non-Thiel subs. I'm temporarily using the PXO5 that previously fed my Thiel SW-1 to mate my  F112s in mono/augment mode with my CS3.7s. The JLA's integrate very well even though I haven't bothered to match levels for this temporary configuration. I did, however, use a rough calculation and the JLA phase controls to phase align the system. This yielded impressive improvements in soundstage and slam and, overall, the best sound ever from my system.

A simple and cost-effective solution for Prof's system, running the subs mono/augment, might be: (1) set up the JLA E110s for master/slave operation; (2) switch off the internal crossover on the master JLA; (3) install a PXO5 configured for your CS2.7s or 3.7s; (3) adjust the level on the master sub to taste (or use measurements). (4) use the polarity and phase controls on the master to phase align your system.

In this setup, the PXO5s handle everything other than level matching as intended. You'd need an adaptor to connect the XLR line-level PXO5 outputs to the RCA line-level inputs on your master sub. You can skip step 4, but you'd be missing a huge improvement.

Note: The 4th-order crossovers in the JLAs are not compatible with the 1st-order crossovers in Thiel speakers. In augment mode, the PXO5's should provide much better integration than anything you can achieve with the JLA internal crossovers.

Check out white papers on the sounddoctor website for tons of expert information on how to set up and integrate subs.

If you're looking for an active crossover, consider Marchand Electronics as a less expensive, but still audiophile grade, alternative to the JLA CR-1. The XM9 and XM44 series are very reasonably priced, but you have to swap plug-in modules to change the crossover frequency. The XM44 offers a fully balanced option that is not available with the CR-1. The XM66 series is functionally very similar to the CR-1 (no module swapping). If you can tolerate the pro audio appearance, you'll be very happy with the cost savings. I was leaning toward purchasing a fully balanced XM44 until the SI-1 came up for sale on eBay.

Thanks for your welcome, jafant, and for your reply, Prof.

I have some deadlines looming and a family member going through some serious medical issues, so I’ll have to limit my time here. I’ll try to respond to your queries a bit at a time.

My system: AMG Viella turntable with Benz Micro LPS-MR cartridge on an HRS M3X isolation platform; Sonore ultraRendu powered by Uptone Audio LPS-1; Classé CP-800 with on-board phono stage, 192 kHz DAC, and ethernet; Classé CT-2300 stereo amp; Thiel CS3.7s, Thiel SW-1 (retired), 2X JL Audio Fathom f112s, now with solo Thiel PXO5 crossover, but soon to be replaced by Thiel SI-1 Integrator; Stealth cables: hyperphono, Indra v10, Dream; GIK room treatments: early reflection points, corner soffit bass traps, array of 50 Hz tuned bass traps on rear wall with “alpha” blonde wood diffusion grills; Isolation and vibration control: various Stillpoints, BDR shelves and cones; Running Springs Audio Dmitri power conditioner with HZ cord. (Most of my gear was purchased used at steep discounts. Still, it adds up.)

I did run the SW-1 in a horizontal position because my wife didn’t approve vertical. This was challenging, because I had to invent an isolation and vibration control system since I have a suspended wood floor. It worked pretty well, but not perfectly. I’m going to try Auralex Subdude IIs under the JLAs, maybe a simpler and more effective solution? In the end, changes to our living room furniture ruled out the horizontal SW-1 for reasons probably similar to those Prof mentioned.

I worked with (free) acoustic consultants at GIK to develop my acoustic treatments. Most of it is standard treatment of early reflection points and soffit traps in the corners. More notable was severe SBIR cancellation off the rear wall. This was killing bass at the listening position. The tuned bass trap array was the only way to solve this, and as promised by GIK, it works. I made numerous measurements with Room Eq Wizard (REW) to understand my room's acoustics and to inform the GIK consultants.

Room treatments are essential to get the best out of your system (but I understand this isn’t always possible — I’d like to also treat early reflections off our ceiling, but that’s a bridge too far). In particular, low bass will muddy up your high end if you don’t control low-end reverberation times. Adding a sub can make this problem worse. This might be why some people find that adding a sub makes their system sound rolled off on top.

Prof, check out these links related to subs in general and an interesting case study of getting phase and time-coherent integration with JLA subs and a Marchand crossover.

http://www.soundoctor.com/whitepapers/subs.htm
http://www.soundoctor.com/studio/

Unfortunately, in my case, I'd have to put my subs right next to me on the sofa to replicate this solution in my room (heh).  I don't want to push Marchand too hard, in part because I've never heard or touched one. But based on what I've read, they should be just as compatible with JLA subs as the CR-1. Compare the controls on the XM66 http://www.marchandelec.com/xm66.html to the CR-1, I think they're very similar.

Here come some long-delayed replies to several points raised by Prof on 12/9 in his response to my original post to this thread. First, a correction:

I wrote on 12/9: 

The [Thiel] SI-1 uses first-order crossovers; the [JL Audio] CR-1 uses 4th-order crossovers.

This is what I was told during my research before bidding on the SI-1, but it’s not true, as seen on p 10 of this document: https://www.manualslib.com/manual/328036/Thiel-Smartsub-Ss2.html. (For more complete information, read the whole Integration section, starting on p 8.)

Thiel argues that standard LP and HP filters can not produce linear frequency response in the net acoustic output, even with some extreme tweaks. See pp 8-9 for examples in which optimal independent selections of standard LP and HP filters combined with optimized level and phase adjustments fail to deliver linear acoustic output. The standard LP and HP filters used in most active crossovers, on the other hand, are designed to deliver matched (mirror-image) voltage profiles, not matched acoustic output. For example, the JL Audio CR-1 uses 4th-order Linkwitz-Reilly filters that guarantee matched voltage profiles.

Thiel’s goal for the SI-1 in crossover mode (augment mode is another story) was to match the total response of the main speakers to the total response produced by a generic sealed sub driven by an ideal (according to Jim Thiel) 4th-order LP filter with Q = 0.5 damping. Evidently, “total response” refers to the net acoustic output determined by the voltage profiles of the HP and LP crossover filters in combination with the distinct acoustic properties of the sub(s) and main speakers. To accomplish this goal, he uses a nonstandard, highly customized HP voltage profile to make the total response of the main speakers match the shape of the target sub response (see Fig. 11 on p 10 in which the red curve shows the customized HP voltage profile for reflex mains that are typical of the Thiel line). Once this is accomplished, “straightforward” phase alignment and level adjustments should complete the integration (this is IMO; it’s not stated explicitly by Thiel).

BTW, the first-order crossovers in Thiel’s main speaker line are designed for first-order acoustic response, not first-order voltage profiles. From Thiel sales literature for the CS3.7: “The Crossover is a true first order *acoustic type* that provides the utmost in spatial and depth imaging performance as well as overall realism. This is the only type of crossover that provides complete accuracy of amplitude, phase, time, and energy, and therefore the only type that does not distort the musical waveform.” The design of first-order voltage crossovers is relatively simple; the design of first-order acoustic type crossovers is significantly more difficult and expensive. Presumably, this is one reason why very few speaker designers undertake phase and time-coherent designs.

I wrote:

Note: The 4th-order crossovers in the JLAs are not compatible with the 1st-order crossovers in Thiel speakers.

Prof’s challenge to this comment is well taken. First, the words “not compatible” are too strong. I thought I changed incompatible to “inconsistent” before posting, but clearly I did not. Better yet, I should have written “inconsistent with Thiel’s emphasis on phase and time coherence in speaker design (through the use of first-order acoustic type crossovers). This is accurate, but one might still infer from my comment that the Thiel SI-1 can deliver phase both and time coherence. I did think this was true at the time of my post, based on the misinformation about the SI-1 using first-order filters and my assumption that Jim Thiel would only bother with first-order filters in pursuit of his signature phase and time coherent designs. However, according to my correction above, it appears the main advantage of the SI-1 is more accurate frequency response relative to active crossovers that use voltage-type matching. Is this distinction important in practice? It’s not clear. Jim Thiel presumably thought so, but Soundoctor makes some good arguments that suggest owners of active crossovers with voltage-type matching shouldn’t lose any sleep over it; see here http://www.soundoctor.com/whitepapers/subs.htm.

This still leaves a few more questions/comments from Prof that I should answer. I’ll address these, as best I can, in another post; this one is already pretty long.

Thanks, @prof

Jim Thiel's concept for the SmartSub line and the crossovers that go with it was to save the end user from having to get into the weeds learning how to integrate successfully. He also wanted to advance the state-of-the-art for sub integration (e.g., SBIR correction in the SmartSubs and acoustic-type filters in the SI-1). The downside for me in using the SI-1 with the JLA's is that I've had to effectively reverse engineer the SI-1 controls and understand the nonstandard aspects of its acoustic-type filter design. Talk about getting into the weeds! Fortunately, I enjoy this kind of puzzle. If I understand things correctly, the integration is essentially the same as it would be with the CR-1.

You should have it much easier working with the CR-1. There's really no need for you (or anyone else) to follow me through the weeds unless it interests you. Good luck!

Prof wrote:

If using a first order crossover with the sub were about maintaining time/phase coherence, I’m not going to have that anyway. The subs will have some delay, and will be placed behind the speakers, hence time delay. I could ideally get them phase coherent with the Thiels with the phase controls, but they’d still be at least a cycle behind in terms of time coherence (unless I go whole hog and want to digitize the entire signal to allow the mains to be delayed for time coherence with the subs - which I’m not yet willing to do.

Prof is right. Even if he places his subs at the same distance from the listening position as his main speakers, the group delay of his JLA subs means that the best he can achieve is phase, but not time, coherence. This is true for both the Thiel SI-1 and the JL Audio CR-1. I explained how I came to post fake news about the SI-1 on this point in my previous post. Apologies for that.

Prof wrote: And, btw, isn't that what the Thiel subwoofer integrator does? Digitize the entire signal to mains and subs?).

Nope. See p 3 of the SmartSub Owner Information manual, https://www.manualslib.com/manual/328036/Thiel-Smartsub-Ss2.html:

All signal processing [in the SI-1 Integrator] is done with analog circuitry. Digital circuitry is used for user interface, calculation and circuit control functions.

How does the SI-1 accomplish level adjustment and phase alignment, accounting for both SmartSub group delay and variable subwoofer placement, and why do SmartSubs lack a phase control? ... We know from the documentation that the SI-1 includes continuous phase adjustment in its bag of tricks for “Smart” integration. Since Thiel surely knew the group delay of its own SmartSubs, computing the phase adjustment to compensate for group delay at any crossover frequency is simple enough. On the other hand, I suspect that the SmartSub circuitry does not account for variable subwoofer placement. Instead, this issue is probably addressed by the instructions on p 9 of the SmartSub Technical Information manual, http://audio.manualsonline.com/manuals/mfg/thiel_audio_products/sw1_ss2_ss3_ss4.html:

The subwoofer(s) preferably should be approximately the same distance from the main listening area as the main speakers.

To the extent that this instruction is followed, the SI-1’s phase adjustment for SmartSub group delay is sufficient for total phase alignment. The user must enter the main speaker sensitivity into the SI-1, and Thiel knows the SmartSubs’ sensitivity. With this information, automatic level matching is straightforward.

OK, then, how can we implement the “straightforward phase alignment and level adjustments” mentioned in my previous post using the SI-1 with non-Thiel subs or if the subs are not placed at the same distance from the listener as the main speakers? Although in this case phase alignment and level adjustment are no longer smart, they can be implemented simultaneously using Method B, as described toward the bottom of p 4 of the instructions that accompany the SoundDoctor test CD: http://www.soundoctor.com/testcd/Soundoctor_Test_CD_v2-7-2.pdf. In this approach we use the level, polarity and phase controls on the subwoofer, just as we would with a more conventional crossover with voltage-type filters, such as the CR-1. Note, however, that the phase adjustment on non-Thiel subs will be the amount needed beyond the default phase adjustment that the SI-1 computes for SmartSub group delay. Fortunately, we don’t need to know the SmartSub group delay to implement Method B, because I don’t know the value of the SmartSub delay.

My initial attempts at implementing Method B indicate that the group delay for SmartSubs and my JL Audio f112s are surprisingly close (within 5 degrees at 80 Hz). I should do some more careful work to confirm this. If anyone knows the group delay spec for SmartSubs, I’d be very interested to know 

In the end, what did I get with the SI-1? Definitely a good price (thanks for not bidding you guys). Definitely a high-quality crossover with a nice remote. Possibly, improved frequency response with the SI-1’s acoustic type filters.*

BTW, I’m very pleased with the results I’m getting so far. I’m definitely hearing the improvements promised for switching to two subs, for using an active crossover, and for addressing phase alignment.

* I think my use of the DARO (automatic DSP room correction) that comes with the JL Audio Fathom line and the fact that my whole system, including acoustic treatments, is designed to optimize the sound at the sweet spot (vs. throughout the room) undercuts to some extent the Sound Doctor’s argument against worrying about frequency response. Jim Thiel believed the difference between acoustic-type and voltage-type filters matters in the cross-overs within our speakers, so why not also for sub integration? … I have no plans to test this conjecture by installing a more conventional crossover for comparison. So with no evidence to the contrary, I’ll continue to tell myself that the frequency response matters, too, just to keep myself happy.

But, as I said in my previous post, no one using a CR-1 or other crossover should lose any sleep over this IMO.

I forgot to mention what was probably the key feature in favor of the SI-1 for me. It has fully balanced circuitry, not just balanced inputs and outputs. The only other active crossover I found with this feature is a particular variant in the Marchand line, the XM44-2FBA. Since my pre and power amps both have fully balanced designs, I wasn't keen on inserting a device with unbalanced circuitry between them.

@prof 

Congratulations on your new gear. I think you'll like the improvements. You called yourself a sub newb in an earlier post, but IMO, your previous comments suggest you understand the principles just fine. So dive in and have fun.

My impression is that steeper is better for crossover slope, so probably 24 dB per octave is best on the CR-1. On the other hand, it might not matter all that much since your Thiel mains and JLA subs should both have good linear frequency response in the overlap region, even at 12 dB per octave. Easy enough to try both if you like.

As before, I recommend the Sound Doctor's sub setup instructions here (or see hardcopy with your CD): http://www.soundoctor.com/testcd/Soundoctor_Test_CD_v2-7-2.pdf.

As for crossover frequency, here's a range of opinions. Sound Doctor says,

I suggest never going below 80, even if you think your speakers go down to 40, or below. Even in a room where the existing "mains" have a pair of 12" drivers (each) you will get far better results if you ... correctly cross them over at 80, (or higher) ...

The Thiel SI-1 instructions state,

Unless the main speakers are very small, it is usually preferable that this setting not be higher than 80.

There's no explanation for Thiel's recommendation of 80 or below; it might be specific to some feature of the SI-1 or Thiel SmartSubs (probably not specific to Thiel mains because the SI-1 is intended to work with any make of main speakers). Alternatively, it might have something to do with staying well below the frequency at which the period equals the sub's group delay (about 120 Hz for JLA f112s). Above this point, the minimum timing error doubles to 720 degrees. 

Finally, a team of audio pros working with JLA f112's and Thiel CS3.7s chose 100 Hz in this example, http://www.soundoctor.com/studio/.

I decided to follow the Thiel recommendation (because I'm working with all Thiel components except for the subs). I selected 80 Hz as my reference but also stored preset configurations at 60 and 70 Hz for comparison. I can switch between these from my listening position with the SI-1 remote, but to my ears, they are indistinguishable. They all integrate seamlessly, and I couldn't reliably detect differences between these settings –– either in the highs or the frequencies surrounding the crossover setting. Still trying to test this.

I suggest you also consider settings somewhat above 80, e.g., 90 or 100 Hz. In fact, I spoke with a support person at JLA who helped an audio dealer install JLA subs, a CR-1 set to 100 Hz, and a pair of Thiel 3.7s. The dealer said it was the best sounding system he ever heard in his space. I would think your 2.7s should not be all that different from the 3.7s. Maybe I'll try up to 100 Hz, too. Can't hurt.

I found the Sound Doctor's Method B (p 4) for adjusting phase and level to be easy and effective. Rather than depend on my hearing, I used a sound-pressure-meter app, SPLnFFT (available for iOS and Android), to easily identify the settings for maximum cancellation.

Good luck!

@prof 
oops

Finally, a team of audio pros working with JLA f112's and Thiel CS3.7s chose 100 Hz in this example, http://www.soundoctor.com/studio/. 

Actually, these guys chose 80 Hz, not 100. It seems that 80 Hz is the default choice for crossover freq. Still, no harm trying other values.

As in my prev. post, it's interesting that Thiel says 80 and below, while Sound Doctor says 80 and above. Too bad we can't ask Jim his reasons.

@tomthiel I’m a little late, but welcome and thanks for your valuable and insightful comments.

I like Thiel's augment mode where the crossover to the woofer is first order and only the subwoofer is higher order. … I use a stereo pair of Thiel SmartSubs and place them at the proper ear distance for best integration and use the room-boundary controls to adjust for early wall reflections.

My first sub was a single Thiel SW1 and PXO5 in augment mode. True enough, it integrated seamlessly right off the bat. As I gradually added acoustic treatments to improve my room, I generated numerous bass response plots that clearly showed the SmartSub's ability to eliminate front and side-wall cancellation. Very neat! There's no other product I know of that can do that.

I bought my SW1, serial no. -4, from Dave C., a Shaolin martial arts master who used to do graphic arts work for Thiel. Did you knew him? According to Dave, my sub is a pre-release SW1 that Jim used to demo the first-generation SmartSubs at audio events. He'd give it a tune-up before every show ... can't say if that improves the sound today.

Keep in mind that when conventional subwoofer integration is employed, the frequency response at the listener position is optimized at the expense of all other positions in the room. Therefore the average power response in the room is wrong and the resultant sound is artificial.

Bottom line: I think that bass-generation position is very important and I position my subwoofers where they are distance-correct and let any room problems be addressed via Thiel's sophisticated distance controls or room treatment. 

Based on these comments, I tried killing the room EQ on my current subs, and I think you’re onto something. Cutting out the EQ seemed, at times, to remove something not quite right in the bass. Anyway, I'll keep this 'back to analogue' option on the front burner as I continue to tweak.

PS: The Classé DR amps in your system; very cool! I’ve been driving my Thiels with Classé amps since the mid 90s, but I jumped in too late to sample any Dave Reich designs. After being shut down by B&W, Classé was recently acquired by Sound United. It’s another conglomerate, but I think it's a good sign that Dave Nauber is reassembling his design team in Montreal. They had just announced a new generation of amps when B&W cut them off, so they should have a running start getting back on their feet. Together with the good news about Rob Gillum’s new business — Yea! — there’s hope for continued enjoyment from both brands.

Tom, I have serial No. -4 of the SW1, now in mothballs. I'll probably put it up for sale at some point. This is the prototype that Jim used to demo the SW1 at audio shows until the SS2 model was released. I bought it from a former employee (or contractor?) who did graphic arts work for Thiel. I used the SW1 with a PX05 that Rob configured for my 3.7s. It sounded great and blended seamlessly without any fiddling.

Tom, do you know how the circuitry controlled by the rear and sidewall adjustments works to counteract Speaker Boundary Interaction Response (SBIR) effects? SBIR involves cancelation between reflected waves (with negative amplitude) and direct outgoing waves at certain predictable frequencies. SBIR problems cannot be addressed with equalization, so I've always wondered how this unique Thiel technology works. The only thing I can think of would be to sum the original signal with a suitably delayed signal filtered to within a narrow band of the problem frequency.

Another question: Why didn't Jim include phase controls on the SmartSubs? Although it runs counter to the plug and play SmartSub philosophy, a phase control can be used to achieve phase alignment between subs and mains at the cross-over frequency.

@tomthiel: Thanks for your replies to my questions back on 8-24. Sorry for this delayed follow-up, life is keeping my busy these days.

I’m by no means an expert on analog delay, having only read an article you can find by searching for ‘boss talk: I love analog delay’. The article describes the history and workings of analog guitar delay pedals based on bucket brigade device (BBD) integrated circuits. One currently available example is the Boss DM-2W which can manage adjustable delays up to 800 ms and can be had for a not overly-dear price of $150. 

However, analog delay pedals are prized for the distortion they add to the delayed signal, even after filtering mostly high-frequency noise, and that might seem to disqualify them for high-end audio applications. But clean filtering might be easier to attain for SBIR compensation because that involves isolating and delaying narrow-band, low-frequency signals –– not wide-band guitar tones. Maybe Jim worked out a design along these lines.

Worth repeating: This is unsupported speculation from a true non-expert.

@prof I too am sorry for your setback and wish you a thorough recovery.

I would add Technical Audio Devices (TAD) Reference Series as another product line whose design shares many of the same goals with Thiel but with a cost-no-object market strategy. These are crazy expensive speakers even at used prices. TAD’s concentric Coherent Source Transducers and slanted baffles should be familiar to Thiel enthusiasts. A TAS review http://www.theabsolutesound.com/articles/tad-reference-one-loudspeaker-tas-218-1/ describes the crossovers as "asymmetrical [with] non-classic shapes", possibly similar to Thiel crossovers which are acoustically (not electrically) first-order.

I heard a pair of Reference One floor standers http://www.technicalaudiodevices.com/reference/ at an early (first?) Axpona sourced with a master copy reel-to-reel tape of John Lennon solo. It was spooky real. I was running Thiel 2.2s at the time, and had not yet upgraded to 3.7s, treated my room, added subs and a Thiel SI-1 crossover, or upgraded most of the rest of my system. At the time, the TAD system was far and away the most convincing music reproduction I’d ever heard. It was long ago and a different room, so I can’t reliably compare it to my current system. But I can say the gap is much smaller now.

On Thiel dealers: My sales rep in one of two high-end dealers near me was a doctoral student studying voice in our Music department in the mid 1990s. He was quite knowledgeable about audio and sold me on a Classé / Thiel system. I’ve stuck with those two brands ever since with no regrets.

@tomthiel Thanks for the counter narrative on the TAD reference speakers. Very interesting.

The Hales line was perhaps(?) another example of a faux coherent speaker line. Their cabinets featured concrete baffles that were sloped similarly to Thiel's. I auditioned a pair at the other audiophile store in town before deciding to buy my 2.2s. When I told the salesman I was impressed with Thiel speakers, he brought out the Hales, pointed out the sloped design, and told me they were like Thiel speakers except the concrete baffles made them better. They completely lacked the realism that sold me on the 2.2s, so I moved on.

Maybe it was the salesman more than Hales that misrepresented them as being similar to Thiels –– I never bothered to look at the Hales sales literature. Apparently, Hales used a sealed cabinet for tighter bass and touted flat frequency response as their main selling point. They used fourth-order Linkwitz-Riley crossover networks.

There's a (mostly) enthusiastic thread on Hales speakers here: https://forum.audiogon.com/discussions/hales-design-group-speakers-how-good-were-they. But it's much shorter than this thread; so there!

Hope everyone had a fine Thanksgiving (for all those here who celebrated that day yesterday or at all). ...

Here's another back story question: Early on, Thiel and Classé Audio seemed to have a close, if informal, relation in marketing and in-house systems (Classé and Thiel were marketed together mid-1990s at my local high-end salon, tomthiel still has legendary DR-era Classé amps). OTOH, the good fit between Thiel & Bryston has been well covered in recent posts. Over time, I heard fewer mentions from Thiel of pairings with Classé and more of pairings with Bryston (the two brands competed in roughly the same market niche). I speculated that this had something to do with Bowers & Wilkins purchase of Classé and their pairing at Abby Road Studios.

Question: How much was this shift at Thiel due to the B&W acquisition vs. technical differences in the evolution of the Classé and Bryston lines.

Side note: The Classé brand was eventually shut down by the conglomerate that purchased B&W. An attempt to resurrect Classé within the Sound United conglomerate is currently underway. In contrast to newThiel, new Classé is trying to retain most of its design team and culture. It will be interesting to see if the Classé reboot succeeds.

Thanks, Tom. I’ve never had an opportunity to listen to DR-series Classé amps. I think my first Classé amps were the first post-DR generation. They tilted to a warm sound. Maybe that agrees with your comments.

My current-generation CT-2300 has a much more transparent and neutral sound to my ear. From what little I understand, its design is fairly distinct from previous Classé series, including the DRs.

Not sure if serial numbers for all Thiel models are of interest here, but I have the following pairs:  CS 2 2s: 7599,7600 and 3.7s 131,132. I am the original owner of the 2 2s and (probably) the third owner of the 3.7s.

@vair68robert Thanks for your post back on 9/11 with the link to the article comparing the Classé CA-2300 and the C-J MF-2550 SE amps running 3.7s. I've been running 3.7s with a Classé CT-2300 (same as CA-2300 except for the case). Starting with the amp and speakers alone, I added a single Thiel SW1 sub via a PXO5 passive X-over, and finally a pair of JL Audio subs via a Thiel SI-1 active X-over. Along the way, I installed an array of tuned (60 Hz) bass traps on my rear wall to fix a severe SBIR problem –– this in addition to more standard room treatments.

The CA-2300 handled the 3.7s quite well from the beginning, but adding the Thiel SI-1 and the dual subs made the most significant improvement. High frequencies were sweeter and stayed sweet to louder volumes, the sound was more transparent, vastly improved extremely immersive sound stage, etc. Adding dual subs with the active X-over lightened the loads on both the amp and the speakers. It's hard to say how much of the improvement came from each component, but I suspect at least some of it had to do with easing the load on the amp. Consistent with the review, while the Classé amp performed quite well on its own, this suggests that there might be room for improvement in its load handling abilities.

It's worth noting that the reviewer didn't prefer the MF-2550 SE in all respects. He preferred the C-J Premier 350SA's weight and soundstage and referenced the Classé's (presumably superior) subtlety. I'm not shopping for a new amp right now, so I'll just make a note for future reference. If your room and budget allow, you might consider adding sub(s) with an active cross-over as an alternative to upgrading to a heftier unit if you're otherwise happy with your amp.

vair68robert,

Starting with a pair of 2.2s in 1996, I too have mostly bought used since then, and it's still been a rather expensive (but rewarding) hobby. I'm surprised and pleased to learn that CA-2300s are fetching $5k+ used. That's a bit more than I paid used for my CT-2300 in 2011.

I admire your DIY effort upgrading your electronics. I lack expertise and skill with a soldering iron in that area, but I did take a deep-dive into acoustic measurement and room treatment as well as designing and building an ultrasonic vinyl cleaning system to earn my DIY wings.

Back on the topic of adding subs, I should add that Thiel cross-overs, whether passive or active, do a superb job seamlessly integrating subs with main speakers.

Bob

Speaking of boundary compensation (that is correction for SBIR --- Speaker Boundary Interference Response), Thiel smart subs have SBIR correction to compensate for reflections off the front and side walls, but not the rear wall (behind the listener). It turned out I had a major rear-wall SBIR problem that could not be fixed by adding forward facing subs along or near the front wall of any size or quantity. The only cure was a sizeable array of *tuned* bass traps on the rear wall, tuned in my case around 50 Hz (where regular bass traps have very little effect). The key diagnostic for rear-wall SBIR is that the drop-out frequency shifts with distance from the listening position to the rear wall, but not with sub placement along the front wall. Once that problem was fixed, I did go with a 2-sub solution that works very well.

The distance from the LP to the rear wall gives you the quarter-wavelength of the SBIR dropout. From that you can find the dropout’s center frequency. If that frequency is well above your bass range, you can feel safe that you don’t have a rear-wall SBIR problem.

My room dimensions are roughly 15’ (front wall width) by 22’ x 8’3". I wish I had higher ceilings, but otherwise the room sounds great –– after treatment for early reflections, general bass control, and the aforementioned tuned bass traps for SBIR.

I strongly agree with other posters that emphasize the importance of getting your room’s bass response under control before going too far with adding subs. If your bass reverberation times are too long, adding more bass is likely to seriously degrade your sound.

....

I’m likely the guy with the Thiel SI-1 active crossover that @tomthiel mentioned. Although the passive PXO crossovers also do a fine job blending subs with mains, active crossovers also relieve the mains from reproducing low bass. This audibly improves the sound of my 3.7s (for a number of physical reasons): sweeter highs and plays louder w/o distortion. I was looking at Marchand active crossovers just before my SI-1 came up for sale. I wanted a crossover with balanced circuitry, and one of the Marchand variants was the only one I could find, other than the SI-1, that had this feature. In principle, a properly configured Marchand X-over should bring similar benefits, but the SI-1 was designed by Jim Thiel to make configuration a snap, even for non-experts.

@tmsrdg re Thiel SI-1 (from 12/15)
I purchased an SI-1 about two years ago to integrate a pair of JL Audio f112s with my 3.7s. It's in perfect working condition, and I have yet to experience any problems with it. Sounds like I'm lucky on that score, and hope that doesn't change. @tomthiel, it would be great if a repair option for the SI-1 does eventually arrive.

The SI-1 replaced a PXO that was driving a single SS1. In my experience, both the PXO and the SS1 deliver perfectly (to my ears) seamless integration with no fussing around. However, the active SI-1 noticeably improves the mid and top end performance of my 3.7s relative to the PXO. The high end stays sweet to well beyond any volume I'm interested in hearing. At the time, I was concerned that switching from the SS1 to the JLAs would mess up the integration, but there's no sign of that audible to me.