Thiel is considered by many to be a small signal transducer. It works best on vocal and/or small ensemble work. I agree and chalk that up to broad range requirements for overlapping drivers. OK, fair enough.
I have been dreaming in the way-back machine. I remember product development sessions back to the beginning. Come with me to 1975 at the Georgetown Road Compound, developing the 01 and 02. For nightly listening sessions, the speaker under test was placed in mono on a stand in a lovely-sounding room previously described here. The crossover network looked like a bird's nest in free space. No board, lots of space. We hung it from a ceiling hook by twine to some fixed parts, usually the heavy inductors. Many of the parts were clipped in with roach clips. Jim knew the circuits so well that he could make flying changes during music while capsulating what the change did. In an hour session we would compare a half dozen variables for him to take back to his night-time labfest. We reviewed graphs and compared notes at breakfast.
One ever-present outcome was that the finalized product always suffered from the crossover being bedded down onto a board and put in the cabinet. We always noted it and lived with it. About a month ago, a rabbit hole led to this particular warren by way of a remembrance dream. In the early years, I reviewed and journaled all service and repairs. It was not unusual to get speakers with burned resistors and even charred masonite boards. (No fire because of limited oxygen in the sealed cabinets.) I recognize thermal distortion as an avenue for improvement, and the 10 watt vs 25 watt resistor choice discussed here led to some experiments. Under musical load, resistors get hot enough to burn, and coils can get hot enough to unwind if not mechanically bound. That's serious. Resistance and quality of conductivity change with temperature. Beware of dynamic distortion
On a different progression, I have mentioned my personal health journey that has now been greatly resolved. The root problem is neuronal overload, and the primary cause is exposure to electromagnetic fields. I now own various EMF meters and devices and have mitigated much of the EMF pollution that put me in mortal danger. I, of course, now point those meters at everything, and was amazed at the soup around the 2.2 when rocking. Serious levels.
Put two and two together to equal redesigning the 2.2 crossover into a three dimensional package with more space, separation of resistors from other components, and physical separation from the drivers, in its own enclosure on a 4' umbilical. By the way, the package seems elegant to me, and in visual harmony with the cabinets. I'll post a photo when I get something presentable. For now let's talk about the woofer. It has its own board, positioned vertically with both sides exposed to air. All resistors are on the back to form an updraft cooling chimney, with the other two (midrange and tweeter) boards oriented similarly with all resistors facing inward with air inlets on the bottom and a screen on top. No metal anywhere. The rig is big enough to accommodate the largest of the CSA / MRA, etc. caps, as well as anything smaller. In the present case, I disassembled one (original prototype # -0004) cabinet, removed the crossover and wired only the woofer to the input terminal. The other (-0003) was left as a reference with the internal XO and woofer intact, but with the midrange and tweeter disconnected. I built the woofer section for -0004 by the new layout while using all the original parts in the original orientations. No upgrades, changes or replacements to confuse the issues.
Yesterday I began listening and today I continued with a long comparison followed by a set of measurements. My measurements are, sadly, still all sweep based via FuzzMeasure. Their impulse information is extrapolated mathematically, not from actual transient impulses. I'll get there eventually and get more insightful information when I do. Let's say that the two speakers via my available measuring techniques look to be virtually identical. There are no observable differences in frequency response, group delay, phase plots or waterfalls. My technique is to listen to each speaker in mono, placed beside the other; and then measure in place so that the mic hears what I hear in the same position in the same room. Then I swap right for left and listen and measure again. I am confident that I am measuring the speakers with no meaningful room noise or equipment differences.
So here goes. We have these two matched speakers with the difference being XO layout and removal from the enclosure. The test takes me back to Georgetown Road over 40 years ago to hear that delicate, lovely spatiousness before buttoning it all in the box. One recording was Pink Floyd's remastered "Take it Back" single which is both layered and detailed along with dense, dynamic and driving. #-0004 reference "A" exhibited splatting on every bass+drum hit. It would be attributed to amp clipping or passive radiator and/or woofer bottoming - overload / or just plain too loud for the material. But I didn't turn it down. On -0003 external XO "B", those same hits were audible, but not extreme, in fact they played more like "hard punch" than distortion. And all the nuance of the band and backing vocals hummed right along. Other audition material included the entire album of Patty Larkin's Stranger's World and Stereophile's Test record 3 cuts 3 and 9 - very detailed, dynamic bass-based work. Of extreme interest is that speaker A sounded like a woofer, somewhat woofy, lacking the detail one would get from the midrange and tweeter. Speaker B sounded like a full-range presentation; subjectively it was satisfying on its own terms. I had to check to make sure I hadn't somehow left the midrange or tweeter connected. I hadn't.
For me, today is a game changer. All the other upgrades have provided various levels of insight, even excitement. And their inclusion has always seemed contingent on effective budget analysis for most contribution per affordability. But this change is non-negotiable. It changes the league from small-source vocal ensemble speaker to throw anything at me.The persistent motivation had been how to shoe-horn all those large upgrade parts into very small spaces in the very well-braced 2.2; as well as how to then simultaneously address the re-bracing that would improve performance. There simply isn't room in there to do it all. I first took out the XO to install my bracing upgrade for testing. But I won't go down that rabbit hole yet. This performance upgrade is totally engaging. Next I'll implement the midrange and tweeter sections of the External XO. By the way, Resistor 1 and 2 got hot enough to burn and coil 1 got pretty warm - all mounted on stand-offs on a vertical board in open air. Imagine the environment in the sealed cabinet with a woofer within spitting distance producing its heat and EMFs.
Now it's back to the drawing board to re-orient those two resistors for better convection cooling.
Nice to visit. Thanks again for all the help and inspiration many of you have provided.
TT
