Thiel Owners

First up I would reinstall the original inductors back in..then listen again.

Inductors of any type are non inear devices.

Tom D.

improvedsound- you’re in the deep end, so take care.

My knowledge is anecdotal and experiential. I have experimented with foil inductors enough to learn that I’m over my head. Indeed a foil inductor is a more perfect inductor than any wire inductor. However all components exhibit all characteristics ie, an inductor also has resistance and capacitance plus the products of reflectance interactions between them - and time factors as these interactions play out. So, changing anything changes everything. Flash back to my direct observation of Jim’s learning curve about these matters, as each iteration of each product grew into its maturity. Part of that advancement is about understanding and implementing more aspects of such component interactions.

There are direct aspects that can be adjusted. A foil coil will have much less series resistance and different capacitance characteristics in space and time per equal inductance value. In some cases those changes can be corrected via addition of a series resistor and/or layout changes to compensate for circuit capacitance and resonance variables. I have been able to accomplish some of these requirements in my experimentation and coaching. However, I’ll say it again, I’m a novice and not qualified to re-engineer Jim’s work beyond basic changes.

So I keep the extant inductors because I know them to be best-of-form (at any price), providing an anchor for any other circuit changes. Similarly with caps, ClarityCap provides very predictable performance in all measured aspects - it’s no accident that they are an industrial / aerospace / professional company making products based on solid engineering performance. (I have tested and measured some brands that de-spec some aspects to achieve some euphonic outcome.) A person could spend a lifetime making sense of the correlations. I’m keeping it simple enough for me to make sense of it within my constraints.

Regarding resistors - the Mills sound better than Jim’s. Fair enough, they cost 5x as much. Jim developed, from first principles, what is called the Ayrton-Perry winding which is non-inductive with minimal parallel capacitance within an inexpensive sand-cast case. His circuits and layout assume those A-P characteristics. I know that Mills MRA-12 is a direct drop-in with the same characteristics. However, I don’t know much for certain about Path, or film or bridges. I do know that altering the type of resistor or any component will have effects that are sonically important. I also know the hundreds, perhaps thousands of hours learning by listening and integrating with fundamentals of physics to get the holistic design expressed in each product as designed.

I do not have the knowledge to help unravel what might be going on when some component is changed. I do know enough to be extremely cautious and make changes that don’t screw anything up. I do not deny that some folks have gotten results they are quite happy with, but the system is complex enough that a dose of luck could be operative. So for myself, caution wins. Trust your ears. If something sounds "off", call it off. Then the fun begins putting Humpty-Dumpty together again.

This very long answer is to recap territory we’ve explored in these 223 pages over that past few years, which new participants probably haven’t read. Keep the faith, learning is good.

@improvedsound Assuming all values match OEM and the connections are consistent with the schematic, that board should be singing. From your images, I see nothing out of place although I can’t see everything. Total resistance of the coax board should be, IIRC, 32 ohms. If all values and connections are good, that suggests the foil coil is the culprit. I also put foil (Erse) in that position, would have to dig into my notes for the gauge.

Could also be you just hear things differently than I do. But there are now at least 3-4 of us who modded at least the coax board on CS2.4 and reported good results. I suggest verifying all values, connections/layout, and total resistance. If those are good, put the OEM coil into that suspect position. 

@improvedsound

My L5 inductor (ie, the one on the coax board, paralleled with 28 uF cap and 2 ohm resistor) is Erse FoilQ, 16 gauge, 0.15 mH.

I have all Mills MRA12s, the 30 ohm position I used paralleled 60.2 ohms for improved cooling.

The biggest difference I see in your pics compared to my build is that I have full capacitance in single caps. I *greatly* benefitted from tagging along with @tomthiel as he explored his “renaissance” project. Not to mention his considerable coaching.

All main caps are Clarity CSA. Coax feeds (14 and 28 uF) are 630 V, subfeed (43 uF) is 250 V, and shunts (100 uF) are 100 V. Each of the feeds and subfeed are bypassed with Multicap RTX at just under 1% (eg, 0.1 uF on the 14 uF cap). The shunts are bypassed with the 1 uF yellow caps common in classic Thiels (Elpac?).

That said, Your use of paralleled CSAs (eg, 18+10) should be fine, certainly not worse than the original CS2.4 which used 27+1. Best wishes resolving the issue.

[edit: post number 11111. Look at what you did, @jafant !!]

@improvedsound

When I completed building the xover boards, I let it run for 24 hours with a dummy load for the components to settle then hook it up to the speaker drivers. At first, it sounded like a AM pocket radio, after a few hours play time, sound quality keep improving until about 100 hours then the SQ become stable and the sound open up with a holographic sound stage... YMMV!