2021-2023 Isiris Update

[JonMarsh]

This week has been a real barn burner, and not about speaker work, unfortunately.

But I think I'll have the epoxy coat garage floor project wrapped up by the end of Saturday, and I'll do a post in the woodworking advice thread with the link to the original YouTube I selected as a guide, and my own comments about the experience and how I'd amplify on the guide. There's more than a few "gotchas" for a project like this...

[Efalegalo]

The RSS315-HFA-8 continue to be out of stock.

They went out of stock just before Thanksgiving. For months they were indicating the speakers will be in stock on 3/25/22. Provided weekly updates that 3/25/22 is still the correct date. However....today...its been pushed back to 5/13/22.

I wonder if they will ever be in stock again.

[Finleyville]

The RSS315-HFA-8 continue to be out of stock. I wonder if they will ever be in stock again.

Well, isn't that a kick in the head.

[JonMarsh]

I expect they will be, though it's too bad they've been delayed this much more.

It's their own brand, and I'd hazard the opinion that the main issue is getting the production done, maybe issues sourcing some parts of the driver.

It's not like say, when a Wavecor part goes out of stock at PE, or an Accuton driver, and the only message is "out of stock" - the kicker on Solen is when they say "out of stock" but don't mention "available on backorder".

These certainly aren't normal times...

[JonMarsh]

Another CD that floated to the top of my current play list is one from the “Chicks” (AKA “Dixie Chicks”) - Taking the long way home. Great songs, and the sonics shine in the current setup. It’s way past time to pickup “Gaslighter”, their 2020 release- many planned music purchases are on hold while I do things like epoxy coating the garage floor! And moving stuff from storage! And guiding my wife in her technique for the stuff she wants to repaint… it’s going to be a very busy spring and summer.

[JonMarsh]

BTW, I'm hearing from people that HAVE recently received previously backordered RS315HFA8, so I think that PE's problem is that they're a victim of their success with this part- I.E., by the time they got the late March shipment in, it was already all sold out and now it will be the next order in May when people can get fulfilled.

[JonMarsh]

Work is progressing on remodeling the garage workshop area, but there is much left to do.

However, in the meantime, I have investigated the presence range dip in the measured response and believe I have some simple crossover "fixes" - for those who prefer dead flat or very close to it. It involves removing one RC shunt zobel network on the midrange section, and adding one capacitor in the current tweeter network. How simple this is raises the possibility of having this as a switched option to support ranges of acoustic environment and program material mastering.

I will not implement and test this, though, until we are in the new acoustic space of the new house, sans the echo chamber behavior of our rental. That should be some time in May...

I have also located the pair of BlieSMa T34B tweeters from storage, which raises the possibility of an "upgrade" for this set. However, I find that the traditional series notch filter approach is working very well with the T34A, and even the T34B has a breakup mode at about 32 kHz which would need addressing. Curiously, the off axis polar behavior of the T34A is superior to the B, based on BlieSMa data sheets. This is another factor worth investigating. Additionally, there are some "subtle" differences in the on axis behavior, based on the data sheets, which would require addressing in the crossover. The B is the flattest between 2kHz and 20kHz, the A having a slight hump in the response between 2kHz and 7 kHz, centered at about 3500Hz. The breakup mode behavior in the data sheets is so similar, though, that a crossover modification for that aspect is likely unnecessary.

[JonMarsh]

Does having the other set of cabinets (top module and woofer module) and the original crossovers moved to the new house count as progress?

At least, I can report that the two opposite thirds of the garage (north and south) have been mudded, sanded, fully primered and the south side painted today- tomorrow is the north side, then assembling steel shelves from Costco on Friday and relocating the big stack of boxes in the dining area (taken out of storage) into the north side of the garage.

And one storage unit is completely emptied and turned in!




Now, to put this in perspective, keep in mind that this unit was packed front to back- you couldn't walk into it at all, in early April. And things came out of it like, a large elliptical trainer machine, for example (in pieces of course- I'm not superman- that thing weighs a ton assembled).

Slow work takes time, but there is definitely progress.

[JonMarsh]

Meanwhile, I am listening to some new disks for "evaluating" the results...

The Sony Music K2 HD mastering release of Sarah McLachlan's "Surfacing",

Ry Cooder, "The Prodigal Son" (picked up on this one from some high end equipment reviews)


Surfacing is quite lovely- the best I've heard it. Of course, it should be!
And the import Steve Wilson Remaster of Yes "Fragile" (yet to spin- saving it for "stereo")

[tktran]

I forgot to ask re: this

https://www.htguide.com/forum/attachment.php?attachmentid=31350&d=1625258709

All drivers measured at the listening positioning, which mic from one position

What's your reason for doing this?

Is this your method of determining the required baffle step compensation, and approximate attenuation required for the other drivers? And how come you choose that particular target response (flat to 1KHz, and down 4dB by 20Khz).

I take it as your in-room target response, or is it the anechoic target response?

best regards,
Thanh

[JonMarsh]

This was just an in room measurement for all the drivers to make sure that the raw response (overlapped) made sense from the viewpoint of sensitivity and a target system design. They were not necessarily windows to the same measurement gate, but each was measured at the same level.



The blue target line assumes flat response to 1kHz and then what may be a "voicing" in the final design (per discussion in other threads, re voicing module) about HF target.


Then, to get a feel for the raw relative phase, I like to measure the woofer and midrange simultaneously,



And the midrange plus tweeter.



There will be some destructive phase combination at the outside of the band, but I've found this kind of check useful to be sure my assumptions about the driver positioning and Z axis relationships are sound for the intended design.

[tktran]

OK, I'm with you. Some QC checks done early.

One of thing I like about VituixCAD is the ability to design at different distance/axis to the measurement axis.

So with a dual channel measurement system like Fuzzmeasure's Automatic Device Correction, VituixCAD allows one to measure at the closest practicable distance for a long time window, and design for a listening distance eg. 2-3m or 5, 10m.

This "any listening positioning you desire" is in the Options tab.

[JonMarsh]

OK, this is not a "design update", but a schematic update to reflect an updated BOM being prepared reflecting an effort for a value oriented approach for the crossover, especially for a couple of forum members that I know are seriously interested in following up on this...


Key concept is component choices, particularly capacitor choices, and paring down the number of parts but keeping a good quality level considering the impact on sonics, as I've generally experienced.

This is NOT the LR2 alternative design, which is still to be finished (partly assembled boards are in bins, getting move over to the new house tomorrow, even!)

This reflects the topology and summed component values of what I've been testing and listening too, but doing things like consolidating the number of caps, specifying certain high value proposition components in the midrange, and alternates in the tweeter crossover.

So, on the woofer, Solen 400V PP caps. For the midrange, mainly Clarity copper connect CSA caps, probably one of the best values on the market right now. I'd even recommend them for the tweeter if controlling costs is mandatory. But if you have a few extra shekels under the sofa cushions, I'm suggesting either Clarity MR, Jantzen Alumen Z, or Clarity CMR caps, in progressing order of financial pain.



This is NOT a final recommendation; that is subject to further testing and listening sans the echo chamber of the rental house.

And you will notice there are a couple of switches in the crossover, which can be used to trim the presence range; in the current crossover configuration I've been listening to, there is about a 2 dB dip in most of what would be called the presence range, between 1kHz and 4 kHz. Open the RC zobel on the midrange, and switch in the 4.7 uF cap in the tweeter crossover, and VituixCAD tells me it should be pretty close to dead flat in the presence region. That is YET to be verified, but let's just say it's advance information based on best available data.

The speaker set that was modified at the rental will be moved to our new place on Tuesday. The other speaker TO BE modified is already there, as are most of the parts to do that.

Also, one more thing, note I will be testing the T34B and providing an adjusted circuit for any brave souls interested in the Beryllium tweeter. It's funny, the FS impedance bump is essentially identical between the two parts, and the breakup mode for the aluminum part is at about 29kHz, and for the Be part at 31 kHz, BUT, there is a roughly 2dB difference in the on axis level between 7kHz and 20kHz, with the Be part being the flatter, higher output part.

[JonMarsh]

We're in the final stages of move out process for the rental house, and the one updated LF cabinet is actually moved into position in our new home for testing.

I have a few more mods to do for the midrange- tweeter cabinet, but hopefully will find time to start in the next week.

Dhar is going to have some time off to start construction on his, around the 4th, so final updates in the cabinet plans must be done soon, too, so I can get the updated construction drawing to him.

And let me tell you, going back and forth between our new home with carpet and a slightly different layout versus the rental with wood laminate reinforces the comparison to an echo chamber- even Liubov was making fun of the old place yesterday afternoon while we were working there, hooting like an owl and listening to the extended decay time!

Curiously, I've been in several Crestwoods in this and the old neighborhood, and the laminate floor seems to be popular... too bad for them.

Oh, and joy of joys, I found my TotalDAC D1 balanced in storage last weekend- when you stash something away years ago, one's memory can get a bit foggy about which stack of. bins it went into, especially when one wrapped it in a heavy duty hoodie for protection!

[Finleyville]

Good luck with your long-awaited final move. I am glad the new place's acoustics are much more conducive to good sound!

[JonMarsh]

You better believe I am too! Another busy day wrapped up…

[Efalegalo]

Its beginning to look a lot like the Isiris :-)

I figured I should share some pictures from my current build. This is going to be a very slow build for me, but I had a kick-start while focusing on it 100% during the week of July 4th (my work gave us the whole week off).

It took me about 50-55 hours of shop time to get to this point. I'm a slow builder because I'm still very much an amateur woodworker, and my woodworking has been solely limited to building speakers :-).

The top for me was VERY challenging..... VERY challenging. Not just because I'm an amateur builder, but I really do think it is difficult to build. In hindsight, it's worth every penny to pay Steve Manning to build it.

Perhaps I was being too difficult on myself, insisting on getting perfect cuts - but that's because I didn't want to use any polyurethane-based glue (which has excellent gap filling properties, but it is not as strong as traditional wood glue).

PS - I made many incredibly stupid cosmetic mistakes along the way (e.g., tilting the palm sander, cutting too close to the facet like with the pull saw, etc. ---- so many). But structurally - they are 10/10 (e.g., perfect glue-up, no gaps, fully sealed, etc.).

[JonMarsh]

Isiris in July

well, I never said these would be an easy build, BUT!

Just look at how the cabinets turned out in your final pictures! the proof is in the pudding, and this pudding looks pretty tasty to me!

Congratulations and good work!

[Steve Manning]

As someone I know likes to say "slow work takes time" Your putting in the time and the results are looking very nice.

[JonMarsh]

As someone I know likes to say "slow work takes time" Your putting in the time and the results are looking very nice.

Exactly! Assiduous efforts can produce very nice results!

[5th element]

Wow look at that. Those are coming along very nicely and that plywood looks as if it machines really well. Those are some crisp looking cuts!

[JonMarsh]

Dhar did a fine job.

It's interesting, I'm getting more and more inquiries about this design now, with the update on drivers and crossover. Due to other pressing projects, around the house and for Steve, I haven't been able to get back to them yet, but hope to soon. I have been sending out some 2D drawing packages to a few folks.

[JonMarsh]

Partly in response to the inquiries, and partly for my own satisfaction, I've started the groundwork for updated the Isiris cabinet documentation. My plan is to follow another recent project, with new "ground rules" for the drawing organization, and update the upper edges of the top module and setup cross integration between Shark FX, TurboCAD, and Shapr3D, and make a STEP file set available for those who want to experiment directly with those files, as in comparison to just working with the 2D sheet plans.

There are more interested parties... which I find interesting! And encouraging! I can measure what I've done in the past with CLIO now, but am still looking for my original gear, mislaid somewhere in storage- it has it's benefits, too, complimentary, like TurboCAD or Shark vs Shapr3D.

And to kick this all off, of course I had to buy a new computer, after coming across a special offer from Costco for Mac Studio Max. The M1 Mac mini I've been doing a lot of this work on was originally purchased in a configuration suitable for my music learning activities, and setup that way. Needs must something a bit different for the design work.

[JonMarsh]

I've been continuing to look at how I may best update the documentation, not just for the cabinet, but crossover materials. That's brought me back to evaluating the current status of Fusion 360, which was the tool used for the development of the Kurosawa project by Evil Twin. AutoDesk incorporated Eagle PCB into Fusion 360, which gets varied reviews. I have experience with Eagle PCB for projects at work with the original CADSOFT version. I looked at it after Premier Farnell took it over, but didn't bite. They added more library features and component library options. Some of the original Isiris crossover updates were done in Eagle, like this TA6 update.






So, I've been doing some evaluation using a Fusion 360 trial version, and created a current schematic update for the hybrid Copper connect CSA version, employing as many of the library component features for specific component documentation, efforts not possible with the older version. That turned out pretty fair; adding text to the file is not straight forward, but by default it uses Euro style resistor symbols, which I prefer anyway. I'm going to see if I can import my old model libraries into it, but I think I would lose some of the functionality of the current component models, so that may be a moot point.


But this doesn't look bad, and is fairly functional- additional text was added by two stepping into EazyDraw and putting that on a separate layer and exporting as a PDF. But I have to say the overall process for some tasks, like saving and exporting to a PDF in a local folder seems unnecessarily complex; not very user friendly interface and organization.

NOTE: UPDATED OCT 11, adding missing inductor L9.

[JonMarsh]

I spent a fair amount of time this morning combing through all all the VituixCAD and Fuzzmeasure results as well as double checking the last build boards- things had to get dropped in a bit of a hurry last March when we got the go ahead for 3rd party inspection and closing on the house.

I'm preparing to generate a crossover component layout based on the last LR3 build, using EazyDraw in the same method as I created what I call the "SIMPLIFY" LR2 build layouts.

A couple of small discrepancies were discovered, nothing really major, but things like the difference between a 120uH inductor and a 130uH. And that was an "unwound" litz part. So all this stuff is going to be correlated carefully, AND I'm setting up soon with the last crossover builds and going to measure with the Clio setup and the new Fuzzmeasure setup being put together with a new Earthworks M30 measurement microphone; ACO Pacific isn't responding to emails or quote requests, and their order page has been taken down, so I'm inclined to believe they've gone the way of all things of the flesh, as the old saying goes.

This is from the last measurement set I did at the old rental place, 1/12th octave smoothing. Before finalizing the next midrange build update, this has to be checked and correlated again. This is actually with a build optimized for the T34B-4, but measured still with the T34A-4, which rolls off slightly above 3-4 kHz in comparison; running for that requires taking one parallel RC element set out of the circuit. I am probably going to switch to the T34B for my own set, as I have two pairs of those on hand now, due to an oversight of what was already in storage. "Storage" is a double edged blade- very useful, but also a great way to lose track of stuff- like my ACO Pacific mics and 2014 MBP and the first Motu interface I bought! Still haven't tracked those down...

[JonMarsh]

Reminder- I'm sharing intermediate steps of development schematics, here. Early this afternoon I did some more evaluation and fine tuning and ordered more parts for testing with the cabinets. These changes involve small topology changes and value changes which in some cases even better optimized the cost structure for the combination of midrange and tweeter crossover...

I.E., WIP - stay tuned, but don't order parts yet!!

OTOH, things are looking very flat between 400Hz and 15kHz, and inverting the midrange produces very deep nulls, so of course phase tracking is very tight. The impedance curve is within a 4 ohm load boundary.

I'm assembling a 2nd test cart this week, with an Earthworks M30 test microphone, and should have that up and running by Saturday or Sunday, if deliveries are made on time. More crossover parts are on order from various countries, with deliveries by Friday.

It's looking up to make some progress...

[JonMarsh]

Given all the things happening on the international stage, this seems like a good time to remember and give a two thumbs up to another set of Isiris built starting from my plans back in 2016, in the Ukraine. I have many more pictures of this build, but these three will give you the idea...

[audioslave]

Wow....I wish I hade those veneering skills.

No rush from me....I am definitely in for more than a few pennies:


And many pounds....about 450LB to be precise:



It till be a while till I can start thinking on the crossover...

Mike

[JonMarsh]

Isn't there an old saying that goes something like "In for a dime, in for a dollar!"?

I've reminded Dhar a few times about that...


Now, if you look at that top picture of the cabinets before veneering, note the use of HDF as a top coat on the base wood, to get a nice clean smooth surface. At Avalon, they used Medite back in the old days- no idea about the process, now.

Yesterday was mostly spent pouring over the simulation work, tweaking some topology stuff and component values, and preparing updated crossover layout drawings for the tweeter and midrange for the LR3 configuration. This work will be checked against new measurements in the new house very soon, and I'll also be modding the 2nd cabinet set, and trying out a new damping material on the interior of the midrange cabinet. That one will have a T34B tweeter installed for testing, also.

So, with updated measurements there may be some adjustment of resistor values, based on level checks, but the response in these VituixCAD plots above 500Hz should be quite reproducible for either tweeter. And I expect to have cleaner measurements below, now that the speakers will be in the new living room with carpet and all, and setup more normally. Phase and reverse nulls are totally righteous at this point. I'll also be getting out the turntable from storage- it was sized to be able to handle a speaker of this weight.






The rough stuff up top is diffraction effects on axis, not using a grille assembly with diffraction control yet.

[Evil Twin]

After reviewing my apprentice's work, I found that some changes were in order... his power in the Force needs further development, to be truly worthy for continuing this three way design study upgrade.

Topological and component value changes have been made using VituixCAD with existing data, and three new board layouts have been drawn up. This includes adjustment to the HF crossover for either using the T34A or the T34B, which is also on hand. While the existing data still has diffraction issues in the top end, and room interaction in the bottom end, I believe we are much closer now to the desired solution...







Components have been ordered and received, and construction of new evaluation boards has begun...

[Evil Twin]

I have received some inquiries regarding the cost structure of the crossover filter system in development. This work is progressing as quickly as possible, and shipments of components have been received (another one today) and boards are having components mounted. Naturally, it should be noted that this is not a fixed issue, but one subject to variation, depending on the type and cost of filter components.

We will provide recommended and synergistic BOM proposals... an expected starting point will be a "Budget Mode" version, building on the initial concept for this development of optimizing the value proposition through lower BOM cost.


However, leave no doubt in your mind that this will not be the ending point- our Imperial spies have been insidiously acquiring information from a most unusual source, Tesla, the premier electric vehicle corporation in this sector, with which our personnel have direct past experience and relationships.

That is why, for those with the wherewithal, there will be a documented build employing the "Tesla option", introducing "Ludicrous Mode" for the crossover BOM.

[Evil Twin]

It is said that without Holocam images nothing is real on the Holonet- note this is not to be confused with the meager offerings of the "Metaverse". However, to assuage the growing numbers of enlisted constructors for this project, I will post images of the current "POC" or "Proof Of Concept" Crossover test board construction with most components mounted, but wiring not yet commenced. Note, these are neither "Budget Mode" or "Ludicrous Mode" configurations, but something in between deemed adequate to support listening tests with higher level source equipment, such as an informed Imperial Moff might have on hand.


Low Frequency Filter Board





Mid Frequency Filter Board.





High Frequency Filter Board.

[Finleyville]

What size are those mockups? Will these x-overs need an external enclosure?

[JonMarsh]

These are not mockups but current test builds in progress. The midrange is 12 x 16"; the woofer version shown is 12" by 16" but could be narrowed to 10" x 16", as in the current planning document. The tweeter is much smaller- I am out of town without access to the documents, but it's on the order of 12" long and 8" to 9" wide.

[JonMarsh]

OK, with the progress over the weekend in moving stuff into the garage (no patio furniture boxes there anymore!) and organizing the electronics room, crossover wiring is starting, beginning with the HF board.

As shown above, all the basic parts are mounted.

Just HF schematic is here:






First order of business is the HF notch inductor, L9, as you can't readily buy this part. But you can buy a quality part that you can unwind to create this value.



On the right is the "starting" part, a 100uH AWG15 Litz inductor from Jantzen, purchased through Parts Express, note the part number.

As a homework exercise, boys and girls, get out your calculators or copies of MathCAD/MatLAB or your favorite calculation too, and reverse engineer the Jantzen part. That is, using the inner diameter, and the overall length, and the AWG15 wire gauge for the wire diameter, calculate how many turns and layers it should take to create the 100uH inductor. This is your reality check.

Then calculate what you need for the 10uH inductor. It should come in the range of 20-24 turns- I've seen a lot of variation in online tools, and have to say I wonder how they're doing the math...

But the real point is, you don't need the math to be super accurate, IF you have a basic LCR meter that does good inductance measurements. Alas, my B&K seems to have gotten stashed away with with my ACO Pacific mics, so I ordered a new one today. In the meantime, I have pretty high confidence with the part on the left... based on calculation. But confirmation is better. If you have a sweep generator of some kind, and a scope, you could just measure the notch result, which should be 29kHz with the 3.3 uF cap for C20, (T34A) and 31 kHz with the 3.0 uF cap for C20 (T34B).

Also, you'll need either a solder pot or a high temp soldering iron to tin the Litz wire; if you choose to use a more conventional base inductor, then scraping the insulation off and tinning will be easier. Solder pots, like my Hakko, are recommended but may not be in your budget.

[PANDINUS]

Hi John,
can you hear the L9 engaging? did you test it with your ears?
And will there be a difference when using a recording from a CD, where the frequencies are not so high as to reach the membrane breakup and Hi Res with high frequency content over 30kHz?

[JonMarsh]

One of the key problems with high Q tweeter breakup modes is that it cases resonant amplification of distortion products of frequencies lower down, both HD and IM. This is part of where some otherwise fairly good performing tweeters in terms of their basic motor get into trouble (Peerless DA25TX, among many, for example), and get a reputation for being "harsh".

Here, that concept was first applied to the development of the Natali P MTM loudspeaker, and incorporated into both the woofer and tweeter crossovers, and quite successfully, considering how popular that design has been.


A much more recent example would be the Calliope system kit, as delivered by Steve Manning to the current owner of the HTGuide forum. This uses the DA25TX tweeter, for its very linear motor and wide and uniform off axis dispersion with rising frequency. But it has a nasty breakup mode up around 27-28 kHz.

Here is the response with HD2 and HD3 distortion plotted with the complete crossover for the initial POC build, not ported, no diffraction control measures:





And this is a polar plot set for POC build of this speaker 0-30 degrees in 10 degree increments, which highlights the otherwise excellent performance of the DA25TX:





And this is a polar response set going out to 70 degrees off axis (speaker on turntable),






BTW, a lot of us here do listen to high res material with high sampling rates, and energy above 20kHz. Not claiming to be able to hear that directly, but I can certainly hear the difference that the reduction of HD and IM makes at lower frequencies. And I can measure it, compared with the raw tweeter.


The original configuration for this system used a diamond dome tweeter with its breakup mode well above 50 kHz- no notch network was used. But the motor is not as linear as say, the DA25TX, much less the BlieSMa tweeters. The BlieSMa tweeters also minimize radiation from the surround, a technique which is also a focus of the PuriFi midrange used in this update.

In the last few years, I have seen others using this technique, and even a white paper from PuriFi on this topic. But my work and designs predate them by around 15 years.

[JonMarsh]

To facilitate the whole POC effort, I decided to add some binding posts to the test boards, installed on risers fabricated from the same phenolic material as the crossover boars are built on. I've done that, and wired things up; this is complete for now EXCEPT for the notch circuit inductor, which is built, but I will not install it until the replacement LCR meter arrives today and I can check it. All of the crossover boards are having binding posts added this way. Note, for gluing to this phenolic material, I use JBWeld, which judging by smell, is actually formulated like marine epoxy. It certainly mixes and behaves like the marine epoxies I have used, and accordingly, sticks like the taint of original sin. NO way will this stuff drip or run, just like marine epoxy.

As the binding posts still go below the board a bit, I added some Dayton feet on the bottom, ones I had a box of around are the same type I used on the bass cabinet of the Isiris, because I haven't been living anywhere previously where I could spike those cabinets. That's different now, but I may still leave them as is for now. The Dayton part number is DRFS1, and the distribution part number at PE is 240-712, should anyone be curious.

Now, the original Isiris cabinet build was in 2013, and the feet on those cabinets are in excellent condition, here in 2022, so I can attest to the quality of construction. I've used them for heavy sub cabinets, too. Recommended.

[JonMarsh]

A lot of distractions going on these days, but still making some progress... things are getting organized and setup in the living room finally for playing music and for testing one of the Isiris cabinet sets. The other is yet to be modified.

This is the nearly completed test midrange board. The output side binding post platform had to be re-glued, but that part should be wrapped up drilled tomorrow.

[JonMarsh]

What size are those mockups? Will these x-overs need an external enclosure?

I do have another crossover type in development which is smaller and less expensive, based on the LR2 crossover filter, with midrange phase inverted. This was started as an even more "budget" approach, but I'd frankly want to evaluate it carefully before offering that as an alternative. It would not have as steep a slope on the filter networks, which could have an impact on the sonics and distortion at higher playback levels. For comparison, this is the POC midrange test board in process for evaluation:

[JonMarsh]

The LR2 tweeter board is more complicated, in order to implement necessary/desirable impedance control.

[JonMarsh]

In for a dime, in for a dollar- may as well post a pic for the LR2 woofer board- keep in mind this is physically smaller. The single inductor is the same value as the first inductor in the LR3 board for the crossover frequencies chosen.

[1Michael]

Those Xovers are so darn pretty, they need to be mounted outside the speaker to show them off😍 ET would be proud.

[JonMarsh]

Those Xovers are so darn pretty, they need to be mounted outside the speaker to show them off😍 ET would be proud.

Yeah, like the Wavecor Ardent cabinet build Steve did with the crossovers on the back under a clear cover!

[JonMarsh]

Father Physics is your friend... if you make his acquaintance.

OTOH, one of my favorite sayings is "Trust, but verify". This applies to our understanding of father physics, too.



I unwound initially to just three remaining layers, which is approximately 30T. I measured and checked, and as I would expect, that came in around 27uH (inductance is related to the change in square of the turns) Then I unwound to just slightly under two layers, about 19-20T, checking as I went. Value is the same with 10kHz test frequency, but with an extra digit of resolution.

As many uncertainties as there are in the world today, Father Physics is not one of them.