2021-2023 Isiris Update

[JonMarsh]

For various reasons it makes more sense to me to have a succinct new thread covering the investigations and updates this year for the Isiris. Only a few key items that will be addressed right away, but there might be more soon if the negotiations for early samples of a new midrange are completed.


To summarize what is underway:

• Updating some cabinet hardware, including the mounting for the cabinet bases and inserts and gaskets for the woofers
• New woofers, RSS315HFA-8 wired in parallel replacing the original Aurasound NS12-513A 4 ohm drivers wired in series, expecting a net 2pi sensitivity gain of 6dB
• Implementation of the TA6 crossover concept update, planned since 2014, modified as necessary for the new woofer system. Parts have been in storage since then...
• Remeasure all driver data, including tweeter and midrange.
• Build and test new crossovers
• UPDATE: This has now changed to going as bit further, updating the midrange and tweeter drivers, using new samples of a PuriFi midrange part received earlier this year.

Re-assembly of the first modified woofer cabinet was completed this afternoon... and if anyone thought the previous cabinet with Aurasound woofers was heavy (I did), then, as Bachman-Turner Overdrive would say, "Baby, you ain't seen nothing yet!" with the RSS315HFA-8 drivers, which must be about double the weight of the Aurasounds...

But feels like more...





Impedance plot has no surprises... sometimes I hate surprises! Fb is right where I expected based on Unibox and VituixCAD, about 32Hz.




More measurements to come soon- possibly tomorrow. There's a lot of logistics travel on the agenda most of the rest of the week, though.

[Efalegalo]

Damn, that's pretty!

[Scareurpasenger]

Does this also mean updates for the upper assembly as well?

[JonMarsh]

Does this also mean updates for the upper assembly as well?

Well, originally it didn't, but after this morning, I think it might. I wired funds to a certain European company for some of the first samples of a new 7" midrange driver they're producing. I was considering that for a design update and building a new set of upper modules, but frankly, I'm having problems finding the fossilized bamboo to build another set just like the first ones. So I'll probably mount a sample of the new midranges in one of my existing basic test cabinets, and if it lives up to the very preliminary data sheet, then I'll probably move to updating the existing cabinets. The update candidate tweeter will fit fine in the existing rebate in the cabinets, but the new midrange is a true 7" part, based on dimensions. The Accuton is a 6-1/2".

This could change, I'm going to keep checking local suppliers, but so far it hasn't been encouraging.

[JonMarsh]

Basic RSS315HFA-8 acoustic measurements

Some more basic measurement data on the RSS315HFA-8 update-

Nothing fancy here, had a very busy day logistically on a variety of topics including getting ready for the run to CA tomorrow (and another on Friday). But after a late lunch, I figured it couldn't be too hard to get a few SPL checks at different measurement distances, and get a basic HD2+HD3 check at "reference level", which is 2.83VRMS (monitored with oscilloscope), which for these drivers would produce (in theory) 91 dB in 2pi radiation (IB). Of course, the cabinet isn't IB, but it is big...

This kind of measurement for this kind of cabinet is a bit of a tradeoff in setup because one wants to see proper baffle step, and one doesn't initially want to see tons of room reflections, and having two drivers complicates things (because they can interfere with each other causing dips or near nulls.

So, this SPL measurement is at 1 meter (I'd recommend 2m possibly for a cabinet this size, and did take some additional preliminary data for FRD files for crossover development) and with Half-Hamming window filtering, but a reasonably long window (100msec) and with the cabinet out in the room like it would be setup for a Cardas style playback setup. Mic is centered between the drivers.





No surprises here... and the SPL response matches well what is predicted with Unibox. With Cardas positioning there will be some boundary boost in the low bass below 40Hz, and things should be pretty neutral, though probably a bit warm in the 40-65 Hz area.


The distortion measurement was taken under similar conditions, but a bit closer to NF distance- I.E., 24". No change in drive levels, though mic gain was adjusted downwards.





Interestingly, there's not much difference in the HD2 and HD3 levels... this is promising as regards the overall motor BL force linearity and cone positioning.

Certainly I have no qualms about proceeding based on this data- though I have all the original measurement data, I'm going to redo those for the midrange and tweeter, and double check Z axis origin time alignment. But that will have to wait until the following week, due to planned travel.

[JonMarsh]

Some more measurements completed in preparation for the crossover design

Here are a few of the desired measurements before starting the crossover design, to fully assess if the actual in cabinet response matches my predictions and double checking former measurements:

First, an overlay from one microphone position at the listening position, not changed for any measurements, along with what the expected target response level will be:





Then, a couple of combination files to be used as an aid for double checking Z axis alignment with the individual driver measurements, and to illustrate the characteristics of the cabinet design:


Woofer + Midrange





Midrange + Tweeter





As Evil Twin might say (he's away in the Outer Rim), "Everything is proceeding according to plan..."

[JonMarsh]

Things came up with the real estate stuff that had to be addressed at the last minute, but necessitated skipping the planned trip to CA on Friday.

But this gave me some time at home to work on a crossover trail design based off the TA6 update planned back in 2014. I'd say it's mostly looking pretty good, though NOT inexpensive!

In this VituixCAD layout, I decided to do some of the component breakout I usually leave to preparing the BOM and component choices for the production schematic, as I already had a good idea of how that would to based on the TA6 schematic, for which I already have a full set of components (since 2014!)

Nice looking results, so time to make a donation to Kimmo again. :T





Of course, with a 4 ohm woofer load instead of 8 ohm, that section will be completely redone. However, there were a lot of "adjustments" made so now I need to see what parts overall I will need when I'm ready to do a new build on this.. (which depends on free time availability - much logistics stuff still to do this summer)


Crossover points are ~500Hz and ~2600Hz, based on polar response capabilities and driver distortion profiles at elevated playback levels (fancy way of saying "LOUD")





With the need to up the SPL on the midrange and tweeter, to match the higher woofer sensitivity, the overall impedance does dip to 4 ohms a couple of spots- but the current phase is near zero, I.E., resistive, so this shouldn't pose problems for any normal 4 ohm capable amplifier, much less something like the Cambridge EDGE W, or Halcro DM68.





So, it looks like time to fire up the browser and order some more parts this July 4.

There is also an alternative midrange driver which will be investigated, due to the increasing difficulty of sourcing the C173-6-090; it may actually be an upgrade, which would pose seem interesting questions. Funds have been wired to purchase half a dozen early samples from the manufacturers; these aren't in retail yet.

[JonMarsh]

too much multi-tasking going on- just got back from another logistics trip. Need to unload from that now.

But all the crossover parts I ordered a week ago have been arriving, and getting sorted- ready to start board layouts as soon as I have some spare time.

With Liubov's younger daughter's family arriving tomorrow from CA for a few days, that won't be for a while...

[JonMarsh]

What with local logistics (first meeting with house construction supervisor), visit from my fiancé’s younger daughter’s family, as well as getting ready to drive to SoCal to see my daughter and her family (leaving in about 10 minutes) there hasn’t been much time for audio work, but a set of boards for the LF crossover have been cut, and layout done, and the input and output terminals and inductors mounted- barely got squeezed in!

And a new 4K video disk player arrived, Panasonic DP-UB820, a semi-budget player that can still handle all the HDR formats, and with very flexible audio outputs and discrete decoding built in. We just haven’t had time to worry about things like watching movies or some special TV shows- a new Apple 4K box and remote is sitting here unopened.

I never figured retirement would be so much work!

[Carl V]

8) enjoy your R-n-R

[JonMarsh]

Thanks Carl! It's been a busy couple of weeks, but both get togethers went well, it was good to see everyone!

now back to CA tomorrow morning (just got back here last night) to close out one more storage unit. Gotta one more big one, plus some motorcycles to move!

[JonMarsh]

I have updated the driver mounting hardware- but...

Isiris work is going to be on hold for a while- we were notified by the property management company that the owners want to sell the house we're renting, so 30 days to find a new place and move.

That's going to be the focus pretty much all of September.

It's also a difficult time- my fiancé's older daughter was diagnosed with cancer a few weeks ago and we went up to Seattle to spend time with her family and advise on some topics- which resulted in a new test, and now a full body scan yesterday.

Hopefully we'll get through the month and have a new place to live soon- I have a showing arranged this afternoon, and have already applied, but turn-over on rentals is pretty fast in this area- often just a few days.

And this morning we have a pre-drywall meeting at the construction site with the builder... a full day ahead!

I hope everyone has a great Labor Day weekend! I was just reading this morning about how one of the "traditional" things going on this weekend (buying a new car) is going to have a major crimp in it due to lack of inventory at dealers, and more and larger production shutdowns were announced this week.

Based on the current hospitalization data for COVID in far too many states, including Idaho, maybe they're replacing buying a car with getting sick and trying to get an ICU bed. . That's just about as hard as buying a new car at this time- our state's hospitals are literally full up, and they're looking at things like converting large classrooms to wards.

Just for the record, I go nowhere without a KN95 or N95 mask, but most of my fellow state citizens seem to think it's 2019, and Delta is about Delta blues.

[JonMarsh]

With a Blitzkrieg effort, a new place was leased, many things were moved to storage temporarily earlier this week, and yesterday we got the keys and started moving in- spent a long day on that.

Oh, and my daughter had her 3rd child, Chloe, who looks very healthy but doesn't seem to be very happy being on planet earth yet (I got about a half dozen pictures, all varying states of crying!). Both mother and daughter are doing very well...

And I'll probably be moving the Isiris test cabinet TA1 set to the new domicile tomorrow. If you checked the pantry and kitchen, you'd think we'd been in there for a month already, so you can guess where some priorities lie. I've been out to the storage place to get one of my multi-mode hand carts, and an area rug from my old place in Livermore, and I'll be taking an iPad and a couple of small Homepods over on this next run... gotta have tunes, you know!

As for the Isiris cabinet, it's very heavy now...

“There’s that word again. ‘Heavy.’ Why are things so heavy in the future? Is there a problem with the Earth’s gravitational pull?”

[Steve Manning]

If you checked the pantry and kitchen, you'd think we'd been in there for a month already, so you can guess where some priorities lie.

That's because you have a good Russian woman looking out for you!

[JonMarsh]

Exactly! And she's made a lot more progress this weekend- but we both have!

Isiris head unit and most new crossover parts are relocated, and the bass cabinet is going over today.

This is a brand new place- investment rental- and blinds are scheduled to go in this afternoon, and the guys to do the backyard work and sod came by Sunday afternoon to check things out and clear out the weeds...

And most of the basic gear and furniture is now installed in the AV corner of the great room.

Like I say, I never dreamed retirement would be so much work! And this place is "temporary"...

[SMABB]

Since being in update-modus, which appears to have the possibility of becoming a comprehensive one, I might ask if the Avalon Saga at any point in time has been included as an option or inspiration?
Same basic layout of the bass section although slightly larger on the Saga, but the upper module is completely different in approach.
The rearward tilt continues from the bass module and distance between mid- and treble-units have increased. This in a strange way at a higher starting point for the upper module resulting in a somewhat higher position on X-axis for both upper module units compared to Isis.. sorry Isiris....

Any thoughts on this would be highly appreciated.

[JonMarsh]

I'm familiar with that model, but due to all the stuff going on with the retirement and relocation process and preparing for a new house being built, doing a new top module soon just isn't in the cards. Yesterday we visited the site under construction, and some good progress has been made, but there's a ton of interior work, and speaking to one of the construction managers earlier this month, I know that supply chain issues due to both materials availability and shipping issues is a problem- he forecast an additional month slip back then. I know from other reading, for example, that appliances are a problem...

The Saga information is interesting from an industrial design viewpoint- download the brochure, it's skimpy on details, but has some nice pictures. I would be curious to see the impact on the vertical and horizontal dispersion from the complex front baffle design.

I'm also curious what a glass+carbon tweeter is about...

The price of the Saga in Europe has been reported as about 170K Euros, incl. VAT. So, lop off VAT, convert to dollars, and the number is just a little less. Quite a bit more than the Isis. I don't think from what I can research that anyone has them yet.

Now, the next thing to be tackled (way late!) in in measurements and development is the Osiris upper end, which is a three way mid bass/midrnage/tweeter; for Michael Bell's system, it is used with dual woofers below 100Hz. Midwoofer is the AS250, midrange as selected by Michael is the C168-990, though I now have a pair of C168-890 I'll be looking at also, which is more of a midoowfer. Micheal's tweeter choice is the big SS Be tweeter, the 7140, which I'm measuring in the Osiris top module, but I'm looking at the T34B as a possible upgrade for the Isiris.

Note, I have a baffle width correct test cabinet for the AS250 which incorporates a PR with the idea of testing a simpler three way based on that woofer, (which is damned expensive to begin with) and the midrange and tweeter drivers.

Now, if the testing on the PuriFi midranges I bought looks good, then only relatively minor modifications will be needed to adapt the top modules to that. And the Jantzen and PuriFi tweeter frame diameters are compatible. So, let's just say because of everything else going on, I'm focused on low hanging fruit, after the rental relocation is finished on Sept 30.

A quick reminder regarding regarding the Osiris with pictures Michael took- note, his Isirs are quite a bit deeper, he wanted ported versions. It's the old monkey coffin problem. Construction material is MDF, if I recall correctly. Beautiful paint job.






My Isiris, at Beverly's, back around 2015.





Constructed with fossilized laminated bamboo plywood.

[JonMarsh]

just getting back to this...

Dang, the last two months have been busy, but not with audio stuff!

However, I've measured the proposed PuriFi midrange driver in a small test box, and decided to proceed with that, and I'm moving forward in evaluating using the T34B (packed away in relatively inaccessible storage at the moment) and moving forward with T34A stand-ins.



Quasi near field measurements show very good SPL and distortion characteristics, quite competitive with the C173-6-90, probably superior in fact.




The physical dimensions are not quite compatible for the rebate, but I have a rabbeting bit ordered that should solve that issue.

I note with a little surprise that these are still not listed on the PuriFi web shop and don't seem to be available through PuriFi vendors yet- I guess it helps to know people and order direct for early samples. Maybe they're going to be OEM for a while.

This weekend I "verified" the mechanical compatibility of the tweeter- not that I expected any issues, based on data sheet information- even the input terminals are located in a similar fashion at 180 degrees.




The new tooling should arrive this week, so I hope to have midrange measurements in the system enclosure by the weekend, as well as all the RAW FRD files prepared for the complete system.


Now, some have questioned why I am doing this whole rework-

I'm looking for several things: some performance improvements in distortion, greater output level capability, (especially in the low frequencies) and a dramatic reduction in the driver BOM.

So far, it seems like these goals are not incompatible.

[Bear]

Dang, the last two months have been busy, but not with audio stuff!

However, I've measured the proposed PuriFi midrange driver in a small test box, and decided to proceed with that, and I'm moving forward in evaluating using the T34B (packed away in relatively inaccessible storage at the moment) and moving forward with T34A stand-ins.

[ATTACH=CONFIG]31522[/ATTACH]

Quasi near field measurements show very good SPL and distortion characteristics, quite competitive with the C173-6-90, probably superior in fact.

[ATTACH=CONFIG]31520[/ATTACH]


The physical dimensions are not quite compatible for the rebate, but I have a rabbeting bit ordered that should solve that issue.

I note with a little surprise that these are still not listed on the PuriFi web shop and don't seem to be available through PuriFi vendors yet- I guess it helps to know people and order direct for early samples. Maybe they're going to be OEM for a while.

This weekend I "verified" the mechanical compatibility of the tweeter- not that I expected any issues, based on data sheet information- even the input terminals are located in a similar fashion at 180 degrees.

[ATTACH=CONFIG]31521[/ATTACH]


The new tooling should arrive this week, so I hope to have midrange measurements in the system enclosure by the weekend, as well as all the RAW FRD files prepared for the complete system.


Now, some have questioned why I am doing this whole rework-

I'm looking for several things: some performance improvements in distortion, greater output level capability, (especially in the low frequencies) and a dramatic reduction in the driver BOM.

I'm looking forward to seeing this one come to fruition. I don't have room for 100L+ (x2!) of cabinet, but these statement projects are definitely fun. One might observe that money you've spent is cheaper than money you spend, but I may be the beneficiary (some day!) of the quest to reduce BOM cost. I am curious to see how the T34x (I have a pair of T34As) pairs with the Purifi mid. The two of them now cost less than just the Accuton mid (assuming availability...).

[JonMarsh]

I don't know of any reason availability shouldn't commence in a reasonable fashion at sometime- but then, there are those supply chain bugaboos, and in that kind of environment, I wouldn't be surprised if OEM's will have first dibs. I think it's good enough it should get a lot of attention, but I suspected that, which is why I did a direct order early on before first samples were shipping to anyone for half a dozen.

I have the T34Bs (in relatively inaccessible storage), but I'm curious to see what kind of measured performance I can get with the T34A, using an elliptic notch filter at resonance- if it works as well as for the DA25, then we may be "in like Flint" with the lower cost part.

We'll see soon... now I'm just waiting for the rabbeting bit. They say Nov 5-12. I know the carrier has picked up the package now... so I'm hoping for the 7-8th!

[Bear]

The availability comment was more targeted at the C173-96E (?) -- or was the original using the -95E? I didn't think those were available to us mere mortals. The -96 is still on their site, but I don't see the -E anymore.

[JonMarsh]

Accuton mids...

The availability comment was more targeted at the C173-96E (?) -- or was the original using the -95E? I didn't think those were available to us mere mortals. The -96 is still on their site, but I don't see the -E anymore.

The original midrange driver is a higher level of unobtanium, the C173-6-090. This one has the acoustic dampers and has a much flatter distortion plot, largely free of resonance amplification in the desired passband. Madison appears to still have some... at a fairly dear price.

https://www.madisoundspeakerstore.com/approx-6-midrange/accuton-c173-6-090-6.5-ceramic-cone-midrange/

The 96 is not as clean and not as easy to work with, lacking the cone dampers, but it is less expensive.



And the E version is pretty much gone, as you note.


In the good news department, tracking data shows the special router bit has made it's way to the Nampa transit center from FL... should be delivered in the next day or so.

[JonMarsh]

Voila!!

Delivered yesterday...








Exactly what's needed, but not inexpensive... beautiful bearing from CMT. It should be, considering the price.

[Finleyville]

Mr. Jon Marsh

All the Pittsburgh people (me, Wkhanna, PewterTA) are eagerly awaiting the final iteration of this design. I am almost seriously considering the Tidal Contriva's (stupid insanely expensive, small house $$) for my bucket list speakers. If these prove about 80% close, I'm all in. Probably so will PewterTA at least!

[JonMarsh]

Mr. Jon Marsh

All the Pittsburgh people (me, Wkhanna, PewterTA) are eagerly awaiting the final iteration of this design. I am almost seriously considering the Tidal Contriva's (stupid insanely expensive, small house $$) for my bucket list speakers. If these prove about 80% close, I'm all in. Probably so will PewterTA at least!

I always love a challenge, so I'll do my best to motivate you!

:T

[JonMarsh]

Progress is out most important product...

Many distractions today, but eventually made progress!

We had to drive out to Eagle to pickup a Russian friend of our realtor who is spending the night here while she and Liubov put their heads together- (Liubov needed to see the route- is not all that comfortable working with a GPS and navigating traffic).





The new router and rabbeting bit were assembled today (I like dedicated tools- just a budget basic Bosch model) and since my ear guards are in storage, I figured I'd try my Ear Pod Max headphones with the noise cancelling turned on- worked like a charm.




as calculated, from the inner hole, the 5/8" rebate was pretty much perfect...




And the data sheets didn't lie, the inside holes required were essentially identical.


So, just to get a peak at the new esthetics...






Now I've got to review the cables and connectors and back plates, and decide how to proceed- I'm leaning toward keeping the Kimber cable and existing plates for measurements, maybe for crossover prototyping, but I've got a mess of 4 pole SpeakOn connectors and cables, and 90% sure I'll update to those in putting the final configuration together.

Now I've just got to put this back together and get some measurements in this cabinet.

[5th element]

My god the perspective on the Purifi driver, with its wonky surround, does weird things to the eyes in those images.

[chrisn]

Delivered yesterday...





[ATTACH=CONFIG]31535[/ATTACH]


Exactly what's needed, but not inexpensive... beautiful bearing from CMT. It should be, considering the price.

Maybe my eyes deceive me, but is that a double height guide bearing? I could make use of those...

[JonMarsh]

Yeah, the guide bearing is at least double the usual value, that and the OD, plus meeting the unusual 5/8" requirement are what decided this one for me.

Completely satisfied with how it worked.

[Evil Twin]

Time to realize the power of the Dark Side

Now we are ready to realize the full power of the Dark Side in the three way design study.







I clearly see the potential of these new drivers... after all, black is back!


Do note that this project is a further development of the original "Three Way Evil Design Study" initiated in 2012...

I will be watching closely over Jonmarsh's shoulder to be sure that his ongoing efforts fulfill my original vision...

It would be most unfortunate if they do not- one need only to consider the fate of Grand Moff Tarkin




or Captain Needa, to understand that nothing less than a fully successful effort will be accepted.

[tktran]

Evil Twin,

This thing looks deadly.

Please do harmonic distortion testing 86, 96 and 106dB please. Looking foward to seeing... nothing!

[Bear]

It's a Star Killer that can also do the Kessel Run in 12 parsecs, assuming that it didn't just kill the beast in the maelstrom out of pique.

[dar47]

And that top half looks like a killer 2-way I am needing for a bedroom system.

[JonMarsh]

There's a road map of sorts for that on the Purifi site...

Ah, but then we'd need to switch out the driver to the PTT6.5Wo4-NFA-01 woofer with the 6 mm stroke - and expect to tolerate a bit more midrange distortion. And probably change the crossover frequency and design. But this is doable. The result would be a lot like the two way web shop design using their crossover kit.


I'm writing up an outline about what what concepts and concerns are driving the crossover development- I'm pretty close to having a first pass done. And even at the stage of not having finished the midrange first pass, and still investigating some points regarding the woofer crossover, it's not looking bad...

[JonMarsh]

A surprising amount of work later...

Slow work takes time... and planning.


This is just a quick post to share my thinking and some points about targets and progress.

Oh, and I have another simulated frequency response plot, that doesn't look all that different, but a number of issues have been addressed with positive results.


First, I'll share my base targets and concerns, and how that impacts design.

• New midrange and tweeter, and woofers, hopefully all working together to provide a lower distortion floor, and likely some increased output capabilities
• New crossover points to optimze based on THESE driver capabilities- in practice, this meant the woofer to midrange crossover frequency was lowered slightly, taking advantage of the new midrange driver output capabilty, and the tweeter crossover point was lowered slighly for the same reasons.
• System sensitivity is raised to 86-87 dB range, BUT, as all driver sections are now 4 ohm loads, special attention had to be paid to keep the impedance up, especially in the low frequencies, and hit target responses without creating a current monster like the Avalon Indra or Isis (look at reviews, see LF load reaching 2 ohms in mid bass).
• Phase tracking is always important, and the easiest test of that is reverse null behavior; the updated network fares far better than the one responsible for the first shown SPL graph. Yes, as is often the case, the SPL graphs look quite similar... but the underlying behaivor isn't. That's always a significant part of the challenge, and necessary to avoid off axis lobing.
• Only a single measured axis was used for this "rapid prototyping" of the network design- I've found that the 15 degree axis is usually represenative of a significant part of the total power response, as well as the a plausible listening axis- and as the drivers measure nearly the same at 30 degrees, the 15 degree axis data was used for fast development. Putting these cabinets on a turntable is not realistically within the ken of most mortal men. Of course, ET could manage it, but he wasn't available when measurements were done. We still can't do anything about earths gravitational field.
• Yes, there's lots of little wiggles in the data files, because even in a large room that's what's happen when you capture at 1/48th octave smoothing with reasonable duration measurement windows for capturing data. No makeup or wigs here to make things prettier.
• ALL of the drivers have notch filters for upper end breakup, to tame resonance amplification of distortion products as well as bury as much as possible the resonant peaks from affecting the base response. Looking at the result, you'd never guess ANY kind of BlieSMa T34 tweeter is used in the top end. Breakup peak? What breakup peak? Even the T34B would require this kind of network, and I think it would be interesting to see if there is a benefit to the B version or not if both the A and B have such a network correctly implemented for each dome material type. We'll get to that some day.

Next task is to put together an actual component BOM, and check the bins for parts already on hand!

While I had what I thought was a good first pass crossover design for the RSS315HFAs to go with the original midrange and tweeter, that has gone by the wayside.

[tktran]

Looks great.

I too, have found that sometimes looking at the 15 degrees off axis is instructive, because it's not filled with all the lumps and bumps of the diffraction effects, and I get overly distracted trying to massage it all out.

How did you settle on the 400Hz HP pass on the Purifi? It looks like it could take 200Hz or so..

[JonMarsh]

Looks great.

I too, have found that sometimes looking at the 15 degrees off axis is instructive, because it's not filled with all the lumps and bumps of the diffraction effects, and I get overly distracted trying to massage it all out.

How did you settle on the 400Hz HP pass on the Purifi? It looks like it could take 200Hz or so..

Past experience with the earlier version which ended up using a higher crossover point, and with the X1 SLAMM clone;

wanting to keep the excursion down on the midrange at high SPL,

Wanting to keep the crossover cost as reasonable as possible; (remember, system cost down with improvement in performance was 1/2 of the dual goals)

Well designed motor of the RSS315HFA means that distortion is monotonically decreasing with increasing frequency to beyond 1kHz on the top end; very similar to the Aurasound woofers. And quite different from most 12" "sub" woofers.

location of effective BSC for the bass cabinet versus the midrange module

Keeping a taller source up to 350Hz at least, and minimizing floor bounce effects.


Foiks have commented on why I had relatively high crossover points on the Wavecor Ardents, too, but it was for the same reason.

It's handy to take a electronic crossover with variable frequency points, like a variety of the Behringer models, and like the ones I used to design for the company I worked for back in the late 70's, and explore in real time the basic impacts.

[JonMarsh]

Well, I was expecting the 2nd batch of crossover parts to be delivered today, 2nd day UPS air, but I guess, 'tis the season, and now UPS is saying Friday. Disappointing, as I have the first woofer, midrange, and tweeter boards close to completion...

Although I'm exercising reasonable care in the overall system design to keep costs down, including the crossover, a solid build with key parts of high quality won't be inexpensive- around $800 per speaker for the complete crossover BOM, not including boards and connectors. A lot of that is premium capacitors in key locations in the midrange and tweeter crossovers.

So, a sneak preview of the test board being built for the midrange- some parts not in yet, but it's getting close.




This is pretty compact compared with the original Isiris midrange crossover.

So far, things seem to be on track for bringing this in at about the same component BOM cost as the Wavecor Ardent, about $4k. That's a huge reduction form the $10K of the original version, and I think this one will have some performance advantages.

Cross fingers...

[Mikerodrig27]

Pretty cool! I like that you are challenging yourself to bring the cost down on this design. I haven't used the drivers that you are using but I have had very good luck with Dayton drivers for subs and full-range speakers.

Purify has also been stirring the pot lately. I have been seeing a lot of them. I will be curious to see your impressions once you get it all buttoned up. Looks great!

[dar47]

Wow your trying real hard this time, I don't think I have ever seen an Electrolytic in your xovers before.

[JonMarsh]

Wow your trying real hard this time, I don't think I have ever seen an Electrolytic in your xovers before.

It's for the Fb LCR zobel for the midrange driver, with a resonance down around 110Hz; high enough that I want to kill any impedance interaction in the crossover (especially since it's high Q) and low enough in frequency it needs big components. But will not be really involved in anything handling any significant signal level, over -40dB.

And 660uF in film caps would be pretty outrageous cost wise, even for cheap ones, in relation to the functional impact.

[JonMarsh]

Step by step, inch by inch, now the more interactive phase of fine tuning has started

Sunday morning I finished wiring up the test crossover boards for one speaker (only bought parts for one- this is development phase) and made some quick cables for hooking everything up to the cabinets.




This is not my idea of an ideal room for listening or testing, as the hardwood laminate is everywhere, and it makes things very lively in the midrange. But for 2 meter tests at the nominal seated listening height, things aren't far off what I expected.

Initial results were very close to the VituixCAD simulation, but that plus some listening using the VOX app on my MacBook convinced me to tweak levels a bit, and also to make some slight adjustments to the midrange bottom corner Q, nudging the crossover point down to a bit below 400Hz.

Here's what that resulted in...





The top level SPL line doesn't look all that much different, but think of it like a submarine- it's what's going on below the surface layer that may be more critical.



I did a fair mount of listening with uncompressed cuts stored on the MacBook for this purpose- of course, it's just going through the test system DAC. But it still gives a good idea...


Material like Edie Brickell ("What I am", with Jerry Garcia on guitar), Hiromi (Spiral), Spyro Gyro (East River Blues), my cuban CD, "Hoy Se Cumplen Seis Semanas", Daria "All or nothing at all", Jacque Loussier, "Chorale No. 1_Sleepers Awake_ Live", The King James Version, "Corner Pocket", Emily Remmler "This is me", and some cuts from the first Robert Plant and Allison Krause album.

I have NO regrets at all about moving forward with the Purify 6.5" midrange driver- it's superb. And the pairing that's been achieved with the BlieSMa T34A is quite remarkable even compared with my hopes and expectations. I don't say things like this lightly.

Moving on to checking distortion, the T34A. is showing NO signs of resonant amplification of distortion products, it's curves are very flat, and below -70 dB for both HD2 and HD3. And you can see from the SPL curve that the tuning I simulated with VituixCAD for the notch filter has been successfully realized- the ultra low ESR/DF Alumen caps in the notch filter are a big factor in my opinion.

OTOH, the experiment with the Jantzen cored inductors isn't working out as hoped- it's like there's a flat distortion noise floor at about -50 dB for the midrange and woofer, with HD2 and HD3 nearly identical, so I suspect core distortion, and have ordered some alternative parts to try for the midrange crossover first.

All in all, a good day with some real progress...

[Efalegalo]

They are just so damn pretty!

[JonMarsh]

They are just so damn pretty!

:rofl:

:T

Yeah, I feel the same way, which is probably why I dug one of these out to work on first before cramming the storage units here to the gills!

Obviously, Neil Patel deserves a lot of credit for that!

But I think of the changes in creating the Isiris and this updated version as being influenced by Sly and the Family Stone... "I want to take you higher!"

[Efalegalo]

Parts-Express had 7 RSS315HFA-8 in stock last night. All gone this afternoon. I'm sure this thread had something to do with it :-).