Kurosawa Koncept System

[Evil Twin]

Do not confuse the impressive individual results of the AS190 with th C90-6-724- note the overall calculated crossover response posted earlier with the disturbance in the response between about 1200 and 1800 Hz, due to the midrange behavior.

[Evil Twin]

As shown here...

and here...

[sdl2112]

Nice progress Jon. I'm in a similar place. I've been scratching my head a little from differences in test results from my proto to real cabinet build on the C173. Long story short...the proto cabinet I built for the C173 was very narrow internally, not much bigger then the cut-out (external baffle dimensions were very close). It caused some lift in the baffle diffraction region actually helping the response. In my real cabinet I had to account for the difference. Adding more wool stuffing helped the peak to peak ripple by about 1dB.

[Evil Twin]

The C173-6-90 is sensitive to both baffle configurations and rear stuffing- this I agree. Optimizing the transfer function has to be done with care with a fully realized cabinet system and measurements taken with some care and showing the power response- otherwise, I find it's better to use the 15 to 30 degree response curve if one can only incorporate one.



The C90-6-724 is a different can of worms- I knew about the problems before buying these, and did so in the belief that they could be remediated. So far, that appears to be the case. Some of attempted to address the resonance issues resulting form the magnetic structure with LCR shunt filters, but I believe this is not the correct way to correct what is fundamentally an acoustical problem.

I won't have final data until these cabinets are loaded, but the baffle step will be no different and diffraction should also be quite similar, due to the overall test baffle similarities to the target baffle design.

It is a situation I don't often find myself in, where the drivers are all running full out, no series resistors, and therefore, a somewhat more limited set of opportunities for shaping the response. Still, with the Duelund target and what I have measured so far, I think this is going to work out fairly closely to the predicted design.

[5th element]

Do not confuse the impressive individual results of the AS190 with th C90-6-724- note the overall calculated crossover response posted earlier with the disturbance in the response between about 1200 and 1800 Hz, due to the midrange behavior.

I was was referring to the measurements in post 66.

HD is clean as is the frequency response sweep. The small irregularities around the 1-2kHz range could be diffraction related, either in the loudspeakers mounting, from the measurement system, or from the geometry of the drivers. It will be interesting to see how the off axis compares with and without the rear wave thingymabob!

Edit - I found another website that has a wealth of information and measurements on this driver.

https://www.hifi-selbstbau.de/datenbler-mainmenu-40/tiefmittelt-mainmenu-58/462-accuton-c90-6-724

Their off axis measurements show something different between 1 and 2 kHz, certainly with no overlapping of the off axis curves. They also show remarkably high 2nd order distortion when the drivers are pushed. I was quite surprised by this.

[neuro]

Greetings ET,
Any data on the effect of your damper assembly on the 3rd order distortion artifact at ~3kHz per the Accuton datasheet? Granted, if you clobber the frequency response per Accuton's "filtered" option you also clobber the distortion peak. However, based on diameter, a 90mm driver shouldn't start beaming until ~4.5kHz. So it does not seem unreasonable to expect usable response at 3kHz from such a driver. I want to like the 724, but that distortion! Would be great to learn that the distortion is also caused by tube resonance and can be addressed with a diffraction device. Then again (with no disrespect meant to the spirit of innovation evident in bending these drivers to your will regardless of their faults), why would one simply not bother with all that doubt and use the old stalwart 079...

[Evil Twin]

These are all sensible questions, and we shall have answers in due time... However, it should be recognized that the primary cause of the HD3 distortion spike at 3 kHz is resonance amplification of distortion products due to the intrinsic 9 kHz break up mode. The 724 does not have the damper pads of the C79, which helps with the overall Q of the primary breakup mode.

Review of the polar response shows that the highest frequency for which response is at truly equivalent levels is about 2500Hz; above this, divergence is evident at 3 kHz and continues to build with increasing frequency.

Whether the 724 can be made to produce a satisfactory result is certainly an open point, as the planned crossover technology does not use a high order roll off. The answer will be in the final distortion plots of the system- so getting to the point where those can be made is a priority.

one could justifiably argue that this is a very expensive test to ascertain the usability of the C90-6-724. Cabinet assembly gluing operations have started (yesterday) and an update will be provided soon on the overall work status- at the time being, getting work done is prioritized over talking about it...

[olu78]

Interesting observations of a legitimately high-end driver. Having perused the write-up, to my eyes the 2nd/3rd order distortion performance appears to remain very relative and consistent, with increasing level. As for the high-order distortion products, they are admirably low in level. I look forward to the system performance as a whole.

Sent from my SM-G935T using Tapatalk

[hyperducky]

Imperial Lambda Shuttle or FedEx International Economy Courier?

If this is a typical example, the UK has nothing to fear from Brexit.

Now, a question I have to pose to you, for an order placed Sunday evening, and received today Tuesday morning from the UK, which do you think was the more plausible delivery method to achieve this alacrity in delivery?

Custom carry via chartered Imperial Lambda Shuttle, or...

FedEx International Economy Courier? (at under $30 USD)

To clarify, the order was entered and shipped on Monday, and received on Tuesday morning- overnight service.




And as you might surmise, this is the capacitor order for the first stereo crossover build... including Jantzen, Duelund, and Obligatto capacitors.

I cannot seem to find where the Dark Lord has been purchasing these capacitors from. The Bothan spies have searched this forum high and low and they are missing the post where you or your alter ego Mr Marsh said where you are getting your high end caps from.

[scottvalentin]

I think it is Hi-Fi Collective or something like that - check the Minerva thread I think.

[Evil Twin]

Completing the front and rear panels...

After the outer two front panel layers are clamped up and set at full panel length, with alignment of the through holes, then the top and bottom end cuts are made, at the 4 degree angle matching the overall front panel "lean back". Here, the capabilities of the DWS780 sliding dual bevel miter saw become quite useful...





The front panel assembly is composed of three layers, which must be assembled with precision- the first two are of the same outer dimensions, but the inner front layer is stepped back on the sides, top, and bottom, by the width of those pieces, to provide an interlocking construction.




To facilitate accurate assembly, alignment blocks fabricated from scrap pieces are used in the actual clamping process to assure the correct alignment during gluing.


The sides and rear panels must be fabricated- the sides, at the 22" of the original material, only require the bevel cut which runs top to bottom, and which was implemented during the initial material rip process. But now it must be cross cut at the required 4 degree angle, creating a parallelogram configuration for a nominal enclosure height of 56".




The rear panel is a matter of tedious repetition- with six passive radiators (PR223BD02) on each panel, it is clearly an example of a process which is much faster to implement with CAD tools than with real tools...




It is a good idea to check the dimensioning against the actual PR223BD02, especially the through hole and rebate alignment...






More than half way through the process at this point...






Machining the rear panel to accept the binding posts is a multi-step process-

• First a 1/8" pilot hole is drilled at the center location of the binding post location
• Using the pilot hole as a locator, a 7/8" Forstner bit is used to 3/8" depth for where the binding post contact body sits.
• " pilot hole, to relieve the backside by 3/8" for the ring terminals and nuts
• Then a 1/4" through hole is drilled to just clear the Vampire binding post - adjust this size for whatever terminal post you should pick

[Evil Twin]

Baffle parts and Glueing the cabinets together

The inner baffles and braces are straight forward, though it should be mentioned that the lower open brace is fabricated with special considerations- the overall cut out diameter is sized so that the cutout pieces can be used as the rear baffle of the midrange enclosure.









When properly executed, the interlocking stepped front panel design facilitates component alignment when glueing up the complete cabinet- this does not relieve one of the need to check squaring and alignment, however- and check blocks were fabricated using the DWS780 and confirmed with a carpenters square before starting.


Complete dry assembly is recommended to double check fit and accuracy. Note also that support blocks are used at 1/3 cabinet height intervals to lift the assembly above the bench and permit inserting clamps from underneath.

In the first step, one of the sides is laid down, using wax paper between it and the assembly support blocks, in the case of epoxy runout, so that the cabinet isn't glued to the support blocks.


Initially the side, front, and top and bottom are aligned and clamped- a slow set expoxy such as West Systems 105 is used with 206 harder, using the available dispensing pumps to automatically assure the correct mixing ratios.






This is also the best time to install the 6" Schedule 80 midrange PVC enclosure, roughing the surface of the outside surface of the of the pipe before gluing it in, using Marine epoxy in this case, which drips little and seems to bond well with the PVC- both the flat mounting surface at the end of the tube and two inches of the outside of the tube are coated before inserting it into the recess provided by the inner front panel plate.

The next step is to position and glue in the inner baffle pieces (setting the position per the plans and marking on the cabinet inner wall is a useful aid if they become disturbed in alignment while piecing everything together) and glue and clamp the rear panel.





The last, and almost anticlimactic step at this point, is to glue and clamp the other side panel, which is aligned by the front panel side step. Align the top edge of the side panel to the other parts of the top, as this is most critical for the final alignment and appearance.

Prior to removing the cabinet from the bench, I add a pair of what I call "safety blocks" which the cabinet will sit on for further finishing operations when needed, and which will protect the bottom of the cabinet from floors and concrete surfaces.

[BobEllis]

I can almost see a grin through your helmet, Dark Lord. Nice progress.

[sdl2112]

You are really moving along...ET...:T

Your woodworking skills are to be commended...

Did you use the Dewalt miter for the compound bevels top and bottom?

[Evil Twin]

Did you use the Dewalt miter for the compound bevels top and bottom?

That was the original plan considered, but upon further examination and adjustment of the 3D model, the preferred configuration was outside of it's performance envelope. I considered resorting to copying Mr. Marsh's bevel cut fixture, but this lacked a certain hands on elegance...

This left me no choice but to trust in the Force and the power of the Dark Side... and as my light saber of many decades has been acting up of late, this meant adopting more conventional tools and techniques to realize the facets...





With the updated design, the upper facet begins 14 inches down from the top edge, and extends to the cabinet top 1" on each side further out from the base angle bevel; on the bottom side of the cabinet. the facet begins 17" from the bottom. This asymmetry resolves itself in visual symmetry, in that the relative point of origin for the top and bottom facet relative to the top and bottom woofer is then quite similar, in a symmetric fashion.

The way that the facet is cut using the DeWalt DC190B, is to set the saw for a 45 degree cut (this model has a cast base and fairly accurate settings), set the upgraded 60 tooth Diablo 6-1/2" blade to maximum cutting depth, and cut on the line marked on the side, and one the top, per the parameter described above.

Minor smoothing and clean up is accomplished with a course and fine pitch drywall sanding block setup. My experience is that drywall sanding screen is more effective as a flat block sanding material than conventional sandpaper, in the same way that Abranet are generally more effective and clog free.

[Evil Twin]

Logistics update...

Once again the HiFi Collective have come through, for the second capacitor order. If only we had suppliers as reliable and fast as them for building the Battle Stars...



And the remaining four PR223BD02 passive radiators have arrived...

It seems at times like a third of the work of completing a project like this is managing the component logistics....



Design and construction being the other two thirds...

[Carl V]

Most impressive your progress thus far.
Soon you will have Bonnie & Clyde,
The New Ardents and now Kurosawa.

Whoa

[Evil Twin]

The modular nature of the Accuton Cell approach has necessitated a slight naming update- consider that it is possible to build this three way with either the 9 ohm or 4 ohm version of the AS190. In light of this, it seems reasonable that the Kurosawa concept could be broken down into three possible offerings based on the number of woofers...

With three 9 ohm woofers, the system in construction would be known as the Kurosawa Dai Katana.

With two of the 9 ohm woofers in parallel, (or two 4 ohm in series) this would become the Kurosawa Katana.

A three way with a single 4 ohm AS190 and two PR's is also possibe, with a less expensive footprint in most regards (except crossover cost), and this could be a Kurosawa Wazikashi. This was reviewed and approved by the marketing department headed by Steve Manning...

[Evil Twin]

Greetings ET,
Any data on the effect of your damper assembly on the 3rd order distortion artifact at ~3kHz per the Accuton datasheet? Granted, if you clobber the frequency response per Accuton's "filtered" option you also clobber the distortion peak. However, based on diameter, a 90mm driver shouldn't start beaming until ~4.5kHz. So it does not seem unreasonable to expect usable response at 3kHz from such a driver. I want to like the 724, but that distortion! Would be great to learn that the distortion is also caused by tube resonance and can be addressed with a diffraction device. Then again (with no disrespect meant to the spirit of innovation evident in bending these drivers to your will regardless of their faults), why would one simply not bother with all that doubt and use the old stalwart 079...

Technically, the "old stalwart" is discontinued other than meeting OEM commitments; once channel inventory is depleted, it may be in relatively short supply at some point. This is true for many of the older drivers, and some new series, such as the mid bass used in the Avalon Tesseract. This is confirmed in direct meetings...

Now, one can look at this from an individual driver issue, or one can look at it from a system viewpoint, as Accuton's main vendors insisted they do. This is why the cell series was created with a collection of tweeters, midrange, and woofers that all have their acoustic origin in the mounting plane. This has unquestionable advantages in system design and implementation, though the design and manufacturing challenge is considerable.

Additionally, there are the inherent driver trade offs and work arounds. If one doesn't care for the midrange magnet structure T-Line resonance, and is unwilling to address it oneself, then there is the C168, which is actually shallower than the 724 (as I can tell, holding one of each and comparing them). OTOH, the C168 has a break up mode at a lower frequency, and due to the larger diameter cone, inherently has a lower polar response cut off point (to be determined, I expect results similar to the C172-6-90). This likely will point to using a waveguide with a suitable tweeter to extend the lower working point.

One OEM addresses the 724's limitation with a revised motor structure and a diamond deposition dome structure, raising the frequency of the upper resonance and the onset of resonance amplification of distortion products.

For the DIY experimenter with a limited budget, those are not available options- and so things like this have been implemented and tested for SPL response- which is improved.




Full testing, including distortion, will await the assembly of this system.

The most that can be said for the unmodified version of the C90-6-724 is that it has moved the first break up mode up from 7 kHz (C79) to 9 kHz. How that works with the driver and chamber configuration with a specific crossover filter will be the telling point.

[scottvalentin]

ET you had a very busy weekend - this is very inspiring for an apprentice like myself.

[fish fingers]

cant wait for more feedback on the C90 post tweaks, Dark Lord

[Evil Twin]

Cabinet detailing and epoxy coating prior to Gel Coat

Things are moving forward, though not as I have foreseen, as complications with the Empire reborn and with local working conditions seem to conspire to slow down work....

[5th element]

Just seeing all of those driver cut outs...this is going to have serious bass. Bonnie and Clyde be damned these are going to be formidable.

[sdl2112]

Dang Jon...at this rate your going to finish before me...:E

Yes...that wall of passive radiators should be interesting with some material content...It almost looks like it could break Hoffman's law in some way.

Is that wool stuffing in the mid cavity? I am curious how it effects other parameters like the T/S parameters, impedance curve...in addition to SPL and distortion.

[Evil Twin]

The parameters for the midrange driver sub enclosure will be a matter of some investigation and measurement- the current first pass plan is a wrap around liner of the denim based UltraTouch, lining the cylindrical wall, and a central core of long fiber wool. the modified midrange drivers use UltraTouch captured between the built in screen and the stainless steel screen of the sprout cap (this is a cap designed for growing sprouts in canning jars- when the massive manufacturing capabilities of a galactic empire are no longer available, true innovation lies in finding the right configuration of components in common household items... Available from such Holonet vendors like "Amazonian" or some such. Though why it should be named such is not at all clear to me, as the company does not seem to be based in South America.

The sanding and detailing process is close to completion- now, the surface is a matte smooth finish of a translucent film of epoxy actually embedded into the wood surface, which should gives the gel coat a suitable and stable foundation free of enclosure structural "bleed through". We shall see if the practice is as successful as the theory in this matter.


Yes, the passive radiator array on the back would appear to be the product of the febrile imagination from someone completely in love with the construction and sonics of car audio system, but this was a sound engineering calculation and choice, based on evaluating numerous passive radiators and their tuning options; in this case, it is believed that the results will be substantially better than the Accuton PRs, as each driver set (one woofer, two PR's) is in it's own working volume, and with the chosen alignment tuned below the main driver resonance should achieve nearly flat response to 30 Hz, substantially below the AS190 free air Fs of 39 Hz. Overall, the system does seem to push against the boundary of Hoffman's iron law, but consider that it is in actuality rather large. 1.4 meter in height does not sound very tall unless you actually stand next to it, and it is on a furniture dolly. Then, matters are different, unless you are of a larger stature as some of us are...

The fascinating point of this design, is considering the unusual response profile and wide and consistent dispersion up to 3 kHz, this AS190 + PR223BD02 combination should produce the best monopole source from 30 Hz to 1500 Hz which I've ever constructed, and is well suited to the wideband requirements of the Duelund crossover approach.

The other factor in the multiple passive radiator approach is a point often not understood by those not versed in the mechanical arts- a passive radiator can all to easily become "detuned" under long excursion high output conditions- the actual suspension compliance changes with stress on the suspension and the position of the cone, and this changes the actual tuning of the LF system. Using multiple PR's in this configuration makes the low frequency tuning parameters far more stable, and as noted earlier, can produce relatively prodigious output levels with only 30Watt of drive per active driver (90W total) and with the active driver travel at a peak of only 4mm, out of 9mm.

The other aspect of the rear panel PR array is my personal esthetic distaste for drivers mounted on enclosure sides. This should produce a much cleaner appearance from the front, in my opinion, though it makes the overall fabrication a bit of a challenge. It has a clear benefit in that the numerous passive radiator openings give access to the enclosure rear for loading or updating crossovers, another reason to use knife blade stainless steal inserts for the 8-32 stainless steel screws, so that repeated access does not compromise the enclosure integrity.

[Evil Twin]

To clarify the low frequency design parameters, which apparently had not been posted:

The analysis is presented for a single AS190 driver and two PR223BD02 passive radiators, at a 30W drive level; in actually, three times this number of drivers is being used.

[Evil Twin]

Gel Coat applied

This has been another productive day in spite of the variable temperatures- it all becomes a matter of timing, and whether it is best to do something at 9 in the morning or at 11 or 12.


Here you see one of the cabinets after all detail epoxy fill work is complete and sanding is finished- it is wiped clean using mineral spirits prior to gel coat, and allowed to dry.





At this time, two coats of of catalyzing black gel coat have been applied; though it is promoted as complete coverage in one coat with self leveling, this is not found to be a completely realistic promise, especially if using in practical coating depths on vertical surfaces. Regardless, the coverage quality appears to be quite reasonable at this point- how it looks after sanding will be the key point. As I still have enough material on hand to do two more coats, it is not anticipated that this will be a problem. More Mirka disks are on the way from the fast shipping vendor.





The intention is to have the sanding and further detailing complete by next weekend and be ready for clear coat work. How feasible that plan is remains to be seen... as their are no additional minions available here, I remain the only individual that must be properly motivated...

[Evil Twin]

Electrical and logistic work is completed for the first pass crossover design to be used in the first test articles...



This has been replaced by an updated version with modifications to the woofer and midrange networks.


There is much detail work to be done, a satisfactory outcome is far from certain.

Checking on the status of the gel coat work shows mainly very satisfactory results, but one side and one rear panel in need of some additional work, currently scheduled for this evening. The curing over night is quite satisfactory, and bodes well for the ultimate quality of the finish.

A brief test placement of the unfinished flame maple midrange/tweeter panels on the cabinets hinted broadly at how the combination may look in it's finished state- this was in all regards a motivational moment, and gives credence to the original design concept.

Additional finishing and construction supplies have been ordered or procured...

Black is back...

[BobEllis]

Wow! It takes a lot of work to hammer that midrange into submission.

[fish fingers]

This build is amazing, I'm hooked. Everything is so well thought out.

[Face]

Wow! It takes a lot of work to hammer that midrange into submission.

I don't believe it's submission as much as achieving the correct roll off in both directions for Duelund slopes.

[Evil Twin]

I don't believe it's submission as much as achieving the correct roll off in both directions for Duelund slopes.

You have the correct sense of it- with the Duelund transfer function, the exact slope must be managed down to - 20dB or better for good results, and this requires much more "management" of the transfer function over a wide range, than a conventional midrange band pass - if one had a driver with ideal response from 100Hz to 10kHz or higher, than a simple LR2 band pass would roughly suffice, but with real world drivers, this isn't possible.

[sdl2112]

Hello Jon...

I see here and on the revised Wavecore Ardent with the -02 woofers you are choosing a downward tilt in the response curve. Do you still plan to use the voicing circuit also? I read where you were leaning towards designing for flat and then applying the voicing curve. I am interested in your thoughts...

[Evil Twin]

How the voicing will work out in the actual built system is not yet known with certainty- CAD predictions are useful, sometime very accurate, but in the end it is how it measured and subjectively performs in room that counts.



While the simulation predicts an overall slope from 100 Hz to 10 kHz, the voicing curve is not a specifically linear transfer function, and versions have been built both with the 3.2 dB recommended for very wide power response systems and a 1.8 dB version which might be better suited to systems of average power response, such as two ways or three way systems with upper band limits on the midrange driver due to piston size.

Those options will be examined in detail, and it would not surprise me if the 1.8 dB options were best suited to this system.

Realistically, a truly flat response would be difficult to obtain without somewhat undesirable methods, as at this point the midrange and tweeter have no attenuation networks.

[audioslave]

Far be it for me to question the Dark Lord.....but as a long time lurker and builder of a NeoDcc. The acoustic slopes in the simulations look significantly different than that design. Is that just the inherent properties of the drivers or does this represent an evolution in your thinking about the application of that concept? Maybe its just the bump at 1500 Hz.

I know you had mentioned that you had evolved your opinion on the results of that project. For the Ardent and Isiris you moved to a different topology...to apparent great results. Just curious.

I can almost feel the force choke hold beginning....

[Evil Twin]

A Duelund with slight alterations... pray I do not chose to alter it more!

In any specific crossover they may need to some adaptation to the driver characteristics- in the case of the NeoD CC, I believe from examination that the issue is the inherent time offset between the dome midrange and mid woofers. For this project, there is no appreciable time offset, due to the unique driver designs, but bandwidth characteristics and desired crossover points combine with efficiency issues to create a slightly asymmetric version, in which the woofer carries a bit more in towards the center frequency, but the ultimate roll off for the woofer must return to a value which sufficiently suppresses it's upper break up modes.

Still, this is a Duelund, with a slightly different aleph gain- remember that the tweeter and woofer circuits are always ultimately 4th order, though starting out lower, and the midrange approximates an LR2 bandpass with varying peak gain and spread.

Note that the "bump" in the 1500 Hz area has been largely eliminated due to the driver modifications- this is based on older measurement data.






Not considering baffle step loss, the intrinsic sensitivity of the low frequency driver group is on the order of 95 dB for 2.83VRMS; for this reason the 88 dB midrange and the 7140 Be tweeter, who's 20 kHz sensitivity is about 89 dB, set the sensitivity limits for the complete system.

[deewan]

Question, is this build taking place someplace other than your home on Coruscant? Looks like a lot of vegetation and lack of urban buildings for this build to be taking place on Coruscant.

On a serious note, great build. I've been following for a long time and I'm really looking forward to the final results.

[Evil Twin]

Question, is this build taking place someplace other than your home on Coruscant? Looks like a lot of vegetation and lack of urban buildings for this build to be taking place on Coruscant.

Working in exile on a barbaric world on the periphery of the RIM is not without its challenges- the lack of advanced manufacturing or prototyping facilities being merely one of them....

Stay tuned, some new challenges are being dealt with this week, but progress is underway...

[Evil Twin]

Reset and rework...

Issues with the gel coat curing and properties have forced a reset and evaluation, with new base finishing work in progress after stripping off the old... This amounts to about a 1 week setback, but it is believed that the current path will produce the desired results, eventually...

A three layer "diagnostic" primer system is being used, with a base layer which is light gray, a second layer that is dark gray, and a top layer which is, of course, black. This system is quite revealing of any tendency to over sand, and is further diagnostic of the degree to which that has occurred.

One new tool's functionality has been particularly useful for the more delicate parts of this work.

If all goes well, sanding the primary system will be complete today, and the primary paint substrate layer will be applied...

[Evil Twin]

To give it's due...

These fine products from 3M are also playing a key role in detail sanding work- they should become another additional to Hank's list of products to feature on the forum...

[Evil Twin]

Further crossover development...

The high local temperatures have precluded working exclusively on the cabinetry- so at times my focus has shifted to the crossover once more. This has involved emulation evaluations, including both test cabinet sets and headphone analysis of the idealized curves.

As a result, some updates are being implemented, primarily to the woofer low pass and midrange high pass.

The woofer low pass has been modified to provide greater attenuation of the region where driver cone resonances occur- this removes a touch of perceived haze in the midrange performance. The change can be most clearly seen in the electrical filter characteristics in the new version.




The midrange high pass has been slightly modified, changing the net high pass capacitor value from 100uF nominal to 80 uF - this provides slightly improved phase alignment and also flattens the response from 100Hz to 1 kHz, removing the slight downward slope. My listening evaluations find little benefit in having that characteristic present, particularly not if an overall voicing filter is employed, as proposed by Sigfried Linkwitz and others.

[Evil Twin]

And another crossover update...

Further evaluation and an engineering review cycle (to catch errors such as value substitution of available components into the simulation network design which had not been completed) have resulted in another iteration on the network design and an increased level of confidence in the initial results.




Primary areas of update include the value of L1011 (1.5 mH from 1.3 mH), the net total value of C2041 (104uF from 22uF), the value of L2011 (0.33 mH from 0.36 mH), and the addition of a small loading network (43 Ohms) on the tweeter to lower the transition range response by 1 to 1.5 dB. Further more, the tweeter impedance conjugate network damping resistor value has been decreased so that now it also provides additional attenuation and controlled phase shift, further reducing the low frequency loading in the out of band area.

These cumulative changes have also tightened up the phase tracking in transition region response for the drivers, while in several cases increasing out of band roll off.




Any further updates must await complete system measurements.


Cabinet work is proceeding at a pace limited by the coating process available- one more coat of automotive grade primer will be applied, then the final sanding before the base black EM6600 water based acrylic lacquer base coat. This will be applied in multiple coats, allowed to cure, then followed up with multiple clear coats.




Today I am having lunch with the Princess Elaina, so that we may discuss the proper upbringing and education of the new heir to the Dark Side.

[CraigJ]

Today I am having lunch with the Princess Elaina, so that we may discuss the proper upbringing and education of the new heir to the Dark Side.

You mean Steve?



I'm really enjoying this build and it is quickly becoming my favorite. Wondering out loud if there could be somewhat of a more budget build for us Padawans.

[Evil Twin]

I'm really enjoying this build and it is quickly becoming my favorite. Wondering out loud if there could be somewhat of a more budget build for us Padawans.

That depends on what one considers "budget build". As mentioned earlier, the concept is "down scalable" to the Katana (two woofers, 4 PR's) and the Wakizashi (one woofer, 2 PR's). Depending on where one sources the drivers, (off shore recommended, in my experience) this will save anywhere from $1200 to $3600, while still conveying the essential tonality and room interaction.

Another idea for the Wakizashi or even Katana is to step back to an RS52 midrange, and of course, some reduction in tweeter cost is possible, the D3004/6040 being one possible candidate, or the Transducer labs Be tweeter tested earlier...



In one sense, though, three ways are rarely truly "budget" due to the crossover build cost. Using the RS52 would likely help that, as it would likely require less "massage" to enable compliance to the target curve, and it's already been established that the AS190 has excellent dispersion at higher midrange frequencies.




The higher crossover frequency would take some care, to manage a transition without issues from the woofer breakup mode or the resonance amplification of distortion products, but it likely doable.

[Evil Twin]

You mean Steve?

There is another... I mean one not yet born... but already I sense the Force in her. This was even more apparent while in my daughter's presence today. Perhaps she thought to hide her from me, by moving to Southern California, but this ploy has not worked as she expected. In December we should have a better measure of her full potential after the birthing event.



Steve is strong in the Light Side... however, I cannot turn him from his path.