Ardent D (aka Kurosawa Jr) Winter Camp

[5th element]

That looks much better!

[Scottg]

^what he said.

[JonMarsh]

A ridiculous schedule for my service to the Empire (including all this last three day weekend) has slowed efforts substantially... I can but offer this pitiful example of one of the wired up low frequency crossover boards...







It is fortunate that I dispensed of Palpatine many years ago, or I am sure I would be feeling his wrath...

[Bear]

A ridiculous schedule for my service to the Empire (including all this last three day weekend) has slowed efforts substantially... I can but offer this pitiful example of one of the wired up low frequency crossover boards...


[ATTACH=CONFIG]30306[/ATTACH]




It is fortunate that I dispensed of Palpatine many years ago, or I am sure I would be feeling his wrath...

What type of buss wire are you using in the crossover? I have been attempting to avoid using any, but I think that I have met my match.

[Evil Twin]

What type of buss wire are you using in the crossover? I have been attempting to avoid using any, but I think that I have met my match.

I suppose no wire would be the best, but I have rarely found that to be possible.

In this and other recent builds, I use a mix of wire gauges depending on the length of wire run and mechanical support required- so I have spools of AWG12, AWG14, and AWG16.

I expect that for smaller boards one could completely justify just settling on AWG16 or 14.

[Bear]

I suppose no wire would be the best, but I have rarely found that to be possible.

In this and other recent builds, I use a mix of wire gauges depending on the length of wire run and mechanical support required- so I have spools of AWG12, AWG14, and AWG16.

I expect that for smaller boards one could completely justify just settling on AWG16 or 14.

Thank you, your Lordship. Is that a solid conductor? Any brand/source preferences? I have given up using Supra wire near my soldering iron.

[Evil Twin]

Thank you, your Lordship. Is that a solid conductor? Any brand/source preferences? I have given up using Supra wire near my soldering iron.

Usually Alpha wire from a distributor such as Mouser or Digikey. In an ideal world one might use tinned and dipped Litz wire, such as some special OSCO wire I have for high power inductors for power conversion in the reservoir chargers of my Tie Fighter, (using what in local specification would be 105 AWG31 conductors in a bundled Litz matrix).

But at audio frequencies this effort to avoid skin effect is not justifiable, except in some specific cases for inductors acting as series elements for midranges operating at relatively high frequencies, when the secondary roll off effect due to skin effect should be avoided. In this case, foil inductor constructions should be considered... as for a Duelund crossover, where precise control of the roll off characteristic ideally should be maintained to -20dB.


note the construction of the recently completed 1st article tweeter crossover board.





Also, note that this board does not particularly reflect an overall cost optimized build approach, as in some cases more expensive components than necessary were simply used because they were on hand, such as in Fs Zobel impedance control networks. The complexity is due to having an LCR impedance zobel network, an LCR ultrasonic notch filter, and a lower frequency LCR EQ filter, plus having a network concept for which the ultimate slope is 4th order, but only realized over a wide frequency spread with a very controlled increase in attentuation.

Duelund crossovers are not for the faint of heart or weak of pocket book.


In other positive news this week, my local medical specialist was "persuaded" to move me up in his priority queue for the additional bionic implants I have requested, and Friday I had the preliminary medical scans (using the primitive local sector equipment) and we setup the schedule for the rest of the necessary preparations, to culminate in the bionic restoration in mid June - the 16th in the local calendar.

This timing will allow me to complete a cargo run to the outer rim next week - loading a variety of items, including the new SawStop table saw, was nearly finished this weekend, only a couple of very high value items (worth more than the transit shuttle) are pending loading on the departure date of this Thursday.

Matters are proceeding better than I had foreseen before this week... one must be prepared to seize opportunities as they arise...

Next today will be completing the midrange crossover test article construction.

These two boards will be used for other system evaluation that are currently planned, as the transfer function and characteristics will not change for the midrange and tweeter, only for the different woofers.

In preparation are test concepts for a small CC using the Anarchy 558, and a larger CC or tower using the Anarchy 708. All would use the same midrange/HF concept with Duelund transfer function crossover.

The PuriFi woofers are another possibility- though for now, they are planned mainly for the NatlieP Ultra. Still, preparing an equivalent design should be a fairly simple task, and test enclosures are partially constructed. Now it may be a race to complete the heavy work for test setups prior to my bionic update procedure; follow that, I will have several weeks off when lighter work may be accomplished.

After that, cargo runs to the outer rim must re-commence...

[theSven]

I like that expression dueland is not for the faint of heart or week pocket book. But I’m excited to see a CC coming. It all goes back to your earlier post to under promise and over deliver.

[5th element]

I suppose no wire would be the best,

We don't use wires in the empire.

[1Michael]

I can but offer this pitiful example of one of the wired up low frequency crossover boards...

LOL, yes that is absolutely hideous Jon. Why is it that your pitiful examples look better than my best efforts...:cry:

[Bear]

LOL, yes that is absolutely hideous Jon. Why is it that your pitiful examples look better than my best efforts...:cry:

Yeah. That's why I keep asking for tips....

[Evil Twin]

LOL, yes that is absolutely hideous Jon. Why is it that your pitiful examples look better than my best efforts...:cry:

But then you should examine Steve Manning's work- had I the time, and if he had the time, I would hire him to build my crossovers....

[chrisn]

Usually Alpha wire from a distributor such as Mouser or Digikey. In an ideal world one might use tinned and dipped Litz wire, such as some special OSCO wire I have for high power inductors for power conversion in the reservoir chargers of my Tie Fighter, (using what in local specification would be 105 AWG31 conductors in a bundled Litz matrix).

But at audio frequencies this effort to avoid skin effect is not justifiable, except in some specific cases for inductors acting as series elements for midranges operating at relatively high frequencies, when the secondary roll off effect due to skin effect should be avoided. In this case, foil inductor constructions should be considered... as for a Duelund crossover, where precise control of the roll off characteristic ideally should be maintained to -20dB.

your observations on skin effect are interesting, how low in frequency does this start to show up? I have a pair of RS52s, might have to order some litz inductors to compare with traditional variants and measure, anyone know where this "time" might be sourced? :alol::alol:

Good luck with the bionic restoration, I hope the recovery process is quick and stress free!

[Evil Twin]

Some progress resumes...

Some progress on small parts and details has been achieved, and a relatively clear path for the next assembly operations is coming into focus.

• Baseplate blanks have been fabricated, and are ready for rotary milling operations.
• The necessary additional power tools have been ordered and should be delivered by next Tuesday
• New internal bracing stanchions for the rear panel are fabricated
• New rear panel lower brace and insert target have been fabricated
• The offset riser blocks attaching to the base plate have been fabricated and checked for alignment. (these will perform the adjustment of the cabinet tilt, so that the actual front panel tilt will be 5 degrees, half of the cabinet 10 degree design tilt.)
• The design layout for the baseplate attachment using local 10-24 hardware has been completed and all necessary hardware is on hand for this and the footer installation.
• Detail insert installation for drivers has commenced for the 2nd test article cabinet.
• The first midrange test article crossover assembly is nearly complete and should be finished today...

The base plate layout with spacing for the rise and footer assemblies:




A review of the bionics update is scheduled for this coming Wednesday, and initial on site review by mobile personnel indicates completely satisfactory performance and recovery process is established.

All in all, this is acceptable progress for the time being...

[Evil Twin]

Midrange crossover TA1 completed

As foreseen, midrange crossover is now wired up.

NOTE: This board has one wiring error- do not use it as an assembly guide. Remember this is a work in progress- issue found with testing on July 7





Not the prettiest job, and it could be much more compact and less expensive than shown here. I anticipate a combined midrange/tweeter crossover PCB in the future, should this design study turn out as hoped.

Input is on the righthand side.

[Evil Twin]

Pictures, not words.

It is said that a picture is worth a thousand words- how convenient to avoid writing, unless a true build guide is necessary in the future.


Milcoast adhesive backed paper was used to draw the layout and drilling guide on the baseplate pieces.





This shows one of the baseplates after drilling and countersink operations were completed, and the adhesive paper was removed using a Veritas flush plane.





Here, we see the baseplates glued to the riser blocks (using marine epoxy- which has very high adhesive properties with a wide range of materials) fabricated from the left over piano pin block pieces- the purpose is to introduce a 5 degree tile in the base, so that the front panel angle is reduced from 10 degrees from the vertical to 5 degrees.




The purpose is to explore the effect functionally and esthetically, as employed by Vimberg on their two floor standing models.



The outrigger bars will be attached to the riser pieces as outlined in the drawing above.

[Evil Twin]

More images

More images...


The Makita LXT drill was acquired as a new tool for this project due to it's compact dimensions and high torque and adjustable clutch design.




This view is of the back panel, showing the installation of 10-24 inserts for one of the passive radiator holes.

It should be noted that after the initial insert hole was drilled with the EZ-LOK tool set, a 1/32" larger bit was used to increase the dimensioning of the holes slightly more, to realistically accommodate the inserts working with dense wood materials compared with the normal "target" materials.




This view shows a baseplate assembly taped up to the bottom of one of the cabinets, in preparation for creating marking "dimples" using a 1/4" bit. The pass holes for the baseplate are #12 holes/1/4" in order to accommodate potential "imprecision" in the final location of the holes for assembly. With CNC instead of manual operations, this would not be necessary or advised.




This shows the bottom of the cabinet after installing the insert holes have been drilled.

[Evil Twin]

This, too...


Last, we have the base of the cabinet with the inserts installed. The Makita drill's controllability and accuracy and torque almost made Force Sense abilities of no great import...

It should be noted that the inside of each hole was "painted" with a low drip epoxy prior to installing the inserts.

[Scottg]

..The Makita drill's controllability and accuracy and torque almost made Force Sense abilities of no great import...

Additions for screws:
https://www.homedepot.com/p/Makita-I...5097/203696915
https://www.homedepot.com/p/Makita-I...5383/305513416

(..I use this with my Worx "Switchdriver" - makes a one-tool one-person job far easier with pre-drill and screw. I do wish the Worx was of the "impact" variety though, but wouldn't give-up the single-tool capability of the Switchdriver for it.)

[Evil Twin]

Intermediate testing...

Thank you for the links- however, the Imperial budget auditor may not thank your contributions as much as I do!


Now, I will reveal how messy a "Design Study" is, compared with a final stage only reveal of a new DIY project speaker...


Yesterday's investigation was checking the newly built midrange and tweeter crossovers with drivers in one of the test article cabinets... sometimes testing is completed flawlessly, other times, certain "discrepancies" are discovered. The latter was the case yesterday, though they were quickly identified and resolved.


One of these is with one wiring error, one is with the wiring error corrected, for the midrange. I leave it to the readers to figure out which is which, based on the posted simulations.






Now, the tweeter crossover was built correctly, but the "improved" notch filter, using the classic configuration of parallel RLC, seems to have more sensitivity to tuning and Q than that I used for the Calliope design.




Otherwise, the profile is exactly as I have foreseen... new parts were ordered, simulations rechecked, and the notch filters should be rebuilt by the end of the week.




The combined response meets expectations, aligning well with the VituixCAD simulations.





With the circuit updates in progress (2nd day air still works, some of the time) I expect these boards and the build for the 2nd pair will be ready soon- but both will be tested before installation.

[Evil Twin]

Crossover boards- midrange corrected, tweeter to be updated.

It should be noted that while Steen Duelund was very dedicated to the crossover concept he pioneered and it's acoustic results, he acknowledged that it was complicated and expensive to build with real world drivers, to get the necessary SPL and phase control over a wide range.

[Norm]

Holy crap! Above is why one does not find the Duelund crossover in too many commercial designs. I know of only Gryphon who can suffer that level of hit on the bill of materials cost times the markup.

[JonMarsh]

Well, this set is done with parts either on hand or readily available. Check out the NeoD CC design I did years ago- doesn’t look nearly as scary, and the focus was on HT. OTOH, it doesn’t have the same precision or some of the upper breakup mode filters or Fs Zobels. This is more of a perfectionist version. Steve asked me to come up with an HT version, so to speak, and that is in the works. Fewer parts, smaller less expensive parts.

You don’t want to look at the crossovers for Tidal, Vimberg, or Wilson, yah know... can take your breath away!

I’ve heard a rumor that test cabinets are in process for a three way CC based on a modified version of the upper driver crossover and the same mid, but a different tweeter... ET tends to be closed mouth about that kind of stuff until he’s got a POC and some test results.

[Scottg]

You don’t want to look at the crossovers for Tidal, Vimberg, or Wilson, yah know... can take your breath away!

Peak Consult (..signal is all Dueland w/ a LOT of cap's and pricing on that is crazy, Jantzen for correction).

When they started selling loudspeakers, the price of the speakers seemed really high, but I don't think anyone had any idea just how expensive the crossovers were in any of their speakers.

[Steve Manning]

This ought to put things into perspective ..... here's a couple of pictures of Tidal's crossovers.

[Scottg]

Yup, they use a lot more Mundorf than Dueland.



If anything, Dueland's resistors are actually a pretty good deal as far as value vs. subjective result in a signal-position.

I've found that resistors are typically the worst offender to good sound (in a passive component context for the crossover). If I can get my resistance from a good inductor for the network, then it's by FAR the better option.

Note Tony's comment here on his latest project "Nova Andromeda":

http://www.humblehomemadehifi.com/

[dar47]

Nice!, is that built in switch in the crossover for the voicing?

[Evil Twin]

Nice!, is that built in switch in the crossover for the voicing?

Ah, this one is strong in the Force! That is exactly what it is, switching in and out changes in contour of the high frequency voicing. For a Force adept, it can be activated outside the speaker cabinet at will. For those of more limited capabilities, removing a PR and flipping the switch will be necessary...


I am in the middle of updating the tweeter ultrasonic filter, as the new components arrived today.

[Evil Twin]

I struggle to come up with an adequate term to describe the Tidal crossover construction, that doesn't indulge in sexist terminology- it goes well beyond the concept of what in one day might have been called "manly" construction.

I believe that if I proposed something similar to Steve, it would be grounds for dissolving my association with his efforts... He was just recently questioning my fondness for the 20W Mundorf resistors. And I do agree, Duelund are good value for for the price.

[Steve Manning]

I struggle to come up with an adequate term to describe the Tidal crossover construction, that doesn't indulge in sexist terminology- it goes well beyond the concept of what in one day might have been called "manly" construction.

I believe that if I proposed something similar to Steve, it would be grounds for dissolving my association with his efforts... He was just recently questioning my fondness for the 20W Mundorf resistors. And I do agree, Duelund are good value for for the price.

Hey, if the customer wants it I'm all in ..... we'll just have to provide a refrigerator dolly to move the crossovers around and a cardiac defibrillator when they find out what it's going to cost them.

[Evil Twin]

Hey, if the customer wants it I'm all in ..... we'll just have to provide a refrigerator dolly to move the crossovers around and a cardiac defibrillator when they find out what it's going to cost them.

There is that issue... and potential medical liability problems, too.

This may require careful consideration...

Meanwhile, assembly of more crossover test boards continues...


I have an appointment tomorrow with Imperial specialists who are promising to improve my mobility and agility for any future light saber battles... pray the are able to delivery on their optimistic promises.


Meanwhile, another specialist has confirmed that black is back, and a custom order configuration with triple pickups is possible, though the delivery date will be after I should have completed my rehabilitation from the recent bionic upgrade. Still, it is something to look forward to in these difficult times...

[Evil Twin]

The 2nd tweeter crossover board is built-

This one currently has a 29uH ultrasonic inductor, and I believe it will benefit from further tuning adjustments. However, the baseline performance seems quite satisfactory...



An additional sequence of values will be investigated tomorrow before finalizing that aspect of the filter.

[Bear]

This one currently has a 29uH ultrasonic inductor, and I believe it will benefit from further tuning adjustments. However, the baseline performance seems quite satisfactory...

I've played around with this in the center channel I'm working on. It seems to provide a relatively painless way to implement the downward tilt that is being discussed around here.

[Evil Twin]

the benefit of tight tolerance crossover components

One can see advantages to selecting components with tight tolerances- 2% capacitors and resistors being common in this design study test build.


Here is an example, with the two midrange crossover boards tested and compared- naturally, without otherwise altering the test setup, including gains and microphone position, and with the same tweeter test crossover. Literally, only the midrange crossover board was touched, swapped out for the two measurements.





I find the result to be most satisfactory... one might even say somewhat surprising.

[Evil Twin]

There have been a few questions directed my way, about what I mean by "Design Study" (a term used specifically for several threads I have posted in the past) and why this isn't finished yet or posted as a finished design with all details available.


The more common method for project development and posting is to work out all the details, solve the problems as they are encountered, and then, and only then, post a project summary with all the characteristics and construction plans, and a summary of the measured results and any relevant subjective listening sessions and comments.


When I post a Design Study, matters are handled quite differently and there is no assurance of a successful or complete outcome- or that there even will be a "final solution". The difference is that from basically the start, in a Design Study I am sharing the development process and issues, as well as different considerations to weigh and balance, and how this evolves into a first prototype. In a sense, this is sharing the "behind the scenes" work which is going on in the development process- hence, the name, "Design Study".

A classic example might be the "Three Way Evil Design Study" started in 2006, eventually resulting in the start of construction of a final test system in 2012, with first crossover prototypes in early 2013, with the first "production build" around April 2013.

https://www.htguide.com/forum/showthread.php?23683-Three-Way-Evil-Design-Study&highlight=Design+Study

A 3rd generation was in progress in early 2014,






and in 2017, using VituixCAD and a new round of measurements, a 4th generation crossover was planned and components purchased- however, based on experience in the last 12 months, I expect to update some components in that build. Both the cabinets and all crossover builds and components are in storage in a secret location in the Outer Rim...

So, one might legitimately complain that this Design Study is not yet complete...

But think of it as like life's journey... constantly evolving. Until it isn't.


However, this new design study should be completed in "record time" (within a few more weeks) compared with most of the past one, in spite of the many issues and difficulties on this primitive world...

[theSven]

I still have an iMac of that model in use. It's super slow, but for my son to use for Internet it works great. Great news that this is progressing and should be a solid build plan. It has been a great learning thread with so much stuff I still don't get and spend time googling away to lookup. Looking forward to being able to get a pair of these built.

[Scottg]

-Don't know if the Empire decided to incorporate these in any design, but $26 is certainly in Rebel-scum territory!

https://www.parts-express.com/fountek-fw135h-5-1-4-aluminum-cone-midrange-driver--296-735

[Evil Twin]

Curious you should mention these... for a non-Duelund “budget build” these should be very effective- I will post data this week some time.

this particular Design Study by virtue of using a Duelund Crossover approach has some strict gating issues- with regard to the effective Z axis time origin.

In another sense, this might be looked from a different viewpoint, as “What would happen if the original Ardent and the NeoD CC has offspring, and were bequeathed a larger than expected woofer budget? The woofer part of the design is not budget oriented at all. If there were a better dome midrange at any price than the RS25AN, I might have used it- but haven’t found it. Based on measurements as well as listening tests, I would say that the DA25 out performs the SS 6640 in essentially all regards, though for reaching the optimal distortion profile it needs a small ultrasonic filter...

Incidentally, for those interested, the center frequency is ~1650 Hz and the aleph coefficient is 2x square root of 2, or 2.828. These are the only data points needed to define the crossover target from the Duelund equations.

So, that is to say, this was not necessarily started as a “budget limited” Duelund design, but the only other midrange driver seriously considered was the Accuton C51, and it was ruled out due to the SPL flatness, or lack thereof, and the distortion in the 1kHz to 2kHz area. Needless to say, this is a rather curious state of affairs where two relatively “Budget” drivers are, for this particular application target, the “last men standing”.


If one wished to move the center frequency down substantially, then the C90-6-724 could have been considered, but it takes more crossover work, modification to the immediate driver damping, and doesn’t have the off axis extension preferred for a reasonable center frequency. Other cone midrange have too much delay on the Z axis time origin. It needs to be coincident with the mounting plane, which is the voice coil plane. This is why the cell woofers are used, as they behave as if this were the case, through some unusual construction and response engineering. The cone is driven at the center front from a very stiff titanium voice coil assembly, and the response appears to be engineered to product a flat response for the “correct” baffle with without traditionally baffle step compensation.

This approach is used in the Kurosawa, with a $500 SS 7140 tweeter. This design study uses a center frequency of 850 Hz, and while most crossover components have been bought for an initial trial, some further testing and measurements will be conducted in the Outer Rim prior to construction.

Again, this is an experiment... in drivers, cabinet construction, crossover design and execution, and modified form factor. I am also curious how it will work out subjectively and measurement wise.

[Scottg]

..as “What would happen if the original Ardent and the NeoD CC has offspring.."

Oh my, I didn't realize the Empire was into that sort of research.. 8O

(..in reference to my "laughed" above.)

[Scottg]

..Other cone midrange have too much delay on the Z axis time origin. It needs to be coincident with the mounting plane, which is the voice coil plane. This is why the cell woofers are used, as they behave as if this were the case, through some unusual construction and response engineering..

..and far more "perverse":

Have you experimented with inverting (rear-forward) midbass drivers to move acoustic center forward? (..something Wilson Benesch does.)

[Evil Twin]

This approach is less ugly, and has been used in the past. I have first hand experience with this system, and have built a modified design using the same drivers.

What I am finding is that their are other benefits to "dome woofers" besides the time offset issue...





And in case your memory/timeline doesn't stretch this far back, that is a B&W DM6, sold in the 70's.

[Steve Manning]

This approach is less ugly, and has been used in the past. I have first hand experience with this system, and have built a modified design using the same drivers.

What I am finding is that their are other benefits to "dome woofers" besides the time offset issue...


[ATTACH=CONFIG]30435[/ATTACH]


And in case your memory/timeline doesn't stretch this far back, that is a B&W DM6, sold in the 70's.

Of course one has to remeber the DM6 started life as a GNK series power droid .... thus near and dear to ET.

[Scottg]

This approach is less ugly...

I'm going to have to disagree with that.



(..I've experimented with the rear of the driver "outward", but for a radial/"omni" pattern (up-firing) - but it results in a lot of pressure-loss.)


Either way, I'd bet those Accuton's sound excellent for a number of reasons.. but the cost is well, IMPERIAL. :W

[Evil Twin]

Either way, I'd bet those Accuton's sound excellent for a number of reasons.. but the cost is well, IMPERIAL. :W

Your points are well taken....

On the other hand, the cost of the Accuton AS168 is far less than, say, a Vimberg or Tidal speaker system using them... $31,000 per pair, according to the recent Stereophile review. Yet, this probably represents good value, compared with, for example, an Avalon Indra.


I also found it most interesting to parse that review in comparison to other public and private data...


One interesting case is the impedance curve of the Mino...





Now, to those skilled in the arts, there are certain impedance signatures for specific crossover topologies that are often recognizable.


Take this example, for the Magico M2, which uses elliptical augmented fourth order networks...





If you compare the general shape and magnitude versus frequency, one might notice some resemblance to the Avalon Indra, including the impedance behavior in the mid bass and the location of other peaks and dips.


On the other hand, the impedance of the Mino is quite familiar to me in concept, because of it's strong resemblance to the simulated impedance curve of the Ardent D, but with the peaks and minimums at different frequencies, reflecting the key difference in midrange driver usage.

Another interesting comparison might be made, examining the step response.

The Magico M2 has a rather "classic" profile observed when using high order crossovers with all drivers in the same relative (positive) polarity.




The Mino looks quite different, with the midrange out of phase, as required with a Duelund crossover.





Interestingly, JA's comment indicates he may understand just what it is he's likely looking at, but he does not spill the beans...

Turning to the time domain, the Mino's step response on the tweeter axis (fig.8) indicates that the tweeter and woofers are connected in positive acoustic polarity, the midrange unit in inverted polarity. However, the decay of the tweeter's step smoothly blends with the negative-going start of the midrange unit's step, the decay of which blends smoothly with the positive-going start of the woofers' step. This suggests optimal crossover implementation.

Optimal crossover implementation? For an out of phase wired midrange? Might that be the condition when all of the drivers are in the same relative phase to each other at most relevant frequencies of overlap?


Not at all surprisingly, this is basically what the simulation of the Ardent D looks like.

Based on the crossover correlation tests so far, I expect the completed speaker and the simulations to be in reasonable agreement.

Is there an audible difference or benefit? That remains to be seen....

[kevinlin1013]

Are the drivers DA25/RS52AN/AS168 already final choices or still surveying?