Saint-Saëns planar- AKA SMJ-40- time for a project now?

[JonMarsh]

The last couple of weeks have been very busy with some non-audio related stuff, sort of like training assistance for former colleagues, culminating in their attendance at the IEEE APEC conference in Long Beach this week. It was sort of fun to review stuff, create sanitized and redacted presentations to loan out, and also fully back up my main former work computer, my own iMac Pro. It's great that all the weird software I ended up buying myself due to company budgets is on a reliable working computer and I can take things back up on a moments notice.

BUT!!!

Now I'm taking the SMJ-40 Saint-Saëns back up, and not on a moment's notice. And I need to ship out to Steve the Seiji Ozawa recording of the Cminor Symphony No. 3 that I got from a vendor in Germany, so that he'll have it to enjoy once this project reaches it's logical or illogical conclusion! For now, I can say it's getting closer, and I'm reporting on the most recent work, and showing a little more about how the sausage is being made.

It has been a long awaited moment, to being ready to start loading one of the MF-HF cabinets and be able to test in consideration of the cabinet acoustics.



There are some key bits to have on hand, having already mounted binding posts for the midrange sub enclosure.




This is all for assembling the cabling to connect to the midrange driver- terminal lugs matching the binding posts in the mid subenclosure, and a ratcheting wrench for tightening those nuts down with limited working angle through the front.

This is a picture of the midrange wiring, before having stripped the full outer layer of insulation off the Cardas 101, in order to have better flexibility when positioning the driver. Your choice whether to use crimp on female spade terminals or to solder to the table on the PT-5010; I chose to do the latter.




Very recommended tools for this process are the adjustable spring loaded automatic center punch, and that 13mm ratcheting wrench for the binding post nuts.

[JonMarsh]

One of the useful things for the automatic center punch is to use the drivers to locate the mounting screw holes, then center punch the locations for reference, to facilitate drilling pilot holes for the midrange driver screws (supplied with the PT5010 mounting kits) or for the Rampa threaded inserts I used for the M4 mounting screws for the TPL75 tweeter.



The Rampa inserts are very well engineered, but I did clean out the distributor (temporarily?) when I bought mind.

Here's a view of the midrange cavity, stuffed with Acousta Stuff, and with the wiring connected and ready to go. Pilot holes were drilled after locating them with the automatic center punch with the driver laid in the front panel first.

One thing I forget to think about is the extra space my side wall and back wall damper pads take; this did reduce the effective volume the midrange panels have and changed the tuning somewhat, as we'll see in a bit. Nothing that can't be dealt with...







Now, as I expected, with the baffle design and rear setback, there was some interaction with the front wave of the midrange compared with a flat front baffle. This produces a large dip at about 2.2 kHz if nothing is done about it.

The solution is fairly straightforward, using 1" thick adhesive backed acoustic foam, to cut out pieces sized so that the active channel slots of the midrange are clear, and the remaining space is filled with the acoustic foam.



With an "optional" grille cloth panel in place, the esthetically fastidious should have little to complain about.

😉

[JonMarsh]

Measurements with the slightly altered midrange volume did show some change in the response profile between 400Hz and 1500Hz, so some tweaks to the crossover will be needed, and parts have been ordered, including some HF stuff for finalizing that aspect with a more complete measurements in this cabinet. Parts scheduled to be here by Friday.





A VituixCAD simulation of the updated configuration:



Of course, trust but verify is the name of the game. With this cabinet as is by itself, there are diffraction issues for the tweeter which should go away when stacked with the woofer cabinet.

[technodanvan]

Looking pretty nice Jon!

[JonMarsh]

It's getting there... like that old three stooges routine, "step by step... slowly I turn..."
Nice deep crossover point notches still for inverted midrange polarity...

[JonMarsh]

Definitely in the mode of "slow work takes still more time", I decided not to modify the old midrange and tweeter crossover boards that worked great with the flat baffles, but keep them as a reference, and built a new set with the design for the new measurements. Tedious, but probably wise in the long run.

And though I got in nearly $500 worth of parts on Friday, this is still a "junk box" crossover and I had to use stuff from my original collection plus raiding some other parts bins, and in one case a built but set aside woofer crossover board for the Isiris!


Once I finish breakfast and all the Sunday morning communiques, including with the Chancellor, I'll get right back to this....





Some value changes and some re-arrangment of parts, compared with the first POC build. Cross fingers...

[theSven]

Definitely in the mode of "slow work takes still more time", I decided not to modify the old midrange and tweeter crossover boards that worked great with the flat baffles, but keep them as a reference, and built a new set with the design for the new measurements. Tedious, but probably wise in the long run.

And though I got in nearly $500 worth of parts on Friday, this is still a "junk box" crossover and I had to use stuff from my original collection plus raiding some other parts bins, and in one case a built but set aside woofer crossover board for the Isiris!


Once I finish breakfast and all the Sunday morning communiques, including with the Chancellor, I'll get right back to this....





Some value changes and some re-arrangment of parts, compared with the first POC build. Cross fingers...

I bet the Chancellor is going to be pleased with this progress. The detail and time going into this design is impressive! The time management skills deserve a medal for how well you manage time to do all the things you do.

[JonMarsh]

I will be quite happy if the Chancellor is pleased...

This last week has been unusual due to the volume of communiqués and real time interaction on line with a larger than average number of people- it has had an impact on productivity when I get up at 5 and find myself not having started "productive" work until after 10:30! I hope to do better, today, or I fear that ET may decide it is time to sit me down for a bit of a "heart to heart"!




So, while the quality of assembly and wiring completed yesterday is fairly satisfactory, the quantity fell short of plans and intentions... and I'm updating the POC assembly drawing as I go, doing that interactively with the assembly this time.

[JonMarsh]

Making progress grinding the sausage, have finished construction of the updated MF PT5010 POC design, will start the TPL75 build later today.

Still kept the board fairly small, but it's a bit cleaner build, and less expensive, too! (usually they go in the other direction...)





I did some quick checks, and things seem on target for just the crossover and also as expected with the older TPL75 crossover and it's higher tweeter level:





The phase is fine but the levels don't match, as expected- reverse connection produced a nice deep null.


Distortion check looks fine, too.





There are some design & component tweaks in the new version for the next POC build, but mostly it's just a level adjustment and updates to the impedance flattening LCR shunt network.

No surprises for the distortion results - still in the -70 dB range for HD3.

I'm hoping for faster progress now, but the weather has been a problem for re-arranging the garage and moving the bikes out to storage in the back- need to setup for doing the LF cabinet final tooling and assembly work.

[JonMarsh]

New TPL75 POC crossover board finished layout, starting construction.






Some topology and many component tweaks.

[JonMarsh]

Oops! Already spotted a missing connection- but need to go get wife's car serviced!

[JonMarsh]

Updated this morning... added L30 connection to input ground.

[JonMarsh]

Making progress, finished this MkII TPL75 POC test board. Wife is going out to meet up with a bunch of her Russian friends for a deferred International Women's Day soirée, so I can make lots of "testing noise" this afternoon!




We'll see how this turns out! Will likely need some front panel tuning also.

[theSven]

Nothing like an open day to make some progress. Bring on the noise!!!

[JonMarsh]

Schweeep! Schweeep! Schweeep!

[JonMarsh]

Well, I patched up a little more foam on the front panel....




If you think you are picking up on me aiming this cabinet up towards the ceiling, give yourself a gold star, it's the easy way to test for just the MF_HF enclosure, and with a high cathedral ceiling, measurements at 1m are no problem.


Sans having the baffle effect of the LF cabinet in place, I'm not going all ultra tune on the tweeter response yet, but things look pretty much as expected...




Of course a critical (to me) thing to check is the reverse connection null at the midrange-tweeter crossover point...





The -20 dB notch is at the same level as the VituixCAD simulation, so I'm probably doing something right, even if only accidentally.

Who says luck isn't as good as skill?


Note, the HVAC was running while I was testing, needed to get this done and move on to some other tasks - so the background noise floor for checking distortion is compromised, though in a predictable manner...





Nothing to complain about for now, so time to move on to moving the bikes out of the garage and setting up for finishing the piece part work on the LF cabinets and getting them put together!

[JonMarsh]

This is a bit of an explanatory follow up to the last post, to explain a bit the differences in TPL75 crossover that have been under consideration due to the differences in measured data for the TPL75 depending on whether measured in the POC cabinet in its current configuration or on the original test baffle.

As a refresher, the original test baffle configuration:



If you plug in the data as measured on the MF-HF cabinet to a POC crossover network developed for the free baffle response, this is what you see, due to differences in response arising from the baffle:




The rise in the 6-8K region is due to differences in baffle response, and because the original basic test baffle had a dip there, a small lift was engineered into the original crossover.

With that lift removed, it looks like this:




which is still not where I'd want it to be, of course. This is what you see in the last measurements taken.

Reconfiguring the crossover slightly, and changing the LCR network into a parallel series band reject filter, this result is easily achieved:




However, until I see what the baffle behavior is with all cabinets assembled and stacked, no hardware updates will be made. For now, I'd say matters are in hand, pending seeing what things look like with the complete cabinet stack up.

[JonMarsh]

Well, not much time for audio today, but after lunch I pondered a little, tweaked a little, and ordered parts a little (small order, a few caps and resistors).

I don't know why I do this to myself, things may need to change after a complete tower is assembled to test, but probably I'll sleep a little better tonight after placing the order today with PE. Guess that's how I'm wired!

😏

[technodanvan]

What's going on with the bump in response around 20kHz?

[JonMarsh]

Diffraction and no diffraction control on FP, yet- much smaller bump on the original TPL75 measurements

[letiennam]

I have never heard a plana driver before. If possible, I hope you can share this project in near future. and my next thing is to gradually purchase the GRS PT5010 and PT2522c

[JonMarsh]

This is my priority project at this time- others are sidelined. Next steps are final piece part fabrication of the LF module parts, then assembly.

Note, this version is using Beyma TPL75 AMT tweeter. I do have an alternative planned for folks like technodanvan, with a GRS 2522c, in a rather specific loading configuration with an expanded rear sub enclosure for the tweeter. With some effort, it can be made to work fairly decently, better it seems than the HiFi News data for the FR30 speaker system.

This is the measured performance of the most recent MOD3+ update as I'm calling it now. I plan to also assemble some with Acousta stuff instead of the AST Denim damping, to see if there are any significant differences.

[JonMarsh]

There is also a rear firing tweeter setup in development, using a GRS ribbon tweeter. RT1.0-8.

[JonMarsh]

I have never heard a plana driver before. If possible, I hope you can share this project in near future. and my next thing is to gradually purchase the GRS PT5010 and PT2522c

well, I have familiarity with push-pull planar drivers from way, way back- was selling Magnaplanr speakers at the sound shop I was a partner in- they had push-pull mid-tweeters, but usually single ended woofers (magnets only on one side) depending on the size and price. Made in Minnesota.

https://magnepan.com

[JonMarsh]

My only other suggestion is to be sure to review the earlier posts in this thread. I didn't go much into the theory or construction of push-pull magnetic planar film drivers, because there is so much on the internet already on that topic. What I've tried to highlight is how sensitive they are to rear loading or no rear loading, and how this has a bigger impact on response characteristics than one is necessarily used to seeing with cone drivers, because of the differences in the relationships of the "piston" mass, compliance, resonant frequency, and Q.

Sometimes I forget that most folks won't have the background experience I have selling, using, and measuring Magnepans and the larger Magneplanar speakers made by Magnepan and sold through Audio Research back in the day.

Next things I'm looking closely at is an update to the GRS PT2522C Mod3+ crossover configuration and integration with the measured midrange results in the POC build cabinets. I have the impression that most interested builders want to save money and go with that solution, even though the TPL75 has lower distortion.

[letiennam]

Thank you for your explanation. I learned more about plana speakers. You were right. I am also one of the interested builders who will implement GRS PT2522C Mod3+ solution ........to save money. I need to keep a good relationship with my wife to get the next DIY speakers ))

[JonMarsh]

On this topic, the GRS 2522c tweeters, I did just receive another pair, which will be used in a modified test build with new rear enclosures, this time with Acousta Stuff. Steve has suggested I should do a side hustle building up the modified PT2522c with the special sub enclosures for those interested in pursuing this design, and interested in a pre-built version with documented test results. Seems like a reasonable idea to consider. That means a build like this, with a high density phenolic wall sub enclosure, capped with the original 2522C rear wall.



This is a busy time of year for all kinds of house and financial stuff, seems like in the last two weeks I've had hardly anytime for music and speakers! And spending just an hour here, or an hour there, doesn't create the "immersion stream" of effort and focus that I know from experience results in the best quality of work! So yesterday, with one brief exception, I spent the whole day reviewing the MF-HF cabinet crossover work, going back and reestablishing test signals, signal levels and minimum phase exports, etc, all the way through the current proposed networks, and did some detail tweaks on that for the mid network and the LPAD configurations.





If all goes as planned today, I will finish assembling the auxiliary LCR network for the midrange crossover, reattach connector leads, and test and (if necessary debug) once again for the MF+HF cabinet.

Meanwhile, I'm working on clearing out space in the workshop portion of my garage so that I can setup for the LF cabinet construction- many piece parts are complete (front panels, for example) and basic cuts are done, but there's a lot of detail stuff left before diving into assembly.
















Now, this is DIY, so of course, other's may chose to be a bit less "elaborate" in the enclosure fabrication, perhaps omitting the damping panels similar to those used in the midrange sub enclsoure- those are choices up to you. This is DIY. you know who you are... there are several I know of to date who have bought the drivers.


When this design and build is a wrap, I can provide STEP files and PDF dimensional prints. Or, you can visit Steve at SMJ if fabrication assistance seems like a good idea... he'll get all the data too, and who knows, he may have some clever ideas up his sleeve- remember, he is the ex machinist from US submarines, whereas I'm just a wires and sparks guy of sorts, an out of work Rock and roll musician from East Texas and Colorado.

[JonMarsh]

I thought I was all set to do measurements yesterday on the updated crossover boards, but the old MacBook Pro being dedicated to the Fuzzmeasure test setup decided to die slow in front of me, no longer outputting power on the USBC connectors and not charging the battery. SO, it could still run after a fashion, but not operate the mouse or the MOTUS audio interface!

Well, my travel MacBook, a 13" M1 MacBook Pro is being configured now for this, but I'm awaiting a mini-doc of sorts to setup the necessary connectivity, plus a new bamboo board for the cart top so I can add an onboard display.

You never know what you're going to wake up to these days... even on a Sunday morning!

In a pinch I could have gotten out my other test cart with Pocket CLIO; it has some good points, but doesn't measure above 20kHz, so is not very useful for HD measurements or detail analysis of tweeters, IMO.

[technodanvan]

How old can it be if it has USB-C ports??? Sorry to hear either way, such is the way of sensitive electronics in a compact form factor.

This is DIY. you know who you are... there are several I know of to date who have bought the drivers.

Yes, yes...I should start planning my own at some point soon. I think I'll opt to use the less elaborate cabinet without all the extra damping, but we shall see. I'd like this one to be maximally cost-effective. I also intend to use the Dayton passive radiators I have in storage. I'm unsure I want to tackle the baffles on the mid/hi enclosure though, I may need to enlist the support of a certain forum member.

[theSven]

I know I will be getting help for this cabinet build from that certain forum member... I would say he is a critical part of my DIY builds for me to get a build completed.

[JonMarsh]

How about 8 years old?

That's what's a bit odd- I have older Mac Pros running strong, but then they don't have batteries and battery power management. I think it's something that failed inside the main power management in the laptop.

However, with a Thunderbolt to A/V adapter, my travel Mac, a 2020 M1 13" MacBook seems quite up to the task at hand, so it will become a permanent part of this test cart, after stealing a larger monitor from one of my older Mac Pro's. Gotta tie wrap the wiring back down now...

[JonMarsh]

Yesterday was filled with shopping errands and other tasks, but this morning I setup and got a few measurements done, basically confirming the work in VituixCAD.

The test setup was the previous "fast and easy" way, with the MF-HF module on its back and pointed straight up, with measurement mic a meter away. Fortunately I have high ceilings, however, not high enough to avoid quite a few reflections!

These are what I refer to for revision tracking as V20.4 MF-HF POC set...




This includes an impedance leveling LCR set, which is actually on the Tweeter board, and a new LCR feedforward network on the midrange, which had to be implemented externally- will need a larger build layout in the future, or a similar add on board! Since I already have base boards cut...


Combined SPL response centered between Mid and Tweeter- with extended distance, the vertical window becomes wider and wider. Primary axis is on tweeter, which will be at seated listening height.




All things considered with the measuring setup and not the final baffle configuration with woofer cabinet, pretty satisfactory.

This is what I mean by quite a few reflections:




And a test with HD2 + HD3 distortion:




Standard nominal driver level, 2.83VRMS, measured with storage scope.

Time to get to emptying out the garage (motorcycles) and making more working room for the LF enclosure fabrication work!

[JonMarsh]

Now that the sun is higher in the sky and not glaring on the MF-LF cabinet, this picture better illustrates the test setup.



The whole time of day thing here is sort of weird- geographically, we should be in Pacific Time Zone, like the northern part of Idaho, but for political/economic reasons, we're in Mountain Time zone, which extends all the way east to Colorado. As a result, right now sunrise is about at 7:30 AM in the morning, so it's barely over the horizon at 9, and kids are going to school in the dark. Dumb, if you ask me... but who's asking?

[JonMarsh]

Another experiment to report on, the PT-2522c 300+ MOD, as I call it, using Acousta Stuff for internal damping instead of Denim damping. This build is about 1/8" deeper also.





So, first a 0 - 50 degree polar spread, which can be compared with that for the Denim version posted above.


A DA25TX it isn't, but what I did notice immediately was that with this damping configuration, the bump/dip action in the 6kHz - 7kHz is considerably dampened. This was the case with both samples, which measured similarly.

Some other plots possibly of interest:

An RMS Average of 50 - 0 - 50 degrees, 10 degree data increments (an estimation of average power response over this horizontal arc of 100 degrees):


RMS Average of 10-30 degrees:





I like the "taming" of the 6kHz-7kHz stuff, so I'm going to create some FRD files and export and convert into minimum phase files in VituixCAD, see what we can do, probably focusing on the 30kHz data initially:

[JonMarsh]

OK, freed up a bit of time from other tasks this afternoon and visited the crossover design using the new data for the updated MOD3+ PT2522 configuration. This plot is with 30 degree axis info, more representative of power response, and does include a parallel "notch" filter that is more of a band dip than a notch. Also a tweak or two on the midrange...




Next up, reverse null connection of the midrange:




In my mind, there is no point in tweaking this concept any further until I have a full hardware stack to measure. I'll like just make a phenolic adapter plate to mount one of these on in the space occupied by the TPL75, to do test confirmation.

Now, you know, the "silly" thing is, If you buy premium caps for the tweeter crossover, the crossover will cost a lot more than the basic cost of the tweeters and mounting plates. Of course, there's also the phenolic sub enclosure to be built. Steve thinks I should offer a build and test option for the tweeter assembly for those wanting to streamline their workload- contact me later on if you're interested.

[JonMarsh]

And, to fulfill my equal time clauses, an update to the TPL75 crossover, in which I actually TOOK OUT PARTS!


(Of course, I did change some component values... slightly rethought the topology... and updated the midrange Xover per the adjustments used with the PT2522c MOD3+... but hey... what do you expect?) (it's just my way of doing Kaizen).





I do not claim this is a final configuration- as noted previously, measurements on a full stack will be needed.

I am encouraged by what can be done with the PT2522c with the right effort, but the on and off axis consistency and very low distortion of the TPL75 certainly have a certain value...

[Rigbaby]

JonMarsh Amazing work! Just wanted to say thanks for revisiting the 2522c, your mod has injected some serious rod miles into these little guys.

[JonMarsh]

Thanks for the appreciation- note, full schematics will be posted with final measurements before we have a wrap on this- I tend to not post too much intermediate work lest someone pick that up instead of a final optimization- though I keep scrupulous records of all revisions, and can share development steps with interested parties if desired.

[technodanvan]

Hey Jon, where do you source your phenolic sheet materials?

[JonMarsh]

It's Leecraft BK-2, I source mine usually from Amazon, but you can sometimes find it at Woodcraft, also. It's what I use for the crossover board builds, also.

[Scareurpasenger]

If you want a cheaper source of phenolic, look for synthetic bowling lane material on Facebook. It is some seriously hard stuff and is about 3/8” thick. If you just need a small amount, I can cut up some pieces from the stock I have.

[technodanvan]

That really interesting, I had no idea that was phenolic.

[technodanvan]

Hey JonMarsh, as I gear up for making a pair of these I have a few questions, some of which may determine the direction I'll take in assembly of mine. In no particular order:

1. The top cabinet is the same depth as the lower cabinet despite having a working volume of a fraction of the size available due to the sub enclosure. I take it the top cabinet can hold a fair chunk of crossover parts, but is there any other reason the depth should be the same as the lower cab? This is more for my own curiosity than anything else as I do think it would look more cohesive at the same depth. One idea was to have it slope or curve at the back though, or maybe not have a 'back' to the cab at all for easy access to internals and such during testing.

2. You have added 1" foam around the mid for good reasons, but I'm curious if the bevel in the surrounding wood baffle is still required now that the foam is in place? It may be easier to cut a uniform piece of foam (raising the aesthetics a hair) if we were to simply cut a rabbet around the mid - or even have the foam extend to the L/R edges of the baffle. Would width of the foam make any difference in this circumstance?

3. Regarding the sub enclosure in the top cab - I note it is not as wide as the cabinet itself, meaning it has sides, top, bottom, and back all independent of the exterior cabinet. Additionally, they are well damped internally. This adds a little complexity to the design - I was curious what the impact might be if one were to just use a 'back' piece for the sub-enclosure that extended to each side of the cab, as well as the top and bottom? Damping materials would still be placed internally of course, and perhaps in the rest of the upper cab to control vibrations despite no drivers being present in that rear volume.

4. Do you have any further plans for isolating the top cab from the bottom cab? What would be an appropriate 'gap' between the two, or is it just the 1/4" sheet of phenolic? Would there be significant negative consequences for having removeable bolts to firmly attach the top cab to the bottom, to remove any possibility of it being knocked off by a very large dog?

[JonMarsh]

Interesting questions all around! Here's my first thoughts, before the coffee has really kicked in this morning... ( I may want to come back to this again, because I'm kind of pre-occupied with a meeting upcoming this morning with one of my former colleagues from Europe who heads up the Silicon Carbide team, a discussion he wanted to pursue by video conference after some feedback I gave him by email regarding a recent developer forum).

(I'm clearly still not getting the hang of this whole retirement thingie...)

1. That's the beauty of DIY, you can Do It Yourway... regarding the depth of the top cabinet. Now, I wanted plenty of room for the crossovers, located behind the midrange subenclosure, however big they might wind up to be, and I also planned to have the routing for the input to the top cabinet via a short jumper from the top of the LF cabinet to the input panel on the MF-HF cabinet, (an an internal cable in the LF cabinet from the input terminal; in this case, some Analysis Oval 12 short cables) so having those in a line and the same depth seemed like the way to go, as you can get short high quality jumper cables (or make your own) fairly easily.
Another complicating factor is that I was planning for the option of a rear firing tweeter, and have a network and LPAD design with a GRS RT1.0-8 ribbon (not planar, ribbon) tweeter on the backside- as shown by this version of the back panel drawing. Yes, I have all the parts on hand, but until I implement and test, it's not part of the "official" design. I've only done a rear firing tweeter in one other design, the X-1 SLAMM Klone build.







2. The foam around the opening for the PT5010 was added because there were some diffraction reflections from edge at the interface to the front baffle, which combined out of phase causing a dip in the 2kHz area. The foam pieces are beveled, obviously, with the thickest part at the top. If you cut a straight side piece for mounting the PT5010, you wouldn't have much foam thickness where you really need it, IMO. It might work, but I'd really suggest building a test baffle that way and measuring it by itself to see what the results look like before proceeding down that path. Now, IMO, the ideal thing would be a flat baffle, with the PT5010 mounted flush, but we have this problem that due to the rising response and phase behavior of the PT5010, even with an LR4 crossover, one needs to have the PT5010 set back by about an inch relative to the mounting plane for the tweeter. IF I was willing to do this all over again and build another set of cabinets, I would incorporate a step in the front baffle for the tweeter, with a beveled transition and see how that measures, with stuff to the sides of the mid flat. Like my original smaller development cabinet. But mounting the mid and tweeter in the same plane just doesn't work- the hint is in the cabinet construction that PS Audio also uses, with a stepped Z axis alignment. And it can be seen clearly in the prototype MDF cabinet for the FR20 that I have a picture of partially assembled. So, it's not just me...
But if the midrange was on a large flat baffle, and the opening was done like the 3.5L test cabinet, then the earlier crossovers could be used.

3. It's DIY... but note that the response of the PT5010 is quite sensitive to the rear loading volume, so optimally you need to be sure you maintain the same net volume. Recall, I evaluated 4 different sub-enclsoure configurations. Also, the outer cabinet walls may tend to pick up some vibration from acoustic waves from the LF cabinet, and I didn't want that impinging on the midrange element in any way. This way, there is no need to brace the overall top module cabinet. And my weirdest justification is that if I wanted to try a completely different front panel (not completely out of the question) I could slice off the existing front panel, slice off the midrange sub enclosure, and re-use it. Talk about tortured reasoning, eh?

4. Not yet, but sort of... I have sheet neoprene, and experimenting with the isolation is sort of a final stage tweak, after all the basics are sorted out. Note that the FR30 has a metal sliding track that locks them together. I'm not sure that's a great way to go, but then though there are a number of design decisions Chris made that I think are quite solid, there are others that I diverge from. Like electrolytic caps in zobel networks for the woofer. Keep in mind, he's building for a price that incorporates a fine trade off between performance and profit. I'm not profiting at all, except in (hopefully!) listening pleasure...

[technodanvan]

IF I was willing to do this all over again and build another set of cabinets, I would incorporate a step in the front baffle for the tweeter, with a beveled transition and see how that measures, with stuff to the sides of the mid flat.

Is this sort of like what you are describing? I only ask because this is what I had envisioned. Foam not shown of course, and would extend all the way to the edge on three sides (top, left, and right) and then be chamfered to meet the beveled transition. Note: none of the dimensions are exact here, I just whipped this up to get the general idea. Mid and tweeter would of course be flush mounted in their respective planes.

[JonMarsh]

Yeah, that's pretty much on the money! Now, you can imagine I'm not really keen on rebuilding the existing cabinets, but I have actually given it some thought. That's the right general idea, but the proportions are a bit off... tell you what, I'll dig into Shapr3D and see what I can come up with quickly. Of course, the next obvious question is what specific wood should I use- I could consider 1" bamboo ply for the main part, and then ponder if I would get by with a 3/4" offset using bamboo board, or if I have enough of the Bamboo ply on hand to just go with that...



Curse you Red Baron! (yeah, I haven't forgotten your Air Force connections!)

[JonMarsh]

What I'm thinking is possibly something like this...

[JonMarsh]

A lot of things in the in basket this Friday morning, good thing I started early!


A top priority is refining the concept for a new front panel, and I'm duplicating the development projects so I can entertain multiple concepts if needed, including one with specific dimensioning for the PT2522c tweeter.

But for now, I think that first cup of coffee did the trick, and here's the update:







Key implementation here is "round overs" instead of beveled edges, should be a bit happier in the diffraction department this way. Of course, only the test microphone would know for sure... And there will likely be a strip of adhesive backed acoustic foam on the angled bevel- that's already on hand, too. Some 1/2" stuff, and some deeper more like on the last test build.


As the side walls will be cut back slightly, I think the base front can be 1" if needed. I think I'd rather use the 1" bamboo ply instead of 3/4" bamboo board, but that's partly due to what I have on hand.


I know Steve is working on an FP for Technodanvan with the PT2522 MOD3, but I'll probably model and create drawings for a version of this one with it too- here, it's for the TPL75.

One other comment, regarding the aspect ratio of the midrange sub enclosure, and whether it could be done the same volume but wall to wall of the main enclosure- Not surprisingly, I find that the depth behind the the planar panel mids and tweeter is important to managing the back wave, and consider B&W's top speakers, with their long tail assemblies for rear wave damping- I don't think I'd want to make this mid enclosure shorter and wider. I'll leave that experiment up to others.

[technodanvan]

This is interesting Jon. I'm not sure I'm a fan of the 'double roundover' on the tweeter section but I see where you're going and I like it. I was also considering the placement of a chamfer above the mid, returning the cabinet to the full depth (like the bottom of the MF/HF cab and the LF cab) as you get to the top of the cab. Finally, while I'm normally a fan of roundovers, in this instance I think having a uniform chamfer down each edge - to include the foam - would look rather nice while potentially approaching similar levels of performance.

Edit: Added images.

Edit again: I wonder if using open cell foam without the adhesive back may be a better plan, and in turn not recess the mid at all. Here's my reasoning: assuming the mid flange is on the order of 0.125"-0.25", acquire a thin wood plate of the dimensions needed for the foam that fits around the midrange and extends to the L/R boundaries of the baffle. Embed neo magnets in the baffle and in the thin material, and adhere the open cell foam using 3M spray or similar to the wooden plate. Note that portions of the foam that overlay the sides of the mid would not have adhesive on them. This would allow the foam to be less permanent - one could remove the plate containing the foam and create a new one if it becomes damaged or worn with age.

[JonMarsh]

The piece at the top will create diffraction reflections and make using foam mandatory. The way I've drawn it, is very similar to my 3.5L mid enclosure test cabinet, and it did not need foam at all. I understand the issues about esthetics, which I why a tried it the way I did on the first build. The FR30 looks similar, but the bevel surround on the midrange fits to the radiating area of the midrange and is formed by a removable synthetic cover, as is the tweeter split waveguide.






Note the Z axis offset of the Mid panel and HF panel. For Steve, note how close the Mid and tweeter are vertically, to optimize the size of the vertical window in the crossover region.

Recall, this is the original 3.5L sub enclosure test box...





Yup, round overs.

And this is how that midrange configuration measures 0 - 50 degrees, polar, with the first POC crossover:




Just sayin...

And adding in the TPL75, though it was far too distant vertically (for normal use) and on a separate panel for this test, and with 1" forward in the Z-axis relative to the PT5010, so mic placement was somewhat critical....






You could say it shows the potential...

[technodanvan]

Lots to think about Jon. I'll engage with Steve and let you know what we decide on. Current plan is to implement 'something' along the lines of what you and I proposed here, take some measurements, and start with the original MOD3 crossover to see where we are.

[JonMarsh]

that sounds like a good plan to me- keep in mind, I’m kind of hitting the reset button here, too, Going back to a configuration more like my original test cabinets. I’ve got one extra piece of 1” BB ply that I unpacked yesterday, and will get ready for cutting soon. OTOH, I’ve got a trip to CA storage to pickup a motorcycle and clear out some other stuff next week, and also re-organizing the garage workshop space at home, so there are a number of “irons in the fire” to take care of- plus ongoing communications relative to SiC product development and applications with a long time colleague, the VP of the SiC group at Siemens- one more detailed update to put together for him this morning.

One thing I can say, I sure don’t have time to be bored and sitting around watching TV- but then, it’s still not even hooked up!