Ensemble - 7+ Channels feat. the King Coax and little brother Rex

[JonMarsh]

First, I'm going to vent a little bit more about Seas data sheets and how they measure drivers.

We'll start first with just the description text for the C18EN001 and C18EN002 coax's:

• OK, to start off, they don't even have the grace in gravity (thank you Jonatha Brookes) to include dimensions of the baffle so you can model what the actual baffle step is probably like.
• Second, they provide 0 degrees, 30 degrees, and 60 degrees data. If I was running Seas, I'd at least make them do 0, 15, 30, and 45. 60 is almost irrelevant considering how midranges and tweeters behave, especially what amounts to a waveguide mounted tweeter.
• Third, overlaying both drivers in black and white gives me a bit of a headache separating out what is what. Use some color folks. Even better, separate charts for cone element and tweeter, for the above mentioned measurement axis, and maybe a simple combo chart with 0 degrees for cone element (note, I don't call it a woofer, though it looks like one) and dome tweeter.

OK, for reference:


C18EN001 Graph







C18EN002 Graph







Maybe it's just me, but the on axis response of the C18EN002 tweeter is just plain nasty as depicted in the data sheet. Diffraction issues of some kind? Reflection interaction with the cone? Measurement problems?

Hard to say, but something that turned me off when I first saw this data sheet.

So, what does the sample Danny brought look like?

[JonMarsh]

First, a little bit about how I go about early driver measurements using Fuzzmeasure.

• First step is to measure what I call QNF, for Quasi Near Field, in a manner such as to bypass baffle step and make the primary signal so high in level in comparison to room reflections, that I usually wind up with a very good approximation or match to competent manufacturers infinite baffle half plane tests
• I have heard "Near field" defined by several as around 1" from the driver, and that is not what this is about- one needs a distance sufficient for integration of the radiation of the full driver surface, and free from anomalies due to propagation delays from the edge. In practice, with typical front baffle sizes and driver sizes, a distance of 6" to 8" works fairly well at accomplishing what I seek. One of my favorite "classic examples" of this is this measurement of the RSS210HF-4.

This is just how I like to do things- the subject is DIYW, Do It Your Way, there are many valid possibilities.


Then, I do "near Far Field" measurements, which are distant enough to show baffle step, but keep the primary sound level high in comparison to room reflections. That often works out to about 1/2 meter, so perhaps the one thing I agree with Seas, depending on baffle and driver size. This will usually work OK for polar checks too, unless, as is the case with this enclosure size, it's really too big, because the front baffle can't be lined up with the axis of rotation. But one can get a rough idea trying at 10, 20, or 30 degrees in that case, still.

Yes, in the driver characterization stage I do polar measurements at 10 degree intervals- to use the compare function in Fuzzmeasure, as well as the RMS Average function- sometimes finding that the most representative axis for net power response of interest (out to 45 degrees) is at about 20 degrees. If that's the case, then I use that as the design axis. The really good drivers have no significant difference in their intended operating range between 0 and 45 degrees. The proof is in the pudding with a crossover connected:


GRS PT5010 in 7L enclosure, 0-10-30-30-40-50 degrees:





GRS PT5010, 7L enclosure, with crossover as developed for SMJ-40/Saint-Saens:



Oh, and yes, that driver puts me in my happy place- for distortion, also.



Next, FINALLY, on to business!

[JonMarsh]

Note: All measurements were done at an amplifier output of 4V peak, about 2.8VRMS as measured on a digital scope.

Amplifier drive level was NEVER changed, but measurement gain for QNF tests was changed to keep good dynamic range early (not overloading the front end, and being able to see how low the distortion floor is), but finding the tweeter was the highest sensitivity part, one that gain level was set, went back and checked Midrange and woofer again for level comparison.

Only HD2 and HD3 distortion products are displayed. Mic gain set for peak level at roughly - 10dBfs, to optimized S/N for distortion measurements. Danny remarked on how quiet my environment is; this is mandatory to get reasonably useful results.


QNF Tests


Anarchy woofer- (this test is a bit compromised, given there are four woofers spread out, but was thought to be an interesting check regardless...)

Half Hamming Impulse filter set to 100msec from onset of convoluted impulse from sine chirp measurement



C18EN002 Midrange

Half Hamming Impulse filter set to 50msec after impulse onset




C18EN002 tweeter

Half Hamming impulse filter set to 30 msec after impulse onset.

[JonMarsh]

Some points worth commenting on-

• The baby Anarchy's have a nice clean motor and well designed cone- this is shown here, especially in distortion results. Their only weakness I've seen in the past is that due to CMS nonlinearity, the odd order distortion rises as you push them harder. Having 4 drivers helps, of course, as none of them have to work very hard...and after reviewing the initial impedance data, we tweaked the mass load on the PRs in a way that should reduce excursion on the Anarchies in the 40 Hz area.
• This test showed a bit less upper range cone "trash" on the midrange element of the C18EN002 than the data sheet graph seems to indicate. However, the distortion plot highlights the travel or linearity limitations of the motor- I really would not expected to see so much rise given the modest drive level and the rated Xmax of 3mm. I'd suggest a higher crossover point than we used with the Minerva, something like 450-500Hz.
• the on axis tweeter response is what I would expect with a good waveguide, and is far, FAR cleaner than the Seas Data sheet. Curious... Most curious. But welcome. OTOH, the break up peak at 24-25 kHz is pretty typical of an aluminum-magnesium dome, and will require a carefully tuned notch filter to flatten and reduce the resonant amplification of distortion products that can be seen in the distortion plots.

Also, worth commenting on, this revived my interest in finishing my stereo Minerva build. Now, I know where the cabinets and crossover parts are, and found one of the PR's, And I know I've seen at least one of the C18EN001's since the move, but the rest?

Well, too much other stuff on the agenda to get distracted now, but definitely something to look forward to... and I do have the oringal Minerva crossover design as a starting point for Danny- bumping up the woofer to mid transition to around 450-500Hz is a small effort... then lets see where the rest lies!

[Hdale85]

They look awesome, it's too bad you can't really get some of the drivers.

[technodanvan]

Thanks for the synopsis Jon, and thanks again for showing me the ins and outs of your process! If you want to post the rest of the data feel free - I likely won't have an opportunity for at least a week.

They look awesome, it's too bad you can't really get some of the drivers.

Thank you! This has been a brainchild of mine for some time and we're finally realizing it. A couple recent photos have really held my focus - something that I've been imagining is becoming real. Driver availability is unfortunate of course, as is the price increase for the coaxial. I've reached out to Erich at DIYSoundGroup and it sounds like the desire to get more manufactured is there, but cost increases make the logistics difficult. They would probably also see a price increase if it happened, I would guess.

I had looked into using the Misco XBL^2 drivers a while back but decided the frame size might look odd next to the selected coax. That said, maybe it deserves another look. I also thought about bringing a Seas MR18REX with me to Jon's place to measure it on the same baffle but ultimately decided I'd try that myself. That is less than half the price of the C18 and may be interesting.

[JonMarsh]

I've been reviewing the original Minerva design variations, and will be importing the data on your systems into an improved concept for that- of course, it will likely wind up with a few more parts, due to more attention to detail, like the tweeter HF notch for the breakup mode. the original Minerva measurements didn't go up that far.

[technodanvan]

I'm okay with a few more parts, I should have a bit more room for crossovers than the Minervas.

I also forgot I had mentioned the Sica coaxials the other day, I have a pair of each of the 5" and 6" versions sitting on a shelf that might pair well with the Misco woofers. I need to do an inventory of the Anarchies I have so I know how far i can go towards seven full channels with those before switching to another model. Erich mentioned he may have more of the 708s sitting around too, so I've inquired about that. All the 704s were purchased in bulk by a larger manufacturer, unfortunately.

Edit: Unrelated to this project, but is this Peerless even MORE affordable now???

https://en.toutlehautparleur.com/ful...-5-9-inch.html

[Hdale85]

Yeah those Anarchy drivers are something special, I was going to look for some for the new car when it comes in as I plan on doing a full DIY setup in there lol. Really sucks they don't sell them anymore. Be amazing if he did another run of them sounds unlikely though.

[JonMarsh]

OK, we've been grinding the sausage- in this case, since we're using a newer version of the C18EN001, went back to the Minerva project, pulled up the old data and crossover, and did a review and went to work. There were "data acquisitions" issues at that time, and no data above 20kHz, so no addressing ultrasonic tweeter issues. My recollection was that there was some other stuff that could bear addressing, but you know how that goes- more components in the crossovers! But also, better chosen ones in the end, IMO.

This was sort of like a wedding effort- that is, something old, something new, something borrowed, and something blue...

• Old: the driver data from 2017 - noisy, and limited bandwidth on top end. Boy, what a difference 7 years of evolution makes, particularly in Fuzzmeasure and VituixCAD!
• New: Impedance compensation for Fs on the tweeter and midrange elements, and other topology and component tweaks
• Borrowed: The LF crossover topology, on loan from the SMJ-40/Saint-Saens
• Blue: How sad I felt, seeing how expensive the Scan Speak 26W-0-00 PR's are these days!

Well, this is like a warm up for Danny's crossover, and working with the somewhat demanding C18EN00x series. The updated crossover has a nice impedance curve, some voicing effect, very good reverse phase midrange notches, and is only deficient in that I don't think you can shoe horn it into the original Minerva crossover location in the base... and it would cost more.

Officially, this is an LR4 on woofer to midrange, and LR6 for Midrange to tweeter. The midrange covers two octaves, from 500Hz to 2kHz.

[JonMarsh]

OK, folks, the sausage grinder has been working some more, and Luigi is ready to share some results...

First, a couple of comments regarding the measurements, and the response character/limitations of the C18EN002, plus a note re the LF configuration.

• Now, the SPL graph for the tweeter section on the Seas data sheet is something of a hot mess for the 0 degree plot, and rather incomplete for the other axis, but direct measurements show somethings didn't change much between the two versions as far as I have measured- yes, the tweeter response looks like a classic waveguide profile, and the relatively flat part between 10kHz and 16kHz, like the -001, is at about 83 dB. With the low sensitivity of the Wavecor SW223BD02, that isn't an issue, and some attenuation is required to match up to the woofer.
• Here, we have a 4x driver series/parallel array, and one of the nice things about that (as used in the SMJ-40) is that it picks up the woofer sensitivity by 6 dB. When all the data made at 1.2 meter is normalized, the sensitivity of the Anarcy 558 array comes out at about 88 db. So, what this means is that a "classic" down tilt voicing is pretty much mandatory.
• Also, measuring at 1/2 meter (to reduce room reflections) means that the measurable impact of the PR setup is reduced significantly, due to the differences in Z axis.
• Doing all this reasonably right while keeping component count "reasonable" (reasonable by whose standards? Mine? Or Danny's?) is a bit of a challenge, and for fun I also included an SPL plot with the Woofer zobel compensation network disconnected and the tweeter notch filter bypassed. For those of the "boom-tinkle" persuasion, who knows?

Full schematic:



Complex SPL, with driver contributions shown:



The Filter transfer functions:




Basic SPL plot, showing voicing





And for giggles, the lower component count version:




BTW, the impedance curve is definitely uglier with this one, particularly in the bass- dips quite a bit lower due to the undamped resonant interaction between crossover and reflex driver. Three ways always have more issues in this regard.

Now, this is just the starting point to put together some POC crossovers and measure the complete system with them. But it ought to come out pretty close...

[technodanvan]

Hey Jon, I'm guessing this doesn't matter since you were using measured data, but these woofers are actually the Anarchy 708s - the bigger Anarchy.

[JonMarsh]

Oops- my bad- but they seemed smaller in real life! Too used to the RSS210HF I guess!

[JonMarsh]

Right. well, no, it doesn’t- it’s just a name, and since there is no good data sheet for these parts, it’s what you measure that counts! but thanks for the correction… still, the crossover is what’s it is.

[technodanvan]

Well thank you for knocking that out so quickly Jon! I’ll see what I have and order the rest of the parts when I get a chance.

[technodanvan]

Right. well, no, it doesn’t- it’s just a name, and since there is no good data sheet for these parts, it’s what you measure that counts! but thanks for the correction… still, the crossover is what’s it is.

Jon,

Is there an easy way to export a list of components from your schematic? I only ask as some values are difficult to read -take R12 for instance, it looks like 100 ohms on the diagram, but I'm guessing it's 10.0?

Edit: Did you see all the new Jantzen caps on PE? Up to 390uF on special order!

[JonMarsh]

Don't guess, read... It IS 100 ohms. it's in a parallel resonant notch filter, and we need a deep notch, and high Q.

No, I don't recall the V2 version having a parts list export mechanism, but I can check. V1 had one, but it just generated a bitmap file with the parts listed. V2 will just generate another bit map, though, I expect, using the same font as in the schematic. All the graphic windows except the Schematic, which is a direct import from VituixCAD, are generated using a Windows utility program for screen capture, as the VituixCAD graph export generally winds up very very small font sizes.

The text is very clear on my 4K screen connected to Mac Studio, from HT Guide, and it came from VituixCAD running on a 2013 Mac Pro on a 4K monitor with windows UI scaling of 200%.

[technodanvan]

Interesting. And C1 and C17 are indeed 330uF? I need to load this into VituixCAD myself to play around...

[JonMarsh]

Interesting. And C1 and C17 are indeed 330uF? I need to load this into VituixCAD myself to play around...

Yes they are- impedance compensation networks to minimize frequency response interactions between the driver impedance and crossover network. \

Keep in mind, this is what I would build, and this shows how it would work. Nothing preventing you from ripping it all up and starting from scratch with ideas and techniques you prefer.

As I mention often lately, DIYW is a totally valid concept. But you should be familiar with filter theory and have some understanding of how driver impedances and response interact with crossover networks- particularly for three ways. Two ways are easier, for example, because the woofer resonance based impedances are relatively far away from the filter crossover frequencies. A three way like this, when the midrange and tweeter have relatively high resonant frequencies present additional challenges. One can ignore them, but then, that's like what happens when one does a simple crossover on a Seas DXT... unexpected bad things happen, and you don't the potential of the driver.

I suggest using the tool application in VituixCAD, convert the FRD files to scaled minimum phase files, re-create the crossover (a good training exercise- of course, I can send it all to you) and try opening parts to see what happens when a circuit leg isn't present.

BTW, I'll mention this again, that 5 dB voicing slope (a common practice by some) is mandatory here due to the sensitivity contrast between the output of the C18EN002 in the range of minimum cone waveguide boost (10-15 kHz) (being about 83 dB) and the sensitivity of the 4x series-parallel woofer array, which boosts sensitivity by 6 dB.

[JonMarsh]

Learning to use the Optimizer tool for setting a target response for a specific driver and honing in on that with the crossover network is very useful(manually preferred, but sometime the Optimizer does OK- other times if you pick the wrong components or topology, it runs away off a cliff)

[JonMarsh]

I created a "redacted" folder with just the files for your King Coax tower, and created a zip file from that and messaged it to you.

A few notes about how to use it:

• Using this project file on a different computer, you will need to re-load the data files into the Driver Tab. This is because VituixCAD uses the full path of the file for its location, even if it's in the same folder or a sub folder. So, just being on a different PC with a different top level path name means it won't be found- similar depending on whether you use your personal data folders or the public directory data.
• The way to do this is by drag and drop the data files into the appropriate slot for the SPL data file for each driver (the minimum phase file which has been scaled from the FRD file from Fuzzmeasure, which has the .txt suffix) and the impedance data file (zma format is what I use- VituixCAD works with most). All will have to be updated.
• When I copy VituixCAD projects to my LG GRAM 17, same process is required.
• If you haven't spent much time on VituixCAD, a review of the user manual would be very useful, I'm guessing.
• BTW, that lower component count version SPL curve I posted above is just disabling the woofer LCR zobel (one of the larger caps you were concerned about) and disabling the tweeter HF notch filter (putting a 10 ohm resistor in their would accomplish the same thing!)
• While you can get NP electrolytic caps in the high capacitance range for those networks, I'd warn you about internal parasitic ESR resistance and the heating loss that comes with it- they don't like high operating currents, and they can literally be worn out from operating at high temperatures due to parasitic losses. So I wouldn't do that for the woofer one, but NP might be OK for the midrange one.

https://kimmosaunisto.net/Software/VituixCAD/VituixCAD_help_20.html

[technodanvan]

Great information Jon, thank you! Last question (for now). Regarding the 330uF caps, do you think it would be better to use several caps together to make up the value, or a single 330 uF cap? I alluded to this earlier, but Jantzen actually has one now. Cost is roughly the same either way, though it'd be a month before the big one would ship.

[JonMarsh]

It comes down to packaging space for the crossover, and if needed, tweaking. I usually use 120uF and 100 uf parts, typically either Dayton or Jantzen CrossCap.

[technodanvan]

All parts ordered from several distributors. I had a hard time finding appropriate inductors for some values due to DCR requirements, but we'll see how it turns out when they arrive. The plan is to construct temporary crossovers along with the already assembled speaker then measure them - I don't see any reason to remove the drivers until we're happy with the internals. My birthday was yesterday and my wonderful wife has ordered me a Macbook Pro, so I guess I'll wait for that to arrive before measuring anything. In the meantime, I need to assemble the other cabinet and figure out how I'm going to finish the aluminum bits and baubles - they look so nice raw that I barely want to risk ruining them!

Regarding the crossovers: I'm a little concerned they won't all fit on the three phenolic boards I cut for each tower. These may need to be pretty three dimensional!

[Steve Manning]

All parts ordered from several distributors. I had a hard time finding appropriate inductors for some values due to DCR requirements, but we'll see how it turns out when they arrive. The plan is to construct temporary crossovers along with the already assembled speaker then measure them - I don't see any reason to remove the drivers until we're happy with the internals. My birthday was yesterday and my wonderful wife has ordered me a Macbook Pro, so I guess I'll wait for that to arrive before measuring anything. In the meantime, I need to assemble the other cabinet and figure out how I'm going to finish the aluminum bits and baubles - they look so nice raw that I barely want to risk ruining them!

Regarding the crossovers: I'm a little concerned they won't all fit on the three phenolic boards I cut for each tower. These may need to be pretty three dimensional!

You could always go for the external option

[technodanvan]

Never!

[technodanvan]

So many bits and baubles...

Caps are mostly Jantzen cross-cap, with Audyn for one value that was out of stock. I went up a quality or two for some of the series caps. Inductors are mostly air core for now, with a couple iron core that you can't really see in the photo. Resistors are all Dayton 20 watt models, and as shown below I also grabbed a bunch of little aluminum stands for those. Figured I'd try them out at least since they weren't too expensive.

Also got some prototyping terminal plates but was a little disappointed to see the terminals didn't come with set screws, or any way to insert wire without soldering. I'd think that would be more useful so I'll replace those at some point. This was my fault, I didn't look closely enough at the photos online.

Probably not related to this thread, but I did go ahead and get a pair each of the MAC drivers. They look really nice up close. From the website:

"They all feature cast frames, anodized aluminum cones, and a slightly extended pole piece. They are vented under the spider, out of the back plate, and the voice coil former."

Given all the designs I heard I'm curious about knowing a little more about them. No idea when I can get a baffle made for the -05, but I may have something suitable for the -04 and -06 already made up.

[technodanvan]

Question for the gurus out there. I've certainly seen all the diagrams showing what happens with inductor values when you place them close to each other in various orientations, however this is for your average inductor. I was curious if the same rules apply with toroidal inductors?

[technodanvan]

It's been more than a minute since I updated this thread and that's largely because I have made zero progress on these since January. I'm wrapping up the bed frame project I started around that same time and in the process I'm cleaning up my garage. I could get half a car in there now! As benches begin to clear I'll see about assembling the other speaker cabinet here in the coming weeks.

A teaser for the frame I've been working on. I'll be nearly as proficient as Steve when it comes to inserts after this.

[JonMarsh]

Half a car, or a whole motorcycle? Appreciate progress in what ever sizes it comes, eh?

[technodanvan]

Half a car, or a whole motorcycle? Appreciate progress in what ever sizes it comes, eh?

Maybe even two motorcycles!

[technodanvan]

The time has finally come to get these parts back off the shelves and onto the production line! Just getting my bearings here as it's been the better part of a year since I assembled one of these. First up is installing inserts for the crossover boards (I am hoping I still have some Rampa inserts somewhere) and cutting and drilling new crossover boards for Speaker #1 as the others were inadvertently coated in spray adhesive last year. Once that is done I'll mark the crossover boards and start laying out the crossover parts to see how everything can fit, and start assembling the box.

Getting Speaker #2 to the same state as Speaker #1 should be relatively trivial, however there are other steps involved that range from 'required' to 'nice to have' if I am to visit Iowa in October again. Given an unknown time frame for some of these it is important to wrap up the cabinets no later than June 30th to give enough time for vendors to catch up.

1. Cut a removable 'door' for the lower crossover board in the lower rear. (reasonably easy, cheap, REQUIRED on both)
2. Align and install inserts for the aluminum bases so they can be installed. (reasonably easy, cheap, nice to have)
3. Sandblast and anodize all aluminum parts. (unknown difficulty sourcing a vendor, unknown availability/timeframe, unknown price, very nice to have)
4. Paint cabinets. (vendor already chosen [assuming they still exist], unknown availability/timeframe, expensive, very nice to have)
5. Acquire chrome hardware throughout (reasonably easy on various website, unknown price, very nice to have)
6. Design/manufacture a reasonably compact, low rolling platform with air-filled tires to easily move these about once complete. (cheap-ish, 1x REQUIRED for Iowa, 2x may be preferable)

[theSven]

Great to see some traction pick up on this build!

[technodanvan]

I pulled the phenolic boards out and threw all the parts for the low pass on the largest of them. That's all I need to do, right? Then throw some wire and solder on there and it's done?

Seriously though, I think I need some structures 3D printed, at least for the toroidal inductors so I can place them vertically. I do not have a 3-D printer though, so I'm open to suggestions. Obviously one is to construct these in exterior cabinets so I have more room, but I'm not sure I like that method.

[Steve Manning]

I pulled the phenolic boards out and threw all the parts for the low pass on the largest of them. That's all I need to do, right? Then throw some wire and solder on there and it's done?

Seriously though, I think I need some structures 3D printed, at least for the toroidal inductors so I can place them vertically. I do not have a 3-D printer though, so I'm open to suggestions. Obviously one is to construct these in exterior cabinets so I have more room, but I'm not sure I like that method.

I find if you throw all the parts in a chilled cooler and shake for ~45 seconds you should be good to go. 😁

On the serious side though, you could look on McMaster at nylon, or other plastic hardware, threaded rod, nuts, washers, etc. for stuff to mount the coils vertically. They also have solid block material that you could fashion what you need?

Steve

[JonMarsh]

Big crossover components and small speaker cabinets are not a match made in heaven...

[technodanvan]

One solution is to use a single 330uF electrolytic to save significant space. I even have them already since they were inexpensive...but I also have all the fancy Jantzen caps already. Decisions, decisions.

[technodanvan]

The last 24 hours has been expensive. I decided to invest in a 3D printer last night for use on several parts in the crossover and elsewhere, and I got a big one capable of doing waveguides in the future. Hopefully will arrive in a week so I can start playing around with it. Then today I dropped off all the aluminum parts at a finisher. What I'm doing is a bit more expensive than I had planned but I'm hopeful it will look very sharp. Four to six weeks turnaround time.

If you count up the pucks you'll note an odd number...and not the right one. Should be 24 + one spare. I forgot a pair of them at home that I was using to hold stuff on my desk. Gotta run back down to the finisher later this week to drop them off.

[Steve Manning]

Boys and their toys ..... maybe one day I can add a printer to the arsenal.

[technodanvan]

Man, I almost ordered one with a laser engraver/cutter built into it but decided that might compromise the printer. Another toy for another day, perhaps?

[technodanvan]

Current plans, though this is not quite proportional to what I'm actually building as I did not update the model since Steve fixed it. Still unsure about the accents on the front, if I do them I'll probably keep them symmetrical.

[technodanvan]

Minor progress so far this weekend. Hoping to get the vinyl and mass loaded foam installed on the sides tomorrow.

The 3D printer arrived way earlier than indicated as well, so I set that up as far as I can considering the first box of filament won't arrive until Monday. Still, I have made some headway with the software. I also discovered an application that uses the Lidar on my iPhone to perform a 3D scan of an object I may try to integrate in this design. After three passes around said object with my camera, the real time 3D model looked absolutely godawful. However, after it processed that data for about 20 minutes (definitely the hardest 20 minutes of my iPhone's life, judging by the battery life and heat produced!) and it was spot on. I don't know how often I'll use it, but I paid for the annual subscription immediately - buying a 3D scanner just doesn't seem very valuable and I was absolutely floored that my phone could do this.

[JonMarsh]

Lots of progress, and a lot roomier working space than I have! I'm jealous!

[technodanvan]

I'd like to reduce that room considerably still...but I keep finding ways (and reasons) to not do that!

[Steve Manning]

Very cool on the 3D scan Danny. I believe I can see where this is going. It would be interesting if you could get enough data and resolution for a print or wood carving of that.

[technodanvan]

It's in CAD right now, though I don't seem to have the necessary skill with Fusion to do what I want just yet. The Bambu software for the 3D printer seems to think it's good enough as-is for a print already, but I was thinking more of a 3D CNC situation...

I think Fusion said it exceeds 25,000 points of data. I don't know if that's a lot.

[technodanvan]

Well I've been printing virtually nonstop for about 48 hours and finally got around to v1 of my big inductor support. Placement is about 2.5 inches apart. Think that should be enough? Space is at a premium, but I 'might' be able to turn one. Not sure that's hugely important on a toroid though.

Next version should use less material and have some more refined dimensions. Final version will use some fancy heat resistant plastic due to the resistor - this is just the cheap stuff for testing.

[Steve Manning]

Well look at your bad self! Very nice Danny 👍👍👍

[technodanvan]

My wife and I have already had semi-serious conversations about getting a second printer...she's finding a lot of stuff to print, and I have dozens of various tools/tool organizers I want to print. Besides all the fancy speaker stuff.

[Efalegalo]

Hi Danny,

Which 3D printer did you end up purchasing?

[technodanvan]

Bambu Labs H2D - their newest, biggest one. I did not opt for the laser modules.

[theSven]

That printer is a beast!!!

[technodanvan]

I had some reservations about going big right away, but I decided I wanted to be sure I could do large waveguides in the future so I jumped in feet first. No regrets, I have so many things to print that it has been running nonstop for almost a week now.

[technodanvan]

Low pass Xover Mk. II. Improvements thorughout, still not quite done though. Tolerances are a bit tighter than expected.Getting a better feel for how CAD translates to real life prints.

[JonMarsh]

Very nice work, Danny! One more tool I don't have anywhere to put at this point!

[technodanvan]

...and Final Low pass board complete. This stuff is tough enough I wouldn't need to use the phenolic...if I had cut the mounting holes evenly, which I did not. So for this project I'll be adhering the 3d structure to the phenolic. In the future I do not think this will be necessary. I made a separate little hat for the 2 mH air core to support a small capacitor and inductor. Worked out pretty well!

Very nice work, Danny! One more tool I don't have anywhere to put at this point!

This really might be my best purchase in years. There are some that are much smaller (and more affordable), though in the case of this structure I used much of the width of my large printer to do it. I really think that in a few months I might get another one if my (and my wife's) ambition keeps up. Just no end to the possibilities.