• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

3D printing a QRD N53 diffuser

Lookin’ good!

Now I see it, I wonder why one would limit oneself to a blocky shape. One could just as well make a smooth shape based on the same geometry. I don’t think it matters very much audibly, but it may look better. Maybe something for v2 ;)
 
Lookin’ good!

Now I see it, I wonder why one would limit oneself to a blocky shape. One could just as well make a smooth shape based on the same geometry. I don’t think it matters very much audibly, but it may look better. Maybe something for v2 ;)
Thanks! You are not the first to think this, there are smoothed QRD-based diffusers available for sale out there. I personally think they look sort of ugly and I think the blocky ones look cool, so I went with the blocky version. But I imagine the smoothed ones probably work a little better.

I don't think it would be hard to create the smoothed version, but I am not totally sure on what type of interpolation to use between the points defined by the QRD equation.

Assuming this project stays on track and goes well, I do plan to print more diffusers, but I'm going to do BAD panels instead of more QRDs, so I can get some good absorption going in the room also. (it should also take way less time and effort.) The current thought is to do two 4'x4'x4" sections of BAD-faced Corning 703, one on the front wall behind my computer screen, and one hanging on the ceiling overhead.
 
Last edited:
I like the looks in all white. Illuminate it with RGB LEDs if you want more color?

PS: BAD panels are nice. I made mine on a lasercutter, quick and cheap.
 
I like the looks in all white. Illuminate it with RGB LEDs if you want more color?

PS: BAD panels are nice. I made mine on a lasercutter, quick and cheap.
I want to make some of these but couldn't find a computer file...
 

Attachments

  • BAD PANELS 600x600.zip
    171.6 KB · Views: 52
  • bad.pdf
    302 KB · Views: 61
  • Binary (MLS) panel 5063 sequence.pdf
    527.7 KB · Views: 59
  • Curved Binary Surface.pdf
    1.8 MB · Views: 54
  • Kinetics_Absorber_Diffuser_Panel vs BAD.pdf
    285.9 KB · Views: 52
Last edited:
Printer running? ;-)
Yes, I'd better go catch it!

Currently printing the 8th piece, which will be done early this afternoon. It would be more, but I had one failure when the filament spool fell over, and a long weekend out of town where I wasn't printing.

After a few tries I switched to 3D honeycomb infill at 3% instead of gyroid at 2%. It adds a bit to the filament budget, but not a lot, and the feel is much sturdier... and each block has a semi-comforting amount of infill in it. The gyroid prints make disconcerting creaking noises when you press on them. At full scale I think that's the least plastic I can get away with.

So far the fit with the pegs seems pretty good, I think the gaps between pieces won't be easily visible between wells, and it should be easy to fill / sand / paint the gaps on the outside. My current plan is to paint the outside faces and leave the rest white, since spray-painting 25 x 2.5 x 2.5 cm wells seems really difficult to do well.

The bad news: My wife realized that the diffuser is going to be very close to the door as you walk into the room, and objects to this on the grounds that it's in the way if you're trying to move about the house. OK, she's not entirely wrong...

Unfortunately this is the only place I can put it that is sufficiently far from the LP. I may end up needing to sell this one and print another, somewhat smaller QRD (or two) after this.

I think these things tend to sell for a pretty penny at this scale, so it's not the end of the world if it fails WAF compliance testing in the end.

PXL_20240731_163944404.jpg
 
Last edited:
Progress update: 14/36 pieces printed with one going now.

Casualties:
Accidentally printed the same one twice due to a quirk of how you select parts in Bambu Studio.
Lost 3 prints so far due to filament tangles / jams. The rollers I printed to hold the filament spools just fell over a few times. The printer sadly doesn't register a filament blockage / snag.

I printed my own spool holder so I wouldn't have to worry about this anymore. It's not as elegant but I have a lot more confidence in it:

1723418557072.png


Here's the progress so far:

1723418671397.png
 
Finished printing and getting ready to assemble! Post-its on the top because turning them over to read the numbers gets tedious pretty fast.

Now that I'm actually putting this together, I would not do the same tab-notch alignment scheme again. They work but they're pretty fiddly.

I realized on another project (after assembling a cheap medicine cabinet) that simple, old-fashioned dowels work fine. Oh well, next time.

I also realized my longest clamps are 24" and not 36" like I thought I remembered, so I'm going to go out and get some cheap Harbor Freight 60" ones to try and eliminate gaps a bit better than I can by hand.

Any day now...!

1725596822436.png


1725596870273.png
 
Last edited:
I did think about that, but the tabs are only meant to give the pieces a little resistance to shear, mainly they're to line the pieces up nicely during assembly. I plan to use a 2-part epoxy to glue them together, which IME results in a bond that's stronger than the part itself. (I tested the bond on a past large print by trying to snap the bond over my knee... the part snapped instead.)

For assembly I'm basically going to use long clamps and boards to glue together 6 pieces at a time, then glue all 6 rows together at once... should work?

You can't see it on this piece but I did also add attachment points for wall cleats and carved out some carry handles also, so once it's glued together the hard part should be over...

Only just stumbled onto this thread. Very cool!

I know a bit about glue-ups. The problem with epoxy is the prolonged clamping time. You can get around this by using another type of glue, e.g. cyanoacrylate (Super Glue) to hold the part while the epoxy sets. CA glue is brittle and prone to failure, so you only want a couple of small dabs.

You mentioned something about painting your part. Do not epoxy over a painted part - because then the strength of the bond relies on the paint adhesion to the surface. I suggest masking off any area you are planning to glue together. Or do a test to check the strength of the bond.

Re: painting. I would spray the paint on rather than dip it in paint. Paint is expensive and I hate wastage - dipping a part in paint would mean a lot of waste. The problem is that spray painting depends on the weather and you need a proper fume hood and dust-free precautions. Hot and humid conditions affects the size of the paint particles and whether they dry in the air before depositing on the part, which in turn affects the finish. Poorly applied paint will have blotchy stippling (paint dries in the air before depositing on the part) or orange peeling (paint too wet or coat too thick). The proper mixture of thinner to paint and painting distance is essential for a good finish.

Spray cans are expensive, but a proper spray booth + spray gun + air compressor setup is even more expensive.

If that's all too hard, then paint with a brush. It will be tedious and the finish will never be as good as spray paint. But it is also the cheapest way to apply paint. If cost matters, and you don't care too much about the finish, then look for a self-levelling paint and brush it on.
 
Only just stumbled onto this thread. Very cool!

I know a bit about glue-ups. The problem with epoxy is the prolonged clamping time. You can get around this by using another type of glue, e.g. cyanoacrylate (Super Glue) to hold the part while the epoxy sets. CA glue is brittle and prone to failure, so you only want a couple of small dabs.

You mentioned something about painting your part. Do not epoxy over a painted part - because then the strength of the bond relies on the paint adhesion to the surface. I suggest masking off any area you are planning to glue together. Or do a test to check the strength of the bond.

Re: painting. I would spray the paint on rather than dip it in paint. Paint is expensive and I hate wastage - dipping a part in paint would mean a lot of waste. The problem is that spray painting depends on the weather and you need a proper fume hood and dust-free precautions. Hot and humid conditions affects the size of the paint particles and whether they dry in the air before depositing on the part, which in turn affects the finish. Poorly applied paint will have blotchy stippling (paint dries in the air before depositing on the part) or orange peeling (paint too wet or coat too thick). The proper mixture of thinner to paint and painting distance is essential for a good finish.

Spray cans are expensive, but a proper spray booth + spray gun + air compressor setup is even more expensive.

If that's all too hard, then paint with a brush. It will be tedious and the finish will never be as good as spray paint. But it is also the cheapest way to apply paint. If cost matters, and you don't care too much about the finish, then look for a self-levelling paint and brush it on.
For the epoxy, I'm using this clear JB Weld stuff that supposedly sets in 5 minutes and cures in 60... not very onerous for this since I only need to clamp 7 times total. Using CA glue to save time is clever though... it is said to bond PLA pretty well and IME does well, especially with a little baking soda as accelerant... but the strength as you say is sometimes not quite enough.

I gave up on the idea of painting most of this thing - it looks decent enough right off the printer and getting paint into all of those wells would be a near impossibility even if I knew what I was doing. I am thinking I might fill & paint the outside perimeter to make it look smoother, but that's it.
 
Last edited:
Oh yes, I forgot to mention - epoxy glues better if it has a rough surface to glue on to. Drill a couple of holes into the parts you want to glue together. The epoxy will squeeze through the holes and when they set, it will be impossible to pry it apart.
 
Oh yes, I forgot to mention - epoxy glues better if it has a rough surface to glue on to. Drill a couple of holes into the parts you want to glue together. The epoxy will squeeze through the holes and when they set, it will be impossible to pry it apart.
This is actually an interesting question. The surface of the 3D printed part does have a lot of texture from the layer lines, but it's quite small, and the layer lines themselves are smooth. I'm not totally sure how adhesives like this work on a micro scale. I know roughening with sandpaper helps too, but I've also found 3D printed parts hold together really well without doing anything to the surface.

The drilled holes idea sounds like a good one, but maybe overkill for this job? The parts themselves aren't exactly off-road tough. You can break them with your hands if you really want to. IME the bond from JB Weld on a 3D printed part is already stronger than the part.

Mega! Can't wait to see that monster finally assembled :)
Thanks! If I get my act together I'm hoping to finish it this Sunday.
 
The drilled holes idea sounds like a good one, but maybe overkill for this job? The parts themselves aren't exactly off-road tough. You can break them with your hands if you really want to. IME the bond from JB Weld on a 3D printed part is already stronger than the part.

I make wooden model ships. I sometimes have to glue metal parts onto wood. Epoxy alone isn't strong enough, but holes drilled through the metal and wood makes it so strong that I can lift the model which easily weighs 10kg (20lb) by the metal parts. It depends on how OCD you are about bond strength. You have already indicated that your part breaks before the glue bond - to me that is plenty strong enough and I probably wouldn't bother drilling holes to increase the bond strength if that were the case.
 
The diffuser is basically done at this point! I did the final glue-up last night. Lessons learned and measurements to come. First impression of the full assembly - much bigger and more imposing than I imagined when I started. :D

Part way through the gluing process: Row 3/6 clamped with first two rows done on the right:

1725986355380.png

An afternoon shot that reveals some of the internal structure of the columns:

1725986402169.png


All rows glued together and preparing for the final glue-up:

1725986455548.png


Only had to remove 3 tabs that ended up being misaligned! :p

1725986494590.png


Waiting for the glue to dry...

1725986537527.png


Touching up a few gaps that opened up in the process of final glue-up:

1725986607122.png

With any luck I can post measurements and audio recordings tonight or tomorrow. Thanks for following along on this bonkers journey.
 
How much does it weigh, all glued up?
 
How much does it weigh, all glued up?
Around 32Kg I think... but it's a heavy 32Kg. Carrying it up the stairs by myself was not easy, it's really wide and tricky to get a grip on.

RIP my walls:

1726037612423.png


I did manage to get it into my office and do some measurements... details to follow. Preliminary subjective and measured results indicate some effect, audibility as ever is subject to doubt. ;)
 
Last edited:
Back
Top Bottom