[a_recording]

This is a thread on Head-Fi for this newfangled 3D printing thing. Inspired by the recent release of the Mr Speakers Alpha Dogs, I recently went out and bought a 3D printer!
My home is that this thread will be a general thread for all the budding makers and tinkerers on Head Fi to post about their own audio related 3D printer projects and to swap notes about their own experiences / builds and 3D printing in general.
If people's projects eventually need their own threads I will also try to link to them here. Hopefully we can get a community of adventurous peeps making some cool things!
 
/////////////////////////////
 
CHANGELOG
29/08/2013: Thread Created

/////////////////////////////
 
There are lots of guides to 3D printers on the internet, but I will put a brief one here concerning 3D printing and audio gear. This will be updated and expanded as time goes on.
 
Why 3D Printers?
3D printing is a additive (layer upon layer) print process for 3D objects. While this technology has been available for some time to enterprises, in recent years we have seen the advent of smaller 3D printers at price points that individual households can now consider purchasing.
3D printers currently can print in a variety of materials including ABS plastic, PLA plastic and Acrylic resin. This has particular importance to the headphone community as the majority of headphone designs are made out of plastic. With 3D printing we have the oppurtunity to print everything from replacement parts for our headphones, to accessories like cable winders and headphone stands, right up to more involved projects like completely reshelling headphones!
/////////////////////////////

Do I need to have 3D design skills?

This was something I asked myself before I purchased a 3D printer. I personally do not have any professional 3D CAD (computer assisted design) skills beyond learning a bit of design & technology in High School, a semester of 3D animation / modelling classes at University and an understanding of how to use design software on computers. Basically if you are comfortable using Photoshop or playing videogames, I cannot see why you could not also learn to use a 3D printer.

/////////////////////////////

What software do I need?
Surprisingly, a lot of powerful CAD software specifically made for 3D printing is freely available on the internet. Since headphones are relatively simply geometric shapes, many of these software suites are quite powerful for the task at hand. I have started my tinkering with a few of the packages from the CAD giant Autodesk, and here are my observations:

Tinkercad
http://tinkercad.com/
Tinkercad is a free browser based 3D modelling program that was acquired by Autodesk. Tinkercad is a basic program with a very unintimidating interface, and lets you very quickly start making designs by combining basic primitive shapes into more complex ones. I have actually managed to make some very precise and complex objects in Tinkercad and the interface is very fast. Because the interface is so fast and intuitive, you can also import shapes generated in other 3D modelling programs and assemble them in Tinkercad for printing with great ease.
Limitations: Tinkercad does not currently support some more advanced features - for instance, curved edges on objects. This is a limitation because generally hard edges create unwanted interactions with soundwaves for audio gear. Another limitation is that Tinkercad does not currently support a high curve resolution for shapes made in Tinkercad like cylinders or spheres. Strangely, importing high resolution curved shapes made in other programs works fine for Tinkercad up to a point, though if the geometry gets too complex I've noticed that Tinkercad will generate artefacts on the 3D shape. I do not know if these limitations will ever be fixed considering that Autodesk now has several other offerings.
123D Design Online
http://www.123dapp.com/
123D Design Online is another free browser based offering from Autodesk. 123D Design is barely another step on the learning curve over Tinkercad with a similar interface, and offers some features that Tinkercad doesn't, including high resolution curves and curved edges / bevels. It also has some clever inbuilt primitive shapes to work with. However, it does not seem as fast or as stable as Tinkercad.
123D Design (Desktop Version)
http://www.123dapp.com/
Autodesk makes another free 3D modelling program for desktops, though it seems to bare little resemblance to the online version of 123D. 123D Desktop has a very stripped down interface and a powerful set of tools much more like traditional 3D printing programs. However it lacks some of the helpful primitives of the above two programs and I find myself wrestling with the very strange camera behaviour and somewhat unintuitive interface. It's possible to use 123D Desktop to generate some very high precision individual shapes and then import them as STL files into Tinkercad (but oddly not into 123D Design online) for assembly.

/////////////////////////////

Which printer should I buy?

Actually, to do 3D printing you do not even need a 3D printer. There are now numerous services online like Sculpteo http://www.sculpteo.com/en/ that will make 3D prints for you - upload a file to them and they can produce a plastic, acrylic - even CNC milled metal. Of course, there is a cost to this service - but for one offs this might work out nicely.

The advantage of having a personal 3D printer is lower material cost over time, and also the ability to very rapidly prototype changes. Something doesn't work or doesn't fit? Just print it again. You can achieve good results with just a few iterations of trial and error.

There are a bunch of 3D printers on the market and I won't review them all here, but it would be nice to have comments and observations in the thread about people's own experiences.

Personally I got myself an UP! Mini 3D, which I did a quick little video / setup review of here:

/////////////////////////////

How much does it cost?

Besides the initial cost of the 3D printer, ABS plastic costs about 5 cents / gram. You can print in a variety of colours and different plastic types.

/////////////////////////////

LINKS TO PROJECTS
a_recording's Fostex T50RP Driver Frame and Fostex T50RP Baffle Replacement

 

[a_recording]

Fostex T50 RP Driver Frame
 
Here's my first project: a 3D printed Fostex T50RP baffle!
 
Part 1: Initial Driver Frame
 
 

 

 
The initial driver frame was made in Tinkercad after taking dimensions of a Fostex T50RP driver. The curved edge smart shape was used from this maker: http://www.thingiverse.com/thing:94631
 
This part can be used to design new, more rigid T50RP baffles or to completely transplant the T50RP driver into other headphone housings. The part can also be printed out by itself to be used as a testing part to determine if the T50RP driver will fit in a given headphone and where the screwholes would go.
 
This design prints out fine on my 3D printer but I am finding that the screw thread looks a little rough. Keep in mind that you need to keep the little rubber / foam square that comes with the T50RP driver to ensure a good seal.
 
I have made this part available for download here: http://www.thingiverse.com/thing:142059
 
This design is shared under a Creative Commons Attribution Share Alike license. http://creativecommons.org/licenses/by-sa/3.0/
 
This means that you are free to download the part and modify the design for your own personal or commercial use. However, please credit me by linking back to the Thingiverse page.

 

[a_recording]

Part 2: 
 
A full Fostex T50RP Replacement Baffle
 
I'm trying to design a full replacement part for the Fostex T50RP baffle.
 

 

 
This design has what I think is a few advantages over the standard Fostex T50RP baffle.
 
- Much more rigid design due to the large ring that acts as a brace for the entire part
- The interior is a smooth, curved surface to eliminate any potential resonances / interactions caused by edges
- The removal of the baffle side port as per bluemonkey's awesome incremental modding thread: http://www.head-fi.org/t/618659/fostex-t50rp-incremental-mods-and-measurements
- The baffle is now a much lower overall height, designed to sit only 2.5mm above the lip of the earcup. This means that it is easier to fit other earpads on this baffle (like the Shure 840 earpad) without stretching and damaging them.
- The deeper ring about the baffle and the fact that the baffle sits flush with the ring ensures a better seal in the earcup.
 
Issues: 
 
With the latest iteration of this design I am getting issues with the entire part curling up on the build plate, possibly because it is too large. Once I can succesfully print off this design I will be able to see how well it works...
 
///////////////////////
 
Update: 31 Aug
 
Worked out the bugbears causing the 3D prints to curl. Printed off two baffles and can now confirm that they fit and seal on the earcup!
 

 
Compared to the stock pair of T50RP's I have on hand (with Shure 840 pads), these baffles remove a lot of the mid-bass bloat and to my ears make the T50RP sound more natural.
 
The sound should be similar, but entirely the same as blocking the baffle side vent and using Shure 840 pads as suggested by bluemonkeyflyer's awesome T50RP modding thread.
 

 
Again here are the differences between the stock baffle and my 3D printed part:
 
- The baffle is now completely flat, removing the lip that causes unecessary stress on smaller earcups like the Shure SRH840
- The baffle does not have a baffle side port like the stock baffle which should improve low frequency response as per bluemonkeyflyer's measurements
- The retaining ring of the baffle now extends far into the earcup. This provides additional bracing to the baffle, making it more rigid, as well as improving the seal between the earcup and baffle
- The baffle is now completely flush with the lip of the earcup, again improving the seal between the baffle and earcup.
- There is no longer an additional layer of damping material between the T50RP driver and the Shure 840's earpad felt.
- Most interior edges of the baffle have been eliminated or reduced to curved surfaces to prevent interaction with sound waves
 
Update 2 September
 
First successful prints off my new roll of black ABS plastic!
 

 
 
I am not sure what is causing the whitish texture on the earside of the baffle. Soon I will start exploring using acetone vapour to finish these parts to a smooth consistency. Right now I am printing at .25mm which is just off the max 0.20mm layer resolution of my printer. Right now each print takes roughly 1.5 hours to print.
 
Here is the 3D printed part compared to the stock baffle:
 

 
The printed part weights 22g versus the stock baffle's 20g.
 
I may try increasing the thickness of the baffle in the 3D print baffle where this is possible, to further improve rigidity.

 

[Armaegis]

Very interesting. I imagine you'll start getting flooded with requests to make custom parts very soon.

 

[a_recording]

Quote:

Very interesting. I imagine you'll start getting flooded with requests to make custom parts very soon.

 
I'm open to that :3 It wouldn't be bad to get the 3D printer to pay for itself. But for the moment it's just a lot of prototyping...

 

[onefatsurfer]

Awesome.  It would be really cool if you did measurements ala BMF in his incremental measurements thread, to document the changes that the 3d printed parts have on the sound.  That is a big undertaking though.  Keep up the good work!  Hopefully in the future I will be able to afford a 3d printer and will join you in this

 

[a_recording]

Awesome.  It would be really cool if you did measurements ala BMF in his incremental measurements thread, to document the changes that the 3d printed parts have on the sound.  That is a big undertaking though.  Keep up the good work!  Hopefully in the future I will be able to afford a 3d printer and will join you in this

Unfortunately I have no ability to do measurements! But once the parts are near final I can start sending them to people who can give me an idea of how they perform.

I have a stock T50RP as reference myself.

 

[a_recording]

Updated the baffle post since I managed to print off the baffles successfully!

 

[DigitalFreak]

Quote:

Very interesting. I imagine you'll start getting flooded with requests to make custom parts very soon.

I knew it would only be a matter of time before you showed up here, lol.

 

[bluemonkeyflyer]

Impressive ingenuity. Well done taking it to th next level.

Thanks for sharing!

 

[hans030390]

Very nice! I'm keeping an eye on this.

 

[bluemonkeyflyer]

Quote:

Part 2: 
 
A full Fostex T50RP Replacement Baffle
 
I'm trying to design a full replacement part for the Fostex T50RP baffle.
 
 
 

 
This design has what I think is a few advantages over the standard Fostex T50RP baffle.
 
- Much more rigid design due to the large ring that acts as a brace for the entire part
- The interior is a smooth, curved surface to eliminate any potential resonances / interactions caused by edges
- The removal of the baffle side port as per bluemonkey's awesome incremental modding thread: http://www.head-fi.org/t/618659/fostex-t50rp-incremental-mods-and-measurements
- The baffle is now a much lower overall height, designed to sit only 2.5mm above the lip of the earcup. This means that it is easier to fit other earpads on this baffle (like the Shure 840 earpad) without stretching and damaging them.
- The deeper ring about the baffle and the fact that the baffle sits flush with the ring ensures a better seal in the earcup.
 
Again here are the differences between the stock baffle and my 3D printed part:
 
- The baffle is now completely flat, removing the lip that causes unecessary stress on smaller earcups like the Shure SRH840
- The baffle does not have a baffle side port like the stock baffle which should improve low frequency response as per bluemonkeyflyer's measurements
- The retaining ring of the baffle now extends far into the earcup. This provides additional bracing to the baffle, making it more rigid, as well as improving the seal between the earcup and baffle
- The baffle is now completely flush with the lip of the earcup, again improving the seal between the baffle and earcup.
- There is no longer an additional layer of damping material between the T50RP driver and the Shure 840's earpad felt.
- Most interior edges of the baffle have been eliminated or reduced to curved surfaces to prevent interaction with sound waves

 
 a_recording,
 
Would it be possible to program your baffle design so that the baffle "face" and "undersurface" are irregular instead of flat and smooth? I wonder if [size=small]small pyramids, for example, would improve the SQ by breaking up and dispersing sound waves that are then absorbed by acoustic foam and/or other materials?[/size]

 

[a_recording]

Quote:

 a_recording,
 
Would it be possible to program your baffle design so that the baffle "face" and "undersurface" are irregular instead of flat and smooth? I wonder if [size=small]small pyramids, for example, would improve the SQ by breaking up and dispersing sound waves that are then absorbed by acoustic foam and/or other materials?[/size]

 
Haha, would you believe that I started working on this idea BEFORE even printing off these baffles?
 
Essentially you are asking if it would be possible to print an acoustic diffuser on the baffle. The answer is yes. From what I read from doing research, pyramids are not the best shape for making a diffusive pattern because they essentially cause severe lobing off the two angled sides as opposed to diffusing the pattern equally.
 
http://www.rpginc.com/docs%5CTechnology%5CWhite%20Papers%5CAcoustic%20Diffusers_The%20Good,%20The%20Bad%20and%20The%20Ugly.pdf
 
What you want is a 2-dimensional "Quadratic Residue Diffuser". The trouble is, feeding the values into the calculator I found here: http://www.subwoofer-builder.com/qrdude.htm I found that at the size you are making the diffuser, the diffuser could only theoretically work on high frequencies outside the range of human hearing, or is at an insufficient distance from the ear to be effective.
 
Of course this is based on diffusers you would say, put on a wall. I have no idea of the effect of putting such a thing in an earcup. I tried such a thing on a smaller scale with a smaller headphone and though it did sound different, the diffuser displaced such a large volume of air inside the enclosure I could not tell if the sound differences were due to diffusion or simply to changes in damping.
 
Once I nail down this baffle design I will do some experiments to see if putting the diffuser in the T50RP earcup will make any changes. Since you have the equipment and knowledge to get the CSD waterfall plots I'm happy to send you some of the prototypes to tinker with
 
As far as the baffle goes, right now the baffle would act as a single large radiating panel coupled with the driver. Ideally you would want the baffle to act as many small (and ideally) irregular panels to break up shifts the resonances far into inaudible high frequencies, so I will definitely see if making an irregular surface would make a difference.
 
It would appear that Bowers & Wilkins had a similar idea with their flowport: http://www.bowers-wilkins.com/Discover/Discover/Technologies/Flowport.html One wonders if you could minimise noise from air moving inside the earcup by creating a similarly dimpled surface inside the earcup.

 

[FraGGleR]

Awesome work.  Next step is to work on a cup ala the Alpha Dogs.  I'm bummed I didn't get my own set done prior to moving and losing access to a 3D printer.  Plus now that Dan has done his, mine wouldn't have been the original.  

 

[Armaegis]

Quote:

 
 a_recording,
 
Would it be possible to program your baffle design so that the baffle "face" and "undersurface" are irregular instead of flat and smooth? I wonder if [size=small]small pyramids, for example, would improve the SQ by breaking up and dispersing sound waves that are then absorbed by acoustic foam and/or other materials?[/size]

 
As a_recording has said, diffusers on such a small scale should not affect anything except the highest frequencies. I know you've found the foam pyramids to make a difference in the BMF mod, but I suspect this is more to do with the larger surface area for absorption rather than diffusion itself.