[Kwisatz]

Well, I've had my LCD-2's for what seems like forever now and the original adhesive rings keeping the earpads on eventually wore out allowing them to slide (and make an adhesive mess on the plate they attach to). Audeze sent me new ones upon request, but for whatever reason, the new ones never seem to last long. Eventually after a long session they'll start to fail and the pads will creep. It'll start slowly at first and then quicker as the adhesive continues to degrade. It could have been a bad batch of adhesive rings for all I know, but the way the pads attach bothered me conceptually. However, it was never quite enough for me to do something about it - especially when I've been doing a lot more speaker listening these days.
 
That changed last week when they failed again. This time it coincided with the addition of a parts cooling fan to my new 3D printer. I figured that now was as good a time as ever to do something to fix the issue once and for all. My first 3D modeling cob job would be a fix for the pad slide. A couple of quick measurements later and I was creating a modified version of the stock black plastic ring that's inserted into the back of the earpad. The new version has internal tabs with pegs attached that fit into the corners of some of the driver plate slots. This basically locks the pad support ring in place preventing any lateral movement, and thus any sheering force to the adhesive rings. Technically it's good enough to keep the pads in place even without an adhesive ring and the pads wont shift at all in use, but it's not designed to have any real prevention of perpendicular forces so the pads will come off easier than you'd probably want them to.
 
Lackluster photo of the modified ring in the pad:

 
Assuming anybody else has ever had a problem with the pads sliding over time and is interested in trying it out, I've posted the STL over on Thingiverse: http://www.thingiverse.com/thing:454526. The photo I have above is of a previous version before I modified the lower standoffs to clear the pad stitching better. It might work for other LCD variants as well, but I haven't measured the driver plate, so I can't really say for sure.

 

[Kwisatz]

Well, I did a tweak to the design to make it useful without any adhesive ring to help keep the pads in place against gravity etc. Now paperclips (or any other sufficiently small ferrous metal rod) span the distance between the top and bottom standoffs. These are attracted to the magnetic field generated by the magnets on the other side of the plate and thus lightly pull the ring against the plate. The force is strong enough to counteract gravity, shaking and other minor forces, but weak enough to pull off easily when wanted. I mostly did it to avoid needing the adhesive rings at all, but I guess it would also work for people that wanted a quick detach system for the different pad types. (Not that there appears to be much interest in the rings in general.)
 
And now some slightly better pics than last time:

 

 

 

 

 

 

 

 

 

 

 
This will probably be my last adjustment to the rings unless something fails prompting a redesign. I'm quite happy with the current results.

 

[pragu]

This looks great! I am also thinking about making some kind of ring support for my LCD-2's pads, but I was thinking of hot gluing the lip support onto the headphones and allowing the pads to be twisted / pulled on.
 
I am also thinking about 3d printing fazors for this assembly. From what Tyll describes, they seem to be static pieces of metal that deflect the sound waves, which should be well within the realm of 3d printing possibilities. 
 
Here is my initial design, (OpenSCAD version) using equilateral triangles (extruded) for the actual fazor bars, however I think I may change it to the slightly rounded design that is shown in this picture. My fazor / pad assembly is for the inner fazor only at the moment, but once I get the balls to take off the dampening material covering where the rear fazor (pictured) will go, I'll measure that space as well and print a fazor for that side too. 
 
It is currently 3mm high, with the fazors reaching a little higher. If your assembly is higher and still comfortable, I may make mine thicker, as the 1mm lip is a bit worrying.
 
Any advice of tips you have would be greatly appreciated! I am relatively new to 3d printing and modeling.

 

[Kwisatz]

Interesting... I haven't been following the LCD changes (keeping up on stuff here is never good for the wallet), so I wasn't aware of the addition of the "fazors". They certainly seem to be easy enough to integrate into the pad side retention ring. Heck, they even lend themselves to magnetic attachment! By putting a metal core down each fazor rib, the fazors (and thus the ring) will be pulled against the driver plate which should also minimize chance of those ribs vibrating against the plate.
 
The addition of the fazors would make printing itself a bit less clear cut though as we're left with overhangs. Your design is easy enough to print if support material is used, but modifying mine would require both a raft and support material I'm afraid due to the centering pins or fazor ribs both being too small of contact area to properly adhere to the bed otherwise.
 
Either that or I'd have to split it and glue the parts together. Although that sounds less than ideal, it might make the metal cored ribs easier to pull off. Basically split off the fazor part that would stick out from the ring. Now the pad facing side is flat with no overhangs like my current design. From there, design the ribs so that there's a channel running through the ribs from the pad facing side up to the last layer of the driver facing side. Push (and possibly glue) a paperclip/whatever into that channel, and then glue the rounded top part of the fazor rib (printed separately) down onto the rest of the rib effectively sealing the metal rod in the plastic. It's a bit involved, but should be more reliable for the metal rod. Designing a hole running the length of the rib isn't going to hold up well in the real world where the tolerances of home printers just aren't quite good enough. It'd likely need to be drilled out, and drilling that long of a hole would be a challenge. In comparison, minor post print tweaking of a channel is easy enough with a sharp blade. I'll probably try a stab at retrofitting my current ring model this weekend.
 
As for advice and tips, I'm new myself. The ring was my first thing that went from idea on paper to 3D modeling to printing. I'm using Autodesk Fusion 360 for modeling, and recently switched from Slic3r to Simplify3D for print prepping. While Simplify3D isn't free, the previews are worth it. Being able to trace through exactly how it'll print before it actually is printing is stupidly helpful. Knowing how the print head is going to cross a given part, how it's going to build up support material, how fast it's going to be doing something at a given point... invaluable really. I also find it generates better gcode for the actual print.
 
For example, here's a preview of your ring running through the virtual print:
 

 

[pragu]

I like the idea of putting ferrous rods and not having to use glue, but I'd rather not use two pieces on the fazors- I don't know much about acoustic engineering, but my gut tells me that if we had even a little glue beading / inconsistent angling / gap in the wave guides it would negate any benefit the fazor might give. 
 
For rev 2, I separated the fazor part from the headphone ring part, so those parts must be glued together, but the fazors remain in one piece. This shouldn't take any support material at all, which is always nice. There are now also 1.3mm channels at the bottom of each fazor bar which should allow you to slot a standard paperclip in each one. Lastly, I increased the height of the upper circle (the one that goes inside the pads) from 2mm to 1mm because I thought 1mm would be too flimsy (how thick is yours?) Hopefully with 6 paper clips it should stay attached, but if not i'm not totally opposed to glue.
 
Here is a dropbox link to all the files. The above links may or may not work now, I'm not totally sure.
 
I'm planning on test printing on Wednesday.

 

[Kwisatz]

Honestly, gluing the tips of the fazors on would be about as complex as gluing a cut off tip of a pyramid back on. It should be relatively easy - super easy if the top pieces were designed to have slight tabs on the bottom to notch into the channel of the fazors where the metal rod would sit. Any effect that gluing would have on the sound by having a few micron misalignment would be an order of magnitude less than the simple shape and size of the fazor in the actual design phase. We're talking multiple millimeters of plastic being used in the fazors with who knows what curve being used to shape them. Meanwhile gluing errors should be at most a few tenths of a millimeter.
 
At any rate, your split would work for your design as it stands. (However, I'm not fond of having ferrous metal parts that could potentially fall out (glue failing) near strong magnets and thin driver membranes.) The main problem with your design currently is it still has no way of keeping itself centered on the driver plate. The rods will hold the pad against the plate, but not keep the pad from sliding around on the plate. Once you put something in the design to retain the ring, you run into the issue with support material again. Fazors have to extend beyond the ring in one direction, and the centering mechanism has to sit beyond the ring in the other direction in the same space that the fazors are. Without cutting something off, you're back to needing support or potentially support + raft.
 
As to the thickness of mine, it is 1.1mm thick with 1.5mm standoffs from the driver plate. Printed with PLA, it's more than sturdy enough. Going a lot thicker would make putting it into the pad more difficult for little practical gain. Feel free to experiment with thicker values though, as it's not like it costs much of anything to do so (other than time).