[akatsuki]

Digikey is letting you place orders for the Peerless 50mm so I just bought 10. They are still back ordered though.

 

[dougms3]

Hopefully my build photos below show somewhat how the driver is being clamped; the copper coloured driver cup holds the driver from the front, just around it's rim and the large cover part clamps the driver around the back of it's outer rim. This is basically how I've clamped all my drivers, with some designs having a ring of thin foam around the sides of the driver, and some clamping the back of the driver in multiple places to in theory help with rigidity, but I doubt it makes any real difference.

I've done a fair bit of headphone repair and usually the drivers will be glued into place, although they often use a type of almost silicone sealant type adhesive, which isn't too hard to remove cleanly.

I'm not too sure what wire I'm using currently. I got it a long time ago from a company that made defence electronics so it's fairly good stuff. We used it on some Eurofighter Typhoon pre-production components if I recall correctly. I just use old-school leaded solder, as I have loads of it and I don't enjoy working with modern solders as much.
edit: this is the wire I've been using - Harbour Industries M22759/11-24

Over £100 per metre is some pricey wiring. At least you don't have to use too much of it per unit. I'm glad to hear you enjoy the results.


Wiring these headphones was a bit of a journey, but it worked out in the end.

Thanks for the insight, I guess silicon is a good option. I would go that route since its forgiving and easy to work with.

 

[Tofty]

It's been a while.

I bought an old pair of Grado SR125s, but the forks on both sides were either cracked or cracking around the metal rod (looks like a design issue, as the "e" version is different).
I therefore printed some simple forks, that make use of the original metal rod, as well as the two plastic side pins that connect the fork to the earcup.

The metal rod is held in place using two 4-40 grub/set screws, one from either side.

Not a printed solution, but I just like how these turned out:

This is a purely cosmetic mod I've done to a recently acquired pair of closed-back Sennheiser HD 569s, since they were looking a bit rough. I wasn't aware that there was a closed-back headphone in the HD 5X9 range, until these came up for sale, but I'm glad I grabbed them when I had the chance.

The original fabric side pieces on these headphones was looking very worn, so I decided to swap them out for some nicer looking fabric. Pretty much anything would work, so long as it could stretch over the slightly domed surface, as these are closed-back headphones and the drivers themselves are contained within a sealed inner chamber, so the design of the outer earcup isn't significantly effecting the sound.

For the earcups; removing the inner ring that surrounds the fabric wasn't trivial, as it's screwed on with two screws in the centre and glued in place around the perimeter from the inside, but thankfully they used a silicon sealant type adhesive and it mostly came off in large pieces. After the adhesives removal, a small amount still remained between the ring and earcup body, but with some careful prying, I was able to remove both sides without damage.

I was able to use the removed fabrics from the sides as patterns for their replacements; in this case a copper and grey paisley viscose fabric. They were glued in place using fabric glue, just in the centre and around the outside, as I was worried the glue might seep through the fabric and discolour it, but it didn't appear to.

The original flocked finish was coming off the earcup fabrics, as well as the headband and earpads. I rubbed the flocking off the headband, to expose the fabric underneath an replaced the earpads with a different type, although this will affect the sound somewhat.

This was the first side I did and unfortunately the pattern doesn't look as good as the second attempt. There's a bit too much empty space.

Overall I'm really happy with how this came out, for what was a relatively cheap and straight-forward mod.

Fancy fabrics on all headphones going forward!

 

[Johannes P]

Hello hello, brand new here! (Well, I've lurked for a long time but decided to register just now.)

I initially saw mwildebeast's Borealis clone and didn't think much of it, but Tofty's refined version really got me interested. That's was before I even had a 3D printer, but now I do. I've been wanting to build my own headphones for months now, and with the Peerless 50mm drivers back in stock at Digikey, that time has come, but I have a few questions.

How are the headphones wired? I feel really silly not being able to figure this out myself. I'll be using a TRS jack on both sides, no bridge cable or anything fancy.

What material (if any) are you using to isolate the driver from the printed parts? I've seen Blu Tack used quite frequently headphone modding, I wonder how effective it is.

How big are you printing holes that will be threaded?

Tofty Headphones 4 are already pretty close to my ideal DIY headphone, but I have a few of my own ideas I want to integrate.

• Angled drivers, no real reason, I just think they're cool. This should be fairly easy to implement.
• Driver and jack removable as one without desoldering, this should also be fairly easy.
• Clean up the yoke-cup connection and use printed pins, with some sanding I think this is doable.
• Completely 3D-printed headband and slider mechanism, now this is difficult. If the headband is too thick it'll be far too stiff, but I need some thickness for the height adjustment mechanism to work. I think this'll be an interesting challenge. If this fails I'll likely fall back to the Grado-style rod mechanism, but with a more elegant swivel limiting mechanism.

Thoughts, ideas, advice? All appreciated, thanks.

 

[Tofty]

Thanks Johannes,

The wiring will depend on what type of split cable you plan on using, as to what connections should be made. If the cable has mono (TS) connections to each ear then the Positive wire should go to the Tip and the Ground should go to the Sleeve of your TRS jack for each ear cup.
Other types of split cables exist that may require different soldered connections to be made. For example; if the cable has fully connected TRS connections for each split side (this type of cable allows the split ends to be fitted either way around on the earcups and still maintain the correct stereo direction, as it provides the stereo signal to both sides) then the left cup should be soldered + to Tip, - to Sleeve, but the Right side should be + to Ring, - to Sleeve.
I'm not sure is some cables may use TRS jacks but be wired mono, so watch out for that I guess.
The driver will have a dot on it which identifies which solder pad is the positive side.

Let me know what cable you plan on using and I'll have a better idea what will work.
I can do a wiring diagram if you think it would be useful.

I didn't use any isolating material in front of or behind the driver, but I did use some closed-cell foam around the sides of the driver. No idea really as to whether it helped or not.

I've tried angled drivers, but it makes designing and printing the parts a real pain and I'm not confident it's entirely worth it.
Having a load of surfaces at 10 degrees-ish from others makes printing without lots of supports very difficult and the surface finish is pretty bad even with supports.

So long as you use the 3.5mm panel mount jacks that mount on the inside and are retained with a nut on the outside, then removing the driver and jack as one piece should just be a matter of slotting the wiring pass-through hole on the main earcup part. Being able to do all the required soldering without having the parts in situ is definitely a good idea.

Simpler and potentially integrated pins on the yolk should be doable. Perhaps something like a small rubber grommet in the earcup would make a neat way of giving some resistance and making up for any irregularities in the pin's roundness.

Good luck with the headband. I've tried a few things and none have been particularly successful. I'm sure what you suggest can be done, but it will likely require a lot of fettling and iteration to get right.

 

[castleofargh]

Beside the obvious struggle to keep things light, comfy but sturdy enough, and having to go through trial and error just to find the right clamping force, I also have the almost entirely baseless belief that PLA isn’t good for a headband. Or really anything with some serious mechanical force applied on it as it will just bend over time.

 

[Johannes P]

Hi, thanks for the reply.

The cable I plan on using is "OPENHEART 4 Core High Purity Silver Cable For Headphones 1 to 2 jack 3.5/2.5/4.4mm Balance Earphones Upgrade Replacement Cable" on Aliexpress, not sure if I'm allowed to link so I won't. It has TRS connectors on the split ends, but if I understand correctly, it is wired as mono.
The description says that "The cable will not fit all headphones, there is a way to confirm whether the cable is suitable for your headphones: Please swap the left and right side of your headphones cable. If it still works and the left and right sound channel are also exchanged, our cable will be suitable for your headphones"

I think you're right about the angled drivers, I've spent a good 15 hours in the past 2 days trying to come up with a design that's easy to print and aesthetically pleasing, and I think I'm ready to give up on it. Angled drivers also introduce 2 additional problems which I didn't initially think of. They make construction of the baffle a whole lot harder without it looking super janky, and it adds to the cup thickness, which I want to avoid.

The obvious solution for making the jack and driver removable as one is to just have a hole big enough for the jack in the main cup, but that's quite unsightly. The solution I came up with is to have an additional cover plate with a hole only big enough for the wires, it'll slot into place and be secured to by the outer ring. Adds slight complexity but should be easy to print, we'll see.

 

[Tofty]

I suppose the benefit of a PLA headband is that even if it only lasts 6 months; you know you can print a replacement and it will cost relatively little.

Personally I like the idea of integrating easily sourced non-printed parts into designs where possible.
I see a trend of people wanting to 3D Print every possible part, which is a fun technical exercise and were I in a situation with only a printer, I'd be grateful that people have pioneered the printing of parts you wouldn't expect. However, it can't really be expected that a printed spring for instance will be anywhere near as durable as a metal one.


I would recommend getting a good multi-meter. They're really useful for all aspects of headphone wiring.
It will allow you to check how the cable has been wired and adjust the jack configuration accordingly.
My assumption would be that since they're using TRS connectors that they have gone with tip + and sleeve -, leaving the ring unconnected.
It's possible that they might have wired the tip and ring together, but I doubt it as that might short the positive and negative if used in a mono wired set of headphones.

 

[Johannes P]

Yeah, I get that printing everything is not always the most practical option (especially for long term use), but the problem is that some parts can only be sourced for a reasonable price from China, which means agonizingly long delivery times. Last time I ordered a cable from Ali, it took a total of 41 days to arrive, though things have likely gotten better since then.

Using a multimeter is great tip, thanks! I can borrow one from work and check the cable wiring, should the cable ever arrive.
The more I work on these headphones, the more apparent it becomes how little I can improve the design. And I don't think that's my incompetence, the design is already really good. It's a bit disheartening that I'm basically just copying the design at this point, but hopefully they'll at least sound great.

 

[Johannes P]

Hello, me again.

I decided to persevere with the idea of angled drivers for no real reason, and I'm happy to report I finally have something satisfactory. Not much is left of the original Borealis design, this is more of a mash-up of HD 800 and Utopia, I quite like the design though.

The main driver cage was printed as one, I sanded the surface the driver clamp mates with to get rid of the "stair steps" of individual layers. The grille was printed as a separate piece and simply snaps into place. The masking tape is there just as a mock-up of the actual baffle design.

 

[Tofty]

That's a really nice design.
What resolution are you printing at? That came out so much cleaner looking than my prints have.

I'm looking forward to seeing it completed and hopefully get a chance to make a pair myself.
What pads are you planning on using?

 

[Johannes P]

Thanks, I printed it with 0.2mm layer height, which I think is a good middle ground between quality and print time. I'll share the files when I have a finalized, hopefully less janky design. I still have to experiment with pads, Beyerdynamic 770/880/990 pads would be perfect aesthetically as they're quite slim, but they don't sound very good to me, too bright.

 

[Johannes P]

I can't believe it's been over 3 months already.

I've done a lot of work since the last update, and TL;DR, I scrapped the entire round cup with angled driver idea due to a multitude of crucial issues.

Firstly, having the driver rather far away from the ear and angled seems to result in a nasty spike in the treble. I don't have a measurement rig to make any objective statements, but it isn't pleasant. Could I just EQ it out? Yeah, but that's cheating and terrible design. The second issue is attaching the baffle both cleanly and consistently enough not to have noticeable channel imbalance in the bass. The cup I had was just plain difficult to work with. I switched from metal mesh to nylon aquarium filter which is easier to work with, but doesn't solve the issue entirely.

The straw that ultimately broke the camel's back though is that PLA yoke pins are a terrible idea. I ruined multiple 3+ hour prints by having the pins shear and get permanently stuck in the holes, and the way I designed the cup doesn't allow to integrate a nut and screw system without looking awful. If anyone wants the CAD files, I'll share them, but I truly do not believe this is overall usable without a complete redesign.

Also, a PLA headband is just not usable, period. Even if you get the geometry perfect, it wears out in less than a week. PETG doesn't fare much better, I haven't tried other materials, I just ended up buying the Beyerdynamic bow, it works.

So what did I do? I started from mostly scratch. I went for a 110x90mm oval cup, no angled driver. A rather simple sandwich style construction but I think it works quite well. Rather than trying printed yoke pins again, embedded nuts and screws are used for fastening. Having a flat baffle makes attaching it significantly easier, who knew? Here's a few pictures, sorry they're rather poor, and the headphones are obviously still very much WIP.

 

[558416]

I can't believe it's been over 3 months already.

I've done a lot of work since the last update, and TL;DR, I scrapped the entire round cup with angled driver idea due to a multitude of crucial issues.

Firstly, having the driver rather far away from the ear and angled seems to result in a nasty spike in the treble. I don't have a measurement rig to make any objective statements, but it isn't pleasant. Could I just EQ it out? Yeah, but that's cheating and terrible design. The second issue is attaching the baffle both cleanly and consistently enough not to have noticeable channel imbalance in the bass. The cup I had was just plain difficult to work with. I switched from metal mesh to nylon aquarium filter which is easier to work with, but doesn't solve the issue entirely.

The straw that ultimately broke the camel's back though is that PLA yoke pins are a terrible idea. I ruined multiple 3+ hour prints by having the pins shear and get permanently stuck in the holes, and the way I designed the cup doesn't allow to integrate a nut and screw system without looking awful. If anyone wants the CAD files, I'll share them, but I truly do not believe this is overall usable without a complete redesign.

Also, a PLA headband is just not usable, period. Even if you get the geometry perfect, it wears out in less than a week. PETG doesn't fare much better, I haven't tried other materials, I just ended up buying the Beyerdynamic bow, it works.

So what did I do? I started from mostly scratch. I went for a 110x90mm oval cup, no angled driver. A rather simple sandwich style construction but I think it works quite well. Rather than trying printed yoke pins again, embedded nuts and screws are used for fastening. Having a flat baffle makes attaching it significantly easier, who knew? Here's a few pictures, sorry they're rather poor, and the headphones are obviously still very much WIP.

@Tofty @Johannes P c̶a̶n̶ ̶y̶o̶u̶ ̶c̶o̶n̶v̶e̶r̶t̶ ̶a̶l̶l̶ ̶o̶f̶ ̶t̶h̶e̶ ̶p̶a̶r̶t̶s̶/̶d̶e̶s̶i̶g̶n̶s̶ ̶t̶o̶ ̶a̶u̶t̶o̶d̶e̶s̶k̶'̶s̶ ̶T̶i̶n̶k̶e̶r̶C̶A̶D̶?̶Are you guys able to explode your designs into the seperate parts and export as STL/OBJ? It allows for unexperienced users to export them into TinkerCAD to modify designs.

Circles made in TinkerCAD are limited to 64 sides so you should import a circle from some other programme.

Maybe for the headband consider resin? I barely know stuff about 3d printing and materials.

 

[Johannes P]

can you convert all of the parts/designs to autodesk's TinkerCAD? Its an extremely basic website for modelling that an 11 year old can understand.

I can export as STL, which can be opened in TinkerCAD, but why? If you want the files for the round cup headphone, I can provide them, but honestly I'd recommend just remaking it from scratch. It'll be less of a pain than dealing with the shortcomings of the current design. The oval headphone is still very much WIP and I'm making some pretty major changes to fix the issues from the previous iteration. I'll think about sharing the files when it's done.

As for using resin, that would require a resin printer, which I don't have. The Beyerdynamic bow isn't ideal but it's good enough.