Thoughts on developing an open-source diy headphone chassis
Oct 5, 2019 at 5:07 AM Thread Starter Post #1 of 13

ericj

Headphoneus Supremus
Joined
Aug 2, 2005
Posts
8,270
Likes
169
Somewhere between homebrewheadphones.com and some of the AKG style and Grado style parts on thingiverse i got thinking about making an entire headphone chassis at home from a minimum of different kinds of parts.

Largely 3d printed. Modular. Lending itself to endless tinkering, so a minimum of glued or easily worn out parts where it matters.

For a while i was thinking about a chassis inspired by Grado and a little bit by Stax.

In the old SR-80 and i suppose a lot of grados that i don't have first hand experience with, the driver gets glued into a tube section that gets glued into a larger tube section and the yoke connects to the larger tube section with some pins that fit into a hole in the yoke and a hole in the tube and have a little integral spacer. It's likely that the total length of the resulting assembly and the flare at the back are a bit like a bass port on a reflex speaker.

The Stax SR-30 works a lot the same way. It's an outer cup and an inner cup. The outer cup includes the baffle to which the driver is adhered. The inner cup has a grille and retains any and all damping materials. The difference is that the only thing that holds the inner and outer cups together is the pins that connect to the yoke.

So i think if i were designing 3d-printable Grado-style headphone cups, I would go with the SR-30 style assembly where the pins hold it together.

I have a little bit of experience with both fused deposition (filament through a hot nozzle) 3d printing and liquid resin 3d printing. Long skinny objects that have really smooth surfaces and don't snap easily are not really something 3d printers are great at. If you're making thousands of something, you have injection molding to lean on, as well as companies that will crank out steel rods with chrome plating and knurled ends.

If you're making ONE of something, carbon fiber rods and tubes aren't expensive on aliexpress and probably glue really well to 3d printable plastics with the right adhesives. You can cut them easily with a dremel. So in a Grado style chassis, cheap carbon fiber rods can provide the pin that connects the yoke to the cup and the pin that connects the yoke to the headband. https://www.aliexpress.com/item/32573564619.html

A 3d printed yoke has to be a bit beefier than one that is injection molded, but that is not a big deal.

The arch of the SR-80 style chassis is a metal band with a cover that is largely cosmetic, and the blocks that connect that to the pins that come off the yoke. The blocks are super easy to print. The metal band is a piece of cake if you have a few hundred bucks worth of sheet metal working equipment, or maybe you can get stainless steel hair bands on aliexpress and perhaps all you have to do is design the block to clamp onto them, or maybe cut them to length and drill a hole in each end. Like these. https://www.aliexpress.com/item/32913953633.html

3d printed headband arches exist but you pretty much have to print them in ABS for long-term durability. Or maybe HIPS or nylon or some other more demanding material. PLA will work great for a while and either break all at once or flow over time until you have no clamping force. PETG is too floppy in my experience, and you can't glue stuff to it easily.

And that's where i come to the K240 style chassis. Or maybe more specifically K270 style. Grado style might be ideal for supra-aural but perhaps isn't ideal for circum-aural.

Maybe after knocking up some jigs, the K240 style wire headband is super easy to make out of the copper-plated TIG welding rods specified for mild steel that are crazy cheap and available in a multitude of diameters anywhere metal exists. I mean you bend it into a ring, take two rings and put a couple more bends into them. Cover it with tubing of . . . . some kind.

AKG injection molds the plastic blocks that connect the outer headband to the cups. I've bought a lot of old and broken headphones in the K240/141/270 family and sometimes one of the wires breaks out of the block. Usually some well-mixed, high quality epoxy provides a durable fix, and i don't see why i couldn't 3d print a block that just has a channel for the wire on each side and a jig to hold the parts until the epoxy has cured.

The K240 style cardan joint is admirable but puts some limitations on the diameter and depth of the driver assembly. It's great if you have a small to medium diameter driver. But i like orthos and electrostatics too and they are often large diameter and picky about what is directly behind them.

The K270 chassis affords a much larger cup volume. It just has a vertical pivot, and relies on the flex of the headband for the horizontal tilt, though it wouldn't be difficult to build in a few degrees of toe for better head fit.
 
Oct 6, 2019 at 10:52 PM Post #2 of 13
As someone who was trying to do something like this myself, I can very much get behind this idea.
 
Oct 6, 2019 at 11:02 PM Post #3 of 13
As someone who was trying to do something like this myself, I can very much get behind this idea.

The more i think about it the more i think a grado style frame will work best for supra-aural, and for circum maybe i can design a cardamatic suspension copy. That patent has to have expired by now, right? Or whatever.

It might be possible to make an AKG type arch with music wire but you'd have to heat treat it. Sounds like a hassle.

There are shops that can laser-cut leather or other materials to order for headband covers.
 
Oct 7, 2019 at 3:01 AM Post #4 of 13
I've done a bunch of 3d printed cups. The headband and a source for good drivers is what is hard to find. Usually you end up with a doner pair of cans that have decent drivers that can be pushed much further. As such you just use the head band that came with it as well. Problem is those head bands usually suck. What material can be used for a clamp in at home diy fassion without heat treatment? 3d printing could work, though I would prefer a metal clamp. You could make a mold and do a fiberglass band or even carbon fiber.
 
Oct 7, 2019 at 1:05 PM Post #5 of 13
The more i think about it the more i think a grado style frame will work best for supra-aural, and for circum maybe i can design a cardamatic suspension copy. That patent has to have expired by now, right? Or whatever.

It might be possible to make an AKG type arch with music wire but you'd have to heat treat it. Sounds like a hassle.

There are shops that can laser-cut leather or other materials to order for headband covers.
I'll need to think about it, but the designs of the HD558/598, Verum One, and anything from HIFIMAN, are inspirations for me. The HD558/598 gave me the idea for angled cups and strategic driver placement, Verum for some aspects of my hypothetical headband (I'm optimizing his design), and HIFIMAN for cup shape.
 
Oct 7, 2019 at 6:57 PM Post #6 of 13
I've done a bunch of 3d printed cups. The headband and a source for good drivers is what is hard to find. Usually you end up with a doner pair of cans that have decent drivers that can be pushed much further. As such you just use the head band that came with it as well. Problem is those head bands usually suck. What material can be used for a clamp in at home diy fassion without heat treatment? 3d printing could work, though I would prefer a metal clamp. You could make a mold and do a fiberglass band or even carbon fiber.

For the moment, you can get a 50mm Peerless headphone driver from Parts Express that probably isn't bad. There are also soviet ortho drivers on ebay.

There are mixed reports of how good or bad some of the aliexpress drivers are. Some claim that they are factory rejects sold by the pound. Some say that the MDR-V6 drivers on aliexpress appear to be legitimate. I ordered a pair of the alleged Onkyo headphone drivers and i have 3 examples of chinese 40mm drivers here.

I think some of the frustration with bare drivers from china stems from not having any data on them and from a belief that as much of the front of the driver should be exposed as possible without the wisdom to accept that it's quite possible that you will have to occlude most of the face of the driver to get it to sound good. The KSC75 doesn't sound good in spite of all the plastic in front of the diaphragm, all that plastic is part of the tuning.

There are good-sounding electret headphones with good bass where the whole face of the driver appears to be blocked by a phase plug.

I think some people will appreciate the project as a way to DIY from a known-good design, like the one offered at homebrewheadphones.com, and some people will appreciate the project as a vehicle to learn the ins and outs of tuning a headphone system. And then others just want to find a home for a set of high-end drivers but it turns out that there are some very pretty turned wood headphones on aliexpress for under a hundred bucks and you could just start with that.

I ordered qty10 of the aliexpress "hair bands" - they're the type of thing that novelty antlers and tiaras get built on.

I suspect that they may work just as well as the steel band that Grado uses on at least their lesser headphones. But i don't have them yet so i can't say for sure.

Since they have the little curve at the end i wonder, if the length is right, if they can just be clamped into the blocks that accept the vertical coming off the yoke.

Maybe for bigger cans you use two.

I'm sure that headbands printed in ABS can work. In fact there's an entire Stax Sigma headband assembly on thingiverse. I'm less sure that printed PETG headbands will work well. PLA printed headbands will work for a while until they don't and heat treatable PLA will work better. Printed in PA (Nylon), ASA, some of the other exotics (polyalchemy elixir, etc) will probably work fine.
 
Oct 7, 2019 at 8:11 PM Post #7 of 13
If you want to buy known-good drivers without buying other whole headphone systems, you can get K702, Q701, or K712 drivers under $40/ea. T50 drivers are something like $50/ea. The beyer drivers are more expensive perhaps because they also come attached to the baffle? The DT770/880/990 drivers are under $70ish each.
 
Oct 8, 2019 at 3:42 AM Post #8 of 13
I'll need to think about it, but the designs of the HD558/598, Verum One, and anything from HIFIMAN, are inspirations for me. The HD558/598 gave me the idea for angled cups and strategic driver placement, Verum for some aspects of my hypothetical headband (I'm optimizing his design), and HIFIMAN for cup shape.

Yeah there are a lot of headphones out there that have angled and strategically placed drivers. A whole lot of Audio-Technica's high end stuff and Ultrasone S-Logic stuff come to mind.

As for Hifiman and Verum, I guess you're referring to big cylindrical cups and a half-circle yoke connecting them? It's easy enough to do but if you are 3d printing it the yoke will have to be a lot chunkier. The metal headband is hard to make at home (got spring steel stock? got sheet metal working gear?) but Hifiman sells an upgrade headband for all their phones for like $80.
 
Oct 8, 2019 at 7:38 PM Post #10 of 13
I have been looking for ortho drivers all over the place and haven't seen anything but ribbon tweeters. What search terms are you using to find these soviet ortho drivers?

I literally typed "orthodynamic headphone driver" into the ebay search bar.

I have a pair of 60mm drivers on their way from kiev right now.

You can also just buy Echo H16-40C headphones (or other variations on the TDS-16) and take the drivers out. Apparently they have dozens if not hundreds of these sitting in a warehouse since the early 90's.

https://www.ebay.com/itm/Hi-Fi-Orth...YH-1-ECHO-NEW-Soviet-USSR-TDS-16/233316905806
 
Oct 11, 2019 at 10:46 AM Post #11 of 13
I have a pair of 60mm drivers on their way from kiev right now.

You can also just buy Echo H16-40C headphones (or other variations on the TDS-16)
Are they any good? I have done some T50RP based cans (open alphas, wood cup varients ect) and gotten to the point of opening the drivers and doing mods to them. Really taking those to the limit of what they can do. I'm looking for better drivers. How do the TDS-16 drivers compare to the T50RP ones?
 
Oct 12, 2019 at 2:20 AM Post #12 of 13
Are they any good? I have done some T50RP based cans (open alphas, wood cup varients ect) and gotten to the point of opening the drivers and doing mods to them. Really taking those to the limit of what they can do. I'm looking for better drivers. How do the TDS-16 drivers compare to the T50RP ones?

The TDS-16 uses 50mm drivers that are inspired by the Peerless/MB ortho drivers of the 80's, which is to say that the drivers have round perforated ferrite magnets and a diaphragm that is pinched in the center. PMB/TDS drivers have a loose-but-flat diaphragm, unlike the Yamaha drivers which have the same kind of magnet and electrode design but an concentrically embossed diaphragm.

So they don't have a ton of bass. They have some bass, and it's good, but it aint a lot and the extension will disappoint bassheads.

The 60mm TDS drivers probably have better bass than the 50mm TDS drivers. Or maybe they don't. I'll know when they get here.

I generally prefer a flat, analytic response without big bass, and the TDS-16 / H16-40C gave that to me with just some poster tak to seal the driver to the baffle. I also put some tape around the pivots so that they don't knock around and make noise.
 
Last edited:
Oct 13, 2019 at 2:29 PM Post #13 of 13
I just got my first 3D printer, which really opens up the possibilities! Everything is easily achievable in this project; it's really the driver that is a stumbling block to DIY. I'm new here, so maybe I have not yet stumbled across a thread that details how to achieve a quality DIY driver?
 

Users who are viewing this thread

Back
Top