My DIY planar headphones (WIP)
Jul 20, 2020 at 11:21 PM Thread Starter Post #1 of 25

IKillPCParts

New Head-Fier
Joined
Jun 26, 2020
Posts
16
Likes
10
Location
England
I just want to preface this with a little paragraph.

I literally do not know what I am doing. I read a little on how planars work, that's the information I've started with. I've seen the DIY planar thread and have taken some inspiration from that to design my own set of drivers, and the headphones around them. Other than the information I've gotten from innerfidelity and that thread linked earlier, I have nothing else. So. Let's fail miserably at this. My initial design goals simply was "make a planar driver" but those have now evolved into "make a set of headphones with planar drivers, and do it on a budget*". I say budget, but, this would require a 3D printer and those can be decently expensive unless you sink a lot of time into it.

So, I started with an initial circular design for the driver, to be mounted in a casing snugly and then having the earpads attach like Brainwavz XL pads on the AKG K612 Pros (which is what I'm using now). Dimensions were a 106mm diameter and 15mm thickness, with the active area being approximately 5200mm2. This design had no area to add in foam/wool/whatever for sound tuning.

3785298eeffc74319b42d6083bb3eb7c.png

Decided not to go with that design because I asked a question about using two 35mm magnets instead of one 70mm magnet and learnt that it would affect audio quality, so I scrapped that. I also realised that having both sides of the driver obstructed with a dense-ish grill like that wouldn't exactly be great for sound (I think?). So I pulled some parts and started messing with V2 of the design.

This time around I decided to ignore any form of symmetry from the front to the back of the driver, instead actually caring about what the structure could do for the sound. So I placed the plate that would end up holding the diaphragm closer to the ear, and shrunk the distance between the driver and the magnet structure plate thing. The diameter remained at 106mm, but the thickness ended up at 20mm, and with an active area of approx. 5350mm2. This time around, I added layers to allow for a 1mm thick layer of wool or felt or similar material to be seated.

bac79201d72535bf4cbab462440ff498.png

I was more than happy with this design until I actually looked at it slapped together and realised, "that's a lot of layers". I would have to rely on my printer being accurate enough and have a good enough surface quality/finish way too many times for something like this. So I iterated and created a V2.1 with minimal changes other than removing unnecessary layers.

480cf908d32ae2f0bd0352aaa29a0336.png

V2.1 was nearly identical to V2 but had much less layers for my printer to screw up on. I was mostly happy with this (having printed a V1 by this point, and now I was testing how to adhere any sort of diaphragm material to the designated holders).

However, because I am rather impulsive with my designs sometimes, I found myself watching a few youtube videos and I came across the Acoustic Research AR-H1. I loved the rounded rectangular design, the grill also fascinated me. I found that I immediately preferred that over what I had already designed. So I set about designing a V3, which so far is the design I've stuck with. With driver dimensions of 105*85*15.27mm (LxWxH) and an active area of approx 4440mm2, I found this one to be the most aesthetically pleasing really, and went forth with settling on this as my final driver design (for now).

f5018f1723341550a9b30b6db9ba07ab.png

V3 was mostly a reaction to seeing Acoustic Research's incredible looking headphones. I did decide to design the casing around the driver though, so I worked on this first. The separated parts that stick out some from the actual driver body are TPU pieces that'll be embedded into the PETG casing whilst printing, to allow the driver some vibration dampening from the headphone casing itself. I could also pack foam around the driver to make it more secure in case the screws aren't enough. I also did a little more research, coming across stuff like 'fasers' or whatever that I decided to implement on the coverings for the magnet array. Speaking of, I also decided to go with a single-sided magnet array mostly due to cost (I got quoted £300 for a set of 70*4*2mm magnets and I immediately thought nope, and I ended up finding 55*4*4mm magnets from a German seller on eBay which I plan on using 7 of per driver). Honestly, a fair share of the motivation for a new driver came from the fact that 70mm long magnets that were thin were nearly impossible to find anywhere unless I talked to some people on alibaba or whatever. And I wasn't a huge fan of that.

Now that I had an actual driver design that I might be happy with, I could start work on a case. My design goal was simple enough, something 3D printable that wouldn't take too long, and be decently sturdy. So I came up with this.

1595300764709.png
1595300805058.png
photo_2020-07-20_01-03-35.jpg
IMG_20200721_041024.jpg

The casing is printed from Prusa's Prusament PETG in Galaxy Black, and I absolutely love its colour. The inner driver holders are printed in some generic TPU and are flexible to allow an assembled driver to be pushed into place. Wiring is run from the back of the casing (side with the grill), through the rectangular hole in the outermost driver mount, and down towards the little extensions where 3.5mm mono jacks are mounted.

After designing the casing, only then I realised that I should actually get myself some material for a diaphragm. So, after scouring the internet for much too long (how many listings for "aluminised mylar film" exist that aren't actually thin enough?) I finally found myself a good source, and I've bought myself 2 rolls of the 5 micron aluminised mylar. I'm sure I would've been pointed to them near-instantly if I made a post but.. That didn't cross my mind. This stuff, I'm going to etch with eBay-bought chemicals into a diaphragm design I have already created and test-printed. I'll post that when I do that.

Now, I still have plenty a ways to go. I need to design a hinge for each side (I'm going to have a look at the Dan Clark Aeon 2's, I absolutely love that hinge design), a way to make that interface with a headband, and, well, a headband. And I'm sure I'm missing a lot of things, but hey, I'm a 19 year old uni student with way too much time and enough ambition to drive stupidity like this. And sorry if this reads like it was typed without proof-reading, it's 4am and I'm sure if I don't do this now I wont ever.
 
Last edited:
Jul 21, 2020 at 11:27 AM Post #4 of 25
So here is the million dollar question. How are you going to apply the etch resist?

Inquiring minds would like to know : )
Etching is going to be an issue, especially if you are after very thin traces. It might payoff to look for a supplier that offers a solution.
I'll be honest, I didn't think much about that just yet. If anything I'm doing everything reactionary and only finding out what to do when I need to do it. So, from quickly looking around, I'm gonna try toner transfer without heat (so acetone and water mixture). If that works, neat, I'll post about it. If that doesn't work, also neat, I'll post about it and look for another solution.
 
Jul 24, 2020 at 10:57 AM Post #6 of 25
So all the parts for one driver have been printed and I need to make some modifications. Also, pulled out the colour printer (an Epson XP-345 or whatever) and discovered that it needs more ink. So I've bought some cartridges for that too.

IMG_20200724_154644.jpg
IMG_20200724_154650.jpg

Modifications to the driver include making the layer that's closest to the ear (visible in the top picture) 0.6mm thinner to actually sit flush with the casing, and I'm most likely going to shrink the casing to be a little thinner and therefore make the topmost red layer (visible in the bottom picture) sit much closer to the surface instead of being embedded like it is currently. However, I do not know how that's going to affect the sound (it shouldn't affect it too much, right? Currently the gap would support some foam or wool to help with sound but.. I'm not too sure how that'd affect sound either. I've read around and it's something about simultaneously reflecting sound back to the ear and also letting some through?). But I digress. For now, I'm going to make the changes to the driver and reassemble. Also, as a mild side-note, currently the whole assembly weighs 100g, but of course that's going to increase with the screws and nuts that'll be added to assemble the driver.
 
Jul 24, 2020 at 4:41 PM Post #7 of 25
Ok scratch the colour printer, the laser print in house will do this much better. After trying and partially successfully sticking a sheet of mylar to paper and running it through the laser printer, it easily printed onto the mylar with no problems at all. Using that I can adjust what I am printing and print out the etch resist directly without needed to jump through hoops with toner transfers and whatnot. However I still need to actually test etching it, so that'll be the next step when the ferric chloride arrives, and after I've found a better way of adhering the mylar to the paper.
 
Jul 25, 2020 at 12:40 AM Post #8 of 25
Ok scratch the colour printer, the laser print in house will do this much better. After trying and partially successfully sticking a sheet of mylar to paper and running it through the laser printer, it easily printed onto the mylar with no problems at all. Using that I can adjust what I am printing and print out the etch resist directly without needed to jump through hoops with toner transfers and whatnot. However I still need to actually test etching it, so that'll be the next step when the ferric chloride arrives, and after I've found a better way of adhering the mylar to the paper.

Glad press and stick is your new best friend. It's basically super thin tacky kitchen wrap.

IDK if you have it in England, but if you do, you should be able to buy it at a well supplied grocery store.
 
Jul 28, 2020 at 8:01 PM Post #9 of 25
Ok so, after some testing and multiple printing trials, I've found a good way to stick the mylar to the paper, but I haven't found a proper way of getting a good etch resist yet.

I've streamlined the process of applying the mylar film to the paper and it goes like this:
1) On a flat, clean surface (I'm using a glass panel) apply soapy water (though with much more soap than water). This is used to 'stick' the mylar film to the glass and allow much easier working.
2) One the mylar is 'stuck' to the glass, use a credit card or something similar wrapped in paper towels to remove creases and bubbles in the film, doing so and repeating until as many bubbles/creases are removed.
3) Once done, apply some (not too much) glue mixture to the film, spreading it out evenly with a sponge. The glue mixture is a 1:1 mix of PVA glue and water.
4) Then, slowly lay out the paper (as if you were applying a screen protector to a phone, start at the bottom and slowly lay it up). As you're laying up the paper, use a credit card or something similar to remove any air bubbles that form.
5) After laying down the paper, I used a hair dryer on gentle airflow but the hottest setting to go over the paper and make sure the glue mix is set.
6) Now I cut out the shape of the paper from the mylar, and I end up with mylar-covered paper.

1595980711837.png

Now is the part where I'm stuck. I've tried using the laser printer, a colour printer, a laminator for toner-transfer, a clothes iron for toner-transfer, and acetone for cold toner-transfer. All partially failed, or fully failed. The closest to success so far was directly printing onto the mylar with the laser printer, but even then the coverage over the lines I want is spotty, therefore if I were to etch using that it'd fail. So I'm a little stuck right now on that part. Otherwise, I've been working a little more on the hinge aspect and have figured out mostly how to put that together, I just need to turn it into a model I'm happy with.
 
Jul 29, 2020 at 2:20 PM Post #10 of 25
I hate to put it this way, but join the club.

A lot of people (including myself) have been looking for a cheap, easy, and consistent method of making aluminum traces.

My approach at the moment is to use a vinyl plotter to cut out aluminum foil. Then to glue that foil to the mylar. But it isn't working all that well.
 
Jul 29, 2020 at 2:29 PM Post #11 of 25
I hate to put it this way, but join the club.

A lot of people (including myself) have been looking for a cheap, easy, and consistent method of making aluminum traces.

My approach at the moment is to use a vinyl plotter to cut out aluminum foil. Then to glue that foil to the mylar. But it isn't working all that well.
Ah well. I'll find a way sooner or later (I hope!). For now though, the failures did let me know that at least one part works. With the 1:1 pva and water mix, even letting the smallest amount of water touch the paper near-immediately makes the glue lose its stickiness so the film peels off easily. So hey, there's a win in that.

I'm tempted to try a few different things to make etch resists. Maybe I could 3D-print the traces that I want as a thin layer and then glue that to the film as the etch resist? Maybe I could just straight up spend a few hours drawing the etch resist on with a sharpie or some other specialised pen. Or, a recent video I've watched used the backing from vinyl to print the traces onto, then using that for the toner transfer. Worked really well for them.

Regardless I have a few other things to work on if I have to pause this for a moment whilst materials come in. I've got 3 PCBs coming in for a 12AU7 Starving Student, and some wiring coming in for a custom set of cables for this pair of headphones.
 
Last edited:
Aug 2, 2020 at 4:37 PM Post #13 of 25
Just a thought: since you can choose the shape, I'd totally go with oval instead of circular. Is there any reason why you want them that way?

Oval is good for electrostatic headphones. Electrostatic diaphragms are under a lot of tension which can cause unwanted resonances. The oval shape uses mechanical advantage in order to help mitigate those resonances.

For planar headphones, the diaphragm is not under a lot of tension and therefore the resonances of the diaphragm tend to be out of the audio frequency range.

We are much more concerned with getting a true planar wave front which means we want the diaphragm to move straight in and straight out with as little warping as possible. An oval shape wouldn't be ideal since the stresses on the diaphragm would be less consistent than a circular or even rectangular diaphragm.
 
Aug 2, 2020 at 4:49 PM Post #14 of 25
Oval is good for electrostatic headphones. Electrostatic diaphragms are under a lot of tension which can cause unwanted resonances. The oval shape uses mechanical advantage in order to help mitigate those resonances.

For planar headphones, the diaphragm is not under a lot of tension and therefore the resonances of the diaphragm tend to be out of the audio frequency range.

We are much more concerned with getting a true planar wave front which means we want the diaphragm to move straight in and straight out with as little warping as possible. An oval shape wouldn't be ideal since the stresses on the diaphragm would be less consistent than a circular or even rectangular diaphragm.
I see. I'm too much a noob to argue that, but my point is about ergonomics, mostly, considering our heads are ovals. I've always prefered the comfyness of senns over ath and beyers, being the latter my faves, though. could there be a middle ground? Beyers always leave a little gap behind my ears, so the seal is always an issue.
 
Aug 2, 2020 at 6:48 PM Post #15 of 25
I see. I'm too much a noob to argue that, but my point is about ergonomics, mostly, considering our heads are ovals. I've always prefered the comfyness of senns over ath and beyers, being the latter my faves, though. could there be a middle ground? Beyers always leave a little gap behind my ears, so the seal is always an issue.
I mean honestly, I could try that D-shaped sorta thing that MrSpeakers has, I think that looks quite nice. The only reason I'm going with this rectangular design is because, well, the AR-H1s look amazing and I wanted to copy them. Well, not just carbon-copy, but it's influenced the design a whole bunch. If anything though, any ergonomic downfalls this design has can be helped with some custom earpads.

Also, just a mild update, I'll be trying proper branded toner transfer paper instead of the knock-off stuff that I bought on eBay, and if that doesn't work I'll be trying actual photoresist film. If I use the latter I'll be modifying my method of getting a flat film surface, instead of using paper I'll back it with an acrylic slab so I can apply all the liquids needed without fear of the film lifting off and wrinkling even more.
 

Users who are viewing this thread

Back
Top