If you have to possibility to built yourself a headphone's driver, what will look like?

Electrostatic.

I am afraid (based on a long history of failures in DIY efforts on these forums) that forming the "cones" on a dyanamic headphone driver and attaching the coils to them is FAR beyond the means of most small operations.

Quote:

Originally Posted by nikongod 

Electrostatic.

 

I am afraid (based on a long history of failures in DIY efforts on these forums) that forming the "cones" on a dyanamic headphone driver and attaching the coils to them is FAR beyond the means of most small operations.

I agree dyanmic drivers will be problematic for DIY.  

You might have already seen this video but it might offer some insight.  Looks like akg uses a UV curable glue.

http://www.youtube.com/watch?v=3LbsQRaQD6A

It's a huge pain to make your own headphone drivers. I started doing it last year, and stopped working on it. The coil equations are found easily enough to calculate impedance. The wire is easy to buy on eBay and through other sources. The problems I had were with the magnets. They were all either too big, or too small. I found the magnet's strength, the weight of the adhesive used, wire gauge, coil diameter, air gap, and even the coil height, all contributed to the drivers damping, efficiency, and overall quality. If I wound too small of a coil, it had a really low impedance and no amplifier was capable of driving it. I didn't get too much into figuring out what was actually going on. Just finding light enough bobbins, and winding mechanisms was hard enough. My final plan was to use silk as the spiders, and cut some baskets from wood. Keeping the coil centered around the magnet was another problem. If I put too much signal onto the coil, it jumped right off of the magnet. It's a big project.

I still think the idea of making hand wound drivers is cool, probably far more work than it's worth though. How will you handle returns if someone blows a driver, and wants their money back? My plan was to also sell a few sets per year. The possible returns was something I did not want to deal with. It's also highly unlikely anyone was going to pay me the $2k per set I might have had to charge to make my money back.

I would replicate the K1000 driver, or even improve it.

Then sell as spare or upgrade parts to all those who want a pair. :k1000:

I'm in no way attempting to discourage you, but it's a TON of work. When I worked for a speaker company (a pretty well known one too!), we spent years developing drivers. The magnets used to come in neodymium slugs from some Asian country, and we had a monster of a machine which turned the neodymium into a permanent magnet. Depending on the amount of power you want to cycle through this device, you may need to worry a lot, or a little about cooling. Some manufacturers use various types of fluids to keep the coils from melting down. Others will use lots of air around the whole driver mechanism. I've even seen drivers built with small heat sinks on them.

One thing which may be useful, is picking a magnet first. It's tough to induce a large enough B-field to create a permanent magnet in metal. Well, not too tough, but not too many of us have the necessary equipment. I say to start with the magnet first for one big reason; you'll be super angry if you wind the perfect coil, then realize it's too small to fit over your magnet. There are also places such as Digikey which currently stock permanent magnets in various metals. Once you have a magnet, you can start looking for a metal basket/pole plate combination. In headphones, it's probably easier to choose a metal cup, and use the permanent magnet as the pole. They'll look something like:
http://www.alibaba.com/product-free/104454224/Speaker_component_Top_plate_U_Yoke.html

How do you plan on doing your impedance analysis?

This is a fun project, but a crap load of work.

I agree that the conventional cone-type driver will be very difficult to implement at home, but a planar driver might be possible.  Either electrostatic, or dynamic.  Maybe a very small version of something like a Magnepan speaker?  I'd imagine even that to be an extremely intricate endeavor.