Soekrates initial test update

Finally got a chance to test most of the electronics involved with the Soekrates build.

The DAM1021 was already updated to firmware version 0.99 so all I have to do now is wait for the Wire LT3042 PSUs and the WaveIO to arrive before I can get music coming from the DAM1021.

I got a nifty USB to RS232 adapter cable with no plug so I could simply connect them to J10 without having to fool around with the plug. I'll probably use something along the lines of this along with a small connector on the back panel of the Soekrates to allow for firmware updates without having to open up the Soekrates enclosure. I did have to install drivers but I can easily put a guide up for anyone who buys the Soekrates.

Soekrates Update 13

Had a chat with a nice chap and have decided on reverting to a single-ended configuration because of Soren's quote that balanced doesn't add much if at all to the sound:

diyAudio Member   Join Date: Jun 2009

Although I don't see the big need for the DAC resistor string to be balanced, balanced signals are most useful for box to box interconnect.... You will get 3 db better S/N ratio, not that it's needed, there should already be 126 db S/N ratio at the resistor string.... __________________ Søren

This will shave off roughly $500 off the total price of the DAC as well as let me make it a lot less tall than what it is currently, so that's exactly what I did. (9in x 7.5in x 3in or 228.4mm x 190.5mm x 75.4mm).

Small project update and wooden headphones side project

As the Soekrates boards are still stuck in America, I finally caved in and burned roughly $500 to import some DAM1021-2 boards from the Soekris EU distributor. While I wait for them, the transformers and the LT3042-based power supplies to arrive, I'll show you a little something special: another little glimpse of a side project I've been working on.

These are wooden headphones based around a very special headphone driver (which I shall not mention in order to avoid driving up prices). Down below, you can see the basic plans. Wooden cups reinforced with a metal ring around the outside with metal gimbals connected to each other with a single metal wire. Shouldn't weigh too much and it should sound good too.

Progress update

Apologies for not posting much at all.

The project is stalled for now, as I had both my proto boards sent to a friend in the US to save on shipping (as I planned to have him ship many things at once). I had everything planned out until I found out I couldn't pay for the shipping label with my PayPal then it started falling apart from there. In addition to that, he can't send them on with his own money due to very complicated reasons so I pretty much have run out of things to do without touching the boards themself.

That said, I'm thinking about the different engineering challenges and compromises that is involved with the use of one board verses the use of two. Use of a single board means that differential output has to pass through two opamps via a single-ended to balanced buffer and it'll have a noticeable (adverse imo) affect on sound quality due to the opamps. This can be somewhat remedied if I swap the opamps but it adds another element of failure for the manufacturing process and then we'll have a whole new debate over which opamps different people want. This can be somewhat fixed if I stick a socket there but it will add noise and then I might as well not change the opamps.

However, having two boards adds its own challenges too. I have to parallel each of the digital inputs and supply power supplies to each board (I'm not sure what effect paralleling the same supply will have as per the latest plan I have), connect the serial lines to each other and increase the height of the enclosure significantly to take the two boards. This is without taking into account any other mods I may or may not perform for sound quality (as detailed on the blog) as twice the number of boards means twice the failure rate.

The reason why I'm troubling over this is because I would much prefer to offer either SE or differential options only as offering both would mean having to increase the price across the board due to smaller production runs meaning more expensive per-unit costs. Overall, I still haven't decided on whether to use dual or single boards (especially because an extra board also increases the price by a large chunk).