link

Who is interested in such a thing?

I'm interested in the MasterLink USB adapter" (MUA). I would like to hear more technical details about the hardware and (of course) the price. I've been planning a major redesign of my audio system/network. I'm not sure exactly what it will be, but it will include a mixture of vintage B&O equipment, as well as equipment from other manufacturers (both new, old, and older). About the only thing that I have figured-out is that nobody seems to offer a suitable solution. As such, I expect to have to develop something custom. Are you still thinking about offering anything based on the "universal link adapter"? I think that that platform has a lot of potential. However, I'm concerned about the "care and feeding" of it from a software build standpoint. I like the idea that the MUA could be used from a Raspberry Pi (or other Linux based PC?). Have you considered including a DataLink interface to the MUA? Glitch  

link

Are you still thinking about offering anything based on the “universal link adapter”?

The problem of a fully integrated solution like the universal link adapter certainly is the long term maintenance. I suspect the demand to be rather low and at certain point probably would not even cover for the ongoing product maintenance. So the product likely would have to be discontinued early leaving customers with outdated devices and me with a bad reputation. Don't want to go that way. If it's really just an easy to use hardware-only product with freely available software examples people would have maximal flexibility for ever and don't have to rely on a vendor updating its software. Hence the USB idea. USB serial converters and USB sound cards will still work in a long time from now so should this device. Would be nice if a certain community grows around that piece of hardware sharing software configuration examples and advancing the project. Pricing wise certainly below € 200. Probably closer to 100 than to 200 but let's see.

link

or other Linux based PC?

Yes, actually should work on anything having FTDI drivers and pyserial support. I actually don't know which platform doesn't. Could even run it from a Mac or a Windows PC.

link

Have you considered including a DataLink interface to the MUA?

Yes, still thinking about it and how to implement it best in that case. Best case Datalink could be directly driven from a general purpose IO pin the USB-serial adapter chip exposes anyway. Worst case it would require a dedicated MCU. Still have to try that. Actually I would like to separate ML and DL into two independent solutions.

link

I would like to hear more technical details about the hardware

I'm thinking about a USB-C connector that supplies 5V power. Internally it is then connected to a USB hub chip which then connects to a USB audio codec and the serial converter for data communication. On the ML side it would also contain the power" circuit so that it could either operate as a master or a slave device. Programming is pretty easy. I can supply some python scripts for basic communication and protocol analysis. Audio would just work like a normal USB sound card.

Great idea!

I’m interested too.

Best regards,

Axel

Sounds interesting ?

Great, thanks for confirming.

Meanwhile I  created a block diagram of the solution in question.
More or less copied over the audio block form the universal link adapter project. Just added a simple USB I2S bridge. While there are several other ways implementing it on the audio side -  those two chips (ADC & DAC) are working well with ML and only require minimal surrounding components.

For ML master/slave selection I probably just add a manual switch on the board. I see no application where this needs to be changed programatically during runtime...

Login to view image

link

Glitch wrote: Are you still thinking about offering anything based on the “universal link adapter”?

(snip) Don’t want to go that way.

Fair enough. It seems like the numbers were against the ULA concept. The subset of people that have the need for such a device, would want to pay a fair price for it, and have the skills and motivation to write the supporting code would be pretty small. However, a similar argument can be made for something like the Beolab 90, yet it exists ;-). Glitch