The mainboard for the AD/DA-Box

[mattamatta]

If learning/buying a hot air station was all that stood between me and DIY (quality) DAC - sign me up!

Hot air rework stations are neat tools, but don't underestimate what can be done with a hand iron and magnifier.  A good microscope is worth its weight in gold, but even lesser magnifiers are greatly useful.  You don't actually need to be able to individually target each pin on something like a QFP, as you can sweep solder across all the pins and then use solder wick to remove excess.  I just did an 0402 part (onto 0603 pads though) today just to see if I could because I've never gone down past 0603 before.  Though I am using a nice boom arm stereo microscope which makes it almost feel like cheating.

Something like the QFN mentioned about is where a hot air rework nozzle shines, though a lot of the hacker/DIY guys doing SMT have been doing parts like that in even unregulated toaster ovens or on hotplates.

 

[grantlack]

What do you mean by "lack of remote deployment possibilities?"

a verbose drunk is just as bad as an unintelligible one.  ;D
i guess that was me frowning at the more "archaic" distance and data limitations of usb2 vs an ethernet solution.  i faintly recall that there were a few members exploring dante, (erikb?) and the brooklyn II card would only require a miniPCI connector.  it may not be viable based on cost, but from what little i understand from pre-NDA disclosures it's at least worth further inquiry.

 

[Andy Peters]

What do you mean by "lack of remote deployment possibilities?"

a verbose drunk is just as bad as an unintelligible one.  ;D

Uh, oh, are we going to have to install a breathalyzer on your computer, which will disallow posting if it detects alcohol on your breath?

i guess that was me frowning at the more "archaic" distance and data limitations of usb2 vs an ethernet solution.

The USB distance limit isn't archaic. (USB is newer than Ethernet!). Just remember that USB was invented to solve a different problem from Ethernet. (FireWire has a similar cable-length limitation.)

 

[grantlack]

The USB distance limit isn't archaic. (USB is newer than Ethernet!). Just remember that USB was invented to solve a different problem from Ethernet. (FireWire has a similar cable-length limitation.)

too true, i suppose i'm looking at things through the "audio interface" lens.  in that regard, ethernet seems to have the forward momentum (and upsides) whereas usb and firewire are becoming old hat.

 

[mhelin]

Regarding XMOS I think you can easily interconnect multiple chips using what they call xCONNECT these days. So you don't have to use the BGA chips for what you want, just put enough smaller (pin count and cores) XS1-L8-64 chips (with TQFP48, LQFP64 or TQFP128 packages available) there.

See https://www.xmos.com/en/published/xs1l_links

 

[rkn80]

There are some news regarding the main board:
a) I will give the xmos devices a try to make an USB Audio 2.0 interface.
Let's see if the TQFP can do the job. If not: soldering QFN is not really difficult as I learned now. It is in fact easier then the TQFPs. You only need a hot air gun. Already a cheap one for a few $ does the job.

b) The power supply is almost designed now. The features are:
Analog voltages: +15v -15v +5v -5v  +48v
Digital voltages: +5v +3.3v
The board is driven by a simple DC 18v to 36v wall wart power supply. Similar to commercial interfaces on the market.

R.

 

[michal_k]

yay for xmos chip!

 

[Esoterimix]

I love what you guys are doing with this... I just discovered the DAC thread last night... unfortunately, my experience is limited to Analog electronics, so I don't have much to add... sounds like Raph is doing a good job of keeping this as simple and straight forward as possible, as it should be... I like the d-sub connector idea... I've been looking for a ad/da that I can have at least 16 in and at least 16 out and be able to use my own pre's... I look forward to seeing this progress... what sample rates are going to be explored?

 

[rkn80]

Hi,

some news on the interface part:
I've now designed the hardware for it. The next step is to explore the software for the amos chips and see what is possible (channel count, sample rates etc.) and wether hardware changes are needed. This I want to do during the christmas holiday.

R.

 

[rkn80]

Latest news on the mainboard: Last week I got the components for the switching power supply and this weekend I built it. After some debugging +48V and +/-15V are working. +/-5V and +5V (digital) will be done soon. There was a wrong connected pin on the transformer. That has to be changed now on all other voltages, too. No idea how this could happen…
I'm also going to use the power supply design for my microphone amplifier. Therefore, it is a kind of universal power supply for audio use.

R.

 

[rkn80]

PSU done. See it here:
https://groupdiy.com/index.php?topic=54610.msg704434#msg704434

 

[fenchelteefee]

What about this guy here designing a 24 channel USB to i2s interface, including the software? You just have to add the Dac boards. No aes, no adat.

https://sites.google.com/site/koonaudioprojects/usb-to-multi-channel-usb2-0
http://www.diyaudio.com/forums/digital-line-level/200448-24-channels-usb-i2s-interface-source-codes-asio-vhdl-schematic.html

 

[Andy Peters]

What about this guy here designing a 24 channel USB to i2s interface, including the software? You just have to add the Dac boards. No aes, no adat.

https://sites.google.com/site/koonaudioprojects/usb-to-multi-channel-usb2-0
http://www.diyaudio.com/forums/digital-line-level/200448-24-channels-usb-i2s-interface-source-codes-asio-vhdl-schematic.html

Unless I've missed something, that guy's design isn't compliant with the USB Audio Class spec, which means a driver is needed, etc, etc, and that makes it a non-starter for use with the usual DAW applications.

It also looks way too complicated and I think he's taking the wrong approach. (He has one hammer in his toolbox.)

The easy way to DIY a multichannel USB audio interface would be to start with an XMOS part. Alternatively, one of the ARMs which support High Speed USB (NXP LPC18xx series, perhaps) glued to an FPGA over the external memory interface works.

-a