Modular multi channel DIY AD/DA Box

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
@chefducuisine: I'm an AES member. I'll have a look at the AES67 doc. But as it is stated on the ravenna website: Ravenna is not AES67, it is compatible. So their might be minor differences. I once downloaded the ravenna white paper but it did not have enough info's to implement it. But it might have changed and I should have a look again.

@ joaquins:
Just kidding.  ;)

I'ld like to focus on the USB and ADAT connectivity replacement first, that is difficult enough...
 
I've some news:
I just ordered a prototype board for the USB connectivity. I hope that the PCB and components will arrive before Christmas so that I can do some soldering and programming during the Christmas holidays.  8)

R.
 
Just wanted to say I spent about 4 hours reading this entire thread, and it is amazing Rafael got as much done as he did, there were a lot of folks asking for the moon, at any rate I am currently building 2 sets (8-ADC/AES) which I will dump into my Lynx AES daughterboard (slaved to the Lynx TWO mainboard) which will give me an extra 8 channels in...

I will let folks know and do a test set up to compare with the Lynx converters as well as do some comparisons with an Apogee converter on a Mac...(I have a converter box that will let me do one channel of AES>Spdif/Optical which I can compare in the MacBook)

Besides that I have a Kurzweil that will allow me to compare it to yet another converter as well as the PCM4222 EVAl board which I can play around with the clocks I2S etc...(which has been collecting dust for many years but it does Spdif/Toslink and I2S as well as AES if I recall...

But first more coffee so I can hand solder the smd...

Wanted to say thanks to Rafael for this project, he has the patience of a god in my humble opinion and has done a great service in answering politely every question and request...

Like the majority of posters here he is quite decent.
 
Hello everybody,

I'm new on this forum and this post is my first one ... did'nt found a place to introduce myself ...  :-\

I read this long, long thread yesterday, and first, I must greet Raphael, it is an inclredible job !

I have lots of questions about all this stuff. I hope Raphael or anabody else can answer.

1) Developped modules :
If I understood well everything I read, there already are :
- DAC module, which take 2 I2S stream line and create 4 (differential) analog out
- ADC module, which take 4 differential analog in and create 2 I2S stream
- AES module, which is a converter from/to AES to/from I2S. How many channel ?
(I deal with I2S channel hier and 1 i2S is 2 analog channel)
Is AES S/PDIF or something else ?
- ADAT module which is a converter from/to ADAT to/from I2S.
how many channels ?
- WordClock module which can select the master clock and give it to each module and the wordClock out
- One motherboard which is just a physical support for data and clock bus, and which generate power supplies
On this par, am I right ?

2) Under developpement modules
I read a lot about lots of modules ... usb, ethernet ... I'm sure Raphael is working on USB (and there are others questions I'm going to ask in the good thread)
... someone is working on the ethernet or is it dead?

3) Connectivity
Is there anywhere a schematic of all possible configurations ?
For example is it possible to have 1 wordClock module (I think it is an obligation), 4 ADC modules, 2 DAC modules and 1 USB module (whenit will be completed), in order to do a big "sound card" ?

4) About digital audio
I read that the clock is one of the most important thing ... can someone explaine in simple words why ?
As clock is a fast signal ... we can't do "Y" duplication with a clock cable, it needs a buffer I imagine ?
can we do a Y duplication with data signals ... for example ... 1 adc channel out goes to usb for recording and a DAC module to be monitored (it is just an example !)

Again, thousand greetings to you guys for this incredible work !

M.
 
8 x channels of ADC and DAC available for sale in the Black Market, just in case anyone's interested...

http://groupdiy.com/index.php?topic=60569.0

 
Some exciting news:
The prototype of a new mainboard is on the way to me. PCB is already shipped, components will be shipped in two days.

R.
 
rkn80 said:
The schematic of the clocking board is done.
Please note that this is a version with the intionention of having it running after soldering meaning that you don't have to program a device.
We discussed DDS and Microcontroller based solutions prior maybe such a solution will come in the future. It would be nice but I've currently no concrete plans to go there.

Raphael

Hi Raphael,

In the schematic, you have the CDCE913 Y3 clock output as the input for 4046 phase comparator.

Because no micro-controller unit or programming is involved, hence Y3 divider value = Y2 divider value = Y1 divider value = 1, the CDCE913 factory default setting.  Which also means that Y3 = Y2 = Y1 = VXCO.

Therefore, your PLL setup is a 4046 phase comparing external word clock (48KHz, 96KHz, or 192KHz) with VXCO.

In TI Application Report "Generating Low Phase-Noise Clocks for Audio Data
Converters from Low Frequency Word Clock" (SCAA088.PDF), the PLL setup is a 4046 phase comparing external word clock with the CDCE913 Y1 clock output divided internally by either 512 for 48KHz, 256 for 96KHz , or 128 for 192KHz. And to set this internal divider value on the fly is by programming via the I2C bus pins (S0, S1, S2)

So... I'm a bit confused. Which one is correct or better?  Perhaps you can explain to this newbie :)
 
Hello,

I would like to understand why, in this post

http://groupdiy.com/index.php?topic=35486.msg583101#msg583101

rkn80 said "I think it is not the badest idea to have an ADAT interface.
I'm not sure anymore wether Mikkel's boards will work because I recognized that the clocking of the wavefront devices is inverted compared to the rest of the AD/DA chip world. "

I'm looking at the data sheets for the Wavefront 1402
http://www.wavefrontsemi.com/DataSheetsFolder/WavefrontAL1402.pdf

and the TI PCM4104
http://www.ti.com/lit/ds/symlink/pcm4104-ep.pdf

and from what i can tell with my very limited comprehension, they both support a wordclock that is high first and left-justified 24 bit data output with most significant digit first. 

Is the problem that the bitclock signal for the 1402 is high first and for the PCM4104 it's low first? Am i reading that right?  Could that be solved with an inverting circuit between the output of the 1402 and the input of the 4104?

Thanks for any help.



 
Elegost said:
Is the problem that the bitclock signal for the 1402 is high first and for the PCM4104 it's low first? Am i reading that right?

In left justified mode, AL1402 outputs a rising edge bit clock and the PCM4104 expects a falling edge bit clock.

Could that be solved with an inverting circuit between the output of the 1402 and the input of the 4104?

I'd probably use a 74LVC1G04 (single inverter logic) in between them, running from the same digital supply as the PCM4104.
 
Hi everyone,

First of all, a big congratulation to Raphael for his incredible work !

I know it's a diy site, but maybe this can be interesting : https://www.audinate.com/products/manufacturer-products/dante-brooklyn-ii

It's a interface made by Audinate (creator of Dante) that allows any DAC / ADC to be connected to a dante network through TDM or I2S. Am i wrong or can it be used to connect the converter board you designed to a Dante network !?

Good luck for the rest,
Cheers
 
Keox35 said:
Hi everyone,

First of all, a big congratulation to Raphael for his incredible work !

I know it's a diy site, but maybe this can be interesting : https://www.audinate.com/products/manufacturer-products/dante-brooklyn-ii

It's a interface made by Audinate (creator of Dante) that allows any DAC / ADC to be connected to a dante network through TDM or I2S. Am i wrong or can it be used to connect the converter board you designed to a Dante network !?

Good luck for the rest,
Cheers

Having just read every post in this thread, (almost) every link, researching the technology discussed, etc. it looks like Audinate requires many hoops to be jumped through to access their technology. NDAs and the like.

It seems for now that the best DIY computer interface solution is USB. If you can wade through XMOS's documentation for code then this seems to be the best affordable product for large channel counts that doesn't require drivers or complex programming: http://www.diyinhk.com/shop/audio-kits/101-xmos-multichannel-high-quality-usb-tofrom-i2sdsd-spdif-pcb.html#/xmos_firmware-192k_8ch_out_8ch_in_spdif_out

It supports up to 32 channels with custom firmware for TDM. One possible solution mentioned in this thread is to use two separate interfaces: one for input and one for output. That is a possibility if you still need to do this DIY and need more channels than the chip supports, and most computers come with 2 USB ports minimum.

In the future I expect that AES67 will be the way to go. Dante now supports it and Ravenna has for a while. There is a product here that does 32x32 via AES67 but it isn't cheap: https://www.archwave.net/new-products-1-1/
The other communication I am interested in is Thunderbolt, but the interface chips are all BGAs, at least for now.

Not sure if any of this has been helpful, but these are my impressions of the state of an almost decade old project as of 2017.

 
Hi Raphael and all you guys

I found myself reading through this thread. Its certainly been a long journey. Am I right in saying that since its been so long now since your lastpost, this project is now buried or metaphorically archived into the museum of Group Diy?  It was such an amazing project and I always felt so ignorant with regard to anything digital related that I had nothing to contribute.  I still watched it eagerly to see its progress and all your pcb's and parts sat neatly on my shelf. I had hoped to make some nice ADC project to add extra inputs to my existing apogee interface. 

Is this project still being developed or has it reached its conclusion?

Cheers
 
Started looking how to connect the ADC card I built some time ago to AL1401 ADAT converter. It seems impossible because AL1401 uses internal fixed 64fs bit clock whereas the PCM4204 fixed 128fs bit clock at 48khz sample rate (single data rate only supported). Some external logic is needed, too bad also xCORE i2s doesn't seem to support that 128fs bit clock.
 
Where did you read about the fixed BCLK?
According to the datasheet it should support 64fs since the device supports i2s.
The master clock SCLKI is fixed to 128fs @ 48kHz.
 

Latest posts

Back
Top