8 Channels ADAT converter

[netnoggin]

One more (important) update. Skylar mentions on the drawing that phase may be reversed. I just ran some tests on this and can confirm he was correct to be concerned. What I found out is some inputs are out of phase with others if wired using the polarity indicated. At this point I don't have a way to know which ones are "correct" and which aren't, but they should at least be all the same.

Here's the way the drawing lays it out (pin number designations follow the standard layout for this class of connector):

J2
Pins 1,2 -    ch1 +,-
Pins 4,5 -    ch3 -,+
Pins 7,8 -    ch5 +,-
Pins 10,11 - ch7 -,+
J3
Pins 1,2 -    ch2 +,-
Pins 4,5 -    ch4 -,+
Pins 7,8 -    ch6 +,-
Pins 10,11 - ch8 -,+

What I found was channels 1, 2, 5, and 6 are indeed out of phase with the others. I reversed those connections and they now all test the same. This really makes sense if you look at the way it turns out, pin-wise.

I will at some point try to figure out how the phasing compares to other equipment, but I thought I would pass this info on right away.

 

[janczmok]

Hi,

received my Focusrite today; starting on the panel design(s) and outwiring.
Wht i am not fully clear of: What power supply do i need to have ?
(Input Side 220V, Output ??)

 

[netnoggin]

output needed = 3:
+16.5 volts
-16.5 volts
+5 volts

Actual current requirements are not known. I didn't have a practical way to measure at the time of my build. I can tell you I used two 15-volt, 1-amp transfomers wired in series (to get 30 volt CT) to feed my homebrew regulator board, and it was enough. I used 1-amp, 3-pin regulators for the + and - 16 volts, and a 3-amp, 3-pin for the 5 volt supply.

 

[Skylar]

One more (important) update. Skylar mentions on the drawing that phase may be reversed. I just ran some tests on this and can confirm he was correct to be concerned. What I found out is some inputs are out of phase with others if wired using the polarity indicated. At this point I don't have a way to know which ones are "correct" and which aren't, but they should at least be all the same.

Here's the way the drawing lays it out (pin number designations follow the standard layout for this class of connector):

J2
Pins 1,2 -    ch1 +,-
Pins 4,5 -    ch3 -,+
Pins 7,8 -    ch5 +,-
Pins 10,11 - ch7 -,+
J3
Pins 1,2 -    ch2 +,-
Pins 4,5 -    ch4 -,+
Pins 7,8 -    ch6 +,-
Pins 10,11 - ch8 -,+

What I found was channels 1, 2, 5, and 6 are indeed out of phase with the others. I reversed those connections and they now all test the same. This really makes sense if you look at the way it turns out, pin-wise.

I will at some point try to figure out how the phasing compares to other equipment, but I thought I would pass this info on right away.

I have verified that channel one is in phase, as shown in my diagram.

Which would mean that, according to your findings, the + and - should be reversed for channels 3, 4, 7, 8.


On my unit, the default settings are 24bit, 88.2kHz.
I read somewhere that this will be different depending upon which firmware revision you have.
Apparently, there were complaints about 24/88.2 being the default, so they rev'd the firmware at some point to retain your last used settings.
I don't know if all of that is true, but the default on mine is definitely 24/88.2

 

[netnoggin]

Thanks for the info, Skylar! So I had a 50-50 shot at the polarity changeover, and guessed wrong. Actually, channels 1&2 were more accessible on the ribbon connector, which is why I chose to flip 1,2,5, and 6. I had a funny feeling I was flipping the wrong ones, but laziness took over. I'll fix it this weekend.

On the LED situation, your firmware must be different from mine because it definitely doesn't start at 88.2 on the rate. IIRC, it doesn't retain last setting but I'll have to check.  The main indicator for me on the bit depth would be if yours steps left to right as you push the selector button. I'm scratching my head for a way to detect the bit depth in software, but can't think of anything at the moment.

 

[netnoggin]

Never mind on the LED situation - I figured it out via software. I forgot Wavelab actually has a bit meter. I recorded audio for each of the LED positions and checked the result.

My original plan was to reverse the order of the bit-depth LEDs from the way they are on the Octopre, because I could and because having both sets of lights incrementing to higher values left to right made more sense to me. But somewhere in the construction phase I forgot my plan. So the result is that my LED order matches the Octopre front panel exactly. Sort of. I have the bit rate row on the left and the bit-depth on the right.

 

[skipwave]

I was going to suggest that you could figure the bid depth settings by recording, even without a bit meter. If cut the recordings to the same length, the file sizes would scale with the bit depth.

"We all take different roads to see the same things." - Taxi Driver Wisdom

 

[netnoggin]

True, except my recording method was a little too ad-hoc for the files to be comparable.  ;D I simply grabbed a mic and uttered something like testing number, number 2, etc. But you're right - done properly that would have revealed it also.

 

[skipwave]

My first test on most new builds is me mumbling through an SM57.  ;D

"Hello world."

 

[blue_luke]

Well I'm back!
Long story of illness and new borns grandson and grand-daughter now!! 

But now to this subject with a question....
I realywant to design a 8 line in to adat out card.
To the point where I think I will try to get a licence for ADAT from wavefront. I will see what it involves...

But a first question, how do we measure jitter, and also very important, where do I measure it.

Do I need a 100K$ osciloscope or there is a trick somewhere?

I am thinking of a rather fancy card using TI  AD converters, as AM are hard to come by, and TI are also very good!, a good buffer for the inputs and a few other goodies....

Is there interest?
Will ya help?
I intend to offer a board or maybe a kit...
Are you interested?
be warned that there will be some SMT soldering to do... but I have issues with that later on..

What do you think all?
I can not pull that one all by myself.

Here is a bit of background on me:
- I am production engineer for a small business that do solar powered trafic signage
- I design most of the stuff for this company
- I have been an industrial electronic tech for the last 20 years (servo-motors PLC programming etc...
- I have been a musician for more than 45 years
- I feel veru old!  :-\

All of this to say that, :yes I know about mixed signal, high precision electronics ( DC to Ku-band design and a stint at metrology), low-noise design etc...
But I am a bit daff when it comes to audio DA and AD and clock etc...

I am saying this not to boast, believe me I don't need it ;D, but to show that somehow there is a chance that I will succeed at this one.

Let me gather my toughts and I will come back with some drawings and ideas...

Luc

 

[jdbakker]

You might consider making a digital front-end for rkn80's recent modular A/D and D/A boards; that would make a lot of people happy.

I'll get back to you on the jitter measurement.

JDB.

 

[shot]

Did anyone figure out pinout for Behringer's ADT1616 converter card?
It looks interesting since it's AD and DA and it has 16 channels!



Luka

 

[walter]

I did check out the ADT1616 in a DDX3216.
50 pin connector.
Pin 1 GND
Pin 2,3,4,6,8,10,12,14,16,18,20,22,24,26,28,30 output pins 3,5,7,9 of LVT244 3.85kHz
Pin 5,7,9,11,13,15,17,19,21,23,25,27 input pins 2,4,6,8 of LVT244
Pin 29,31,36 ?
Pin 32 11kHz squarewave Data input to pin11 of IC6 74HCT04D hex inverter
Pin 33,35,37,39,41 GND
Pin 34 3.125 kHz Data input to pin 1 of IC6
Pin 38 11.76 kHz sinewave on O3 pin 12 of IC11 LVT244 octal buffer/line driver
Pin 40 5.3 vDC to IC3 pin 6 RB0/INT  PIC16C554-04/P
Pin 42 5.3 vDC IC3 pin 3 RA4/T0CKI
Pin 44,45,47 3.3vDC
Pin 48,49,50 5.3vDC
The signal lines appear to be digital, looks like a double helix on my scope. There are two cardslots on the DDX and that signal was on the empty slot as well. The LVT244s are setup as four sets of two in parallel. Running audio signal through it, I couldn't see it with the scope. This card is connected to the DSP board, and the I/O board. So this is an ADAT digital interface, the analog conversion is on another card. The frequencies may not be accurate.

 

[Andy Peters]

The signal lines appear to be digital, looks like a double helix on my scope.

That's a triggering artifact, although if they look like sine waves instead of square waves, your 'scope doesn't have enough bandwidth.

LVT244s are setup as four sets of two in parallel.

That's what they are: dual 4-bit digital buffers, with two output enables (one for each nybble).

Running audio signal through it, I couldn't see it with the scope.

You ran analog audio through the '244s? That won't work!

-a

 

[pedroplanet]

I've just finished my focusrite octpre A/D option converter  ;D

It's working! thank you guys for all the info!

 

[pedroplanet]

I'm back and my converter died, just pink noise coming from it. so sad