Hello to everyone!
I'm trying to design a ADAT-to-analog-converter, but I ran into several problems. I hope there's someone out there who can help me with this. First a short description of my circuit:
I use a AL1402 optoRec by Alesis(Wavefront) semiconductor, connected to a TORX178B opto receiver (47uH and the 100nF capacitor are connected, too). The AL1402 is connected to a Burr-Brown PCM1606 DAC with following signals:
AL1402 PCM1606
DVCO --> 50 ohm resistor --> SCKI
BCLK --> 74LS04 ---> 50 ohm-> BCLK
WDCLK --> 50 ohm resistor --> LRCLK
The AL1402 works in mode 0 (master) and format 01 (left-justified, data valid on falling edge of BCLK?!?).
The PCM1606 works in format 11 (left-justified, data valid on rising edge of BCLK)
A 74LS04 inverter is connected in the BCLK signal to change the polarity of the BCLK signal. (rising/falling edge).
The De-emphasis of the PCM1606 is switched off. (BTW, what is it for at all?)
My problem is: i hear a distorted signal from the analog output. On the scope there are visible peaks in the analog signal, all pointing from signal to GND. Changing the SCKI-source from SVCO to DVCO improved the quality, but there is still a distortion in the signal. So I have several questions:
1. is a TORX178B-receiver possible for ADAT use? Or will I have to use the TORX173?
2. I don't get a stable WDCLK-signal (scope) on the output of the AL1402. Is that normal, or is here my problem?
3. Does anybody have a reference design with the AL1402 or the PCM1606? The datasheet of the AL1402 doesn't show enough interconnections to the DAC...
4. Is the LS04 fast enough for the BCLK-signal?
Any comments are appreciated!
Greetings
KMT
I'm trying to design a ADAT-to-analog-converter, but I ran into several problems. I hope there's someone out there who can help me with this. First a short description of my circuit:
I use a AL1402 optoRec by Alesis(Wavefront) semiconductor, connected to a TORX178B opto receiver (47uH and the 100nF capacitor are connected, too). The AL1402 is connected to a Burr-Brown PCM1606 DAC with following signals:
AL1402 PCM1606
DVCO --> 50 ohm resistor --> SCKI
BCLK --> 74LS04 ---> 50 ohm-> BCLK
WDCLK --> 50 ohm resistor --> LRCLK
The AL1402 works in mode 0 (master) and format 01 (left-justified, data valid on falling edge of BCLK?!?).
The PCM1606 works in format 11 (left-justified, data valid on rising edge of BCLK)
A 74LS04 inverter is connected in the BCLK signal to change the polarity of the BCLK signal. (rising/falling edge).
The De-emphasis of the PCM1606 is switched off. (BTW, what is it for at all?)
My problem is: i hear a distorted signal from the analog output. On the scope there are visible peaks in the analog signal, all pointing from signal to GND. Changing the SCKI-source from SVCO to DVCO improved the quality, but there is still a distortion in the signal. So I have several questions:
1. is a TORX178B-receiver possible for ADAT use? Or will I have to use the TORX173?
2. I don't get a stable WDCLK-signal (scope) on the output of the AL1402. Is that normal, or is here my problem?
3. Does anybody have a reference design with the AL1402 or the PCM1606? The datasheet of the AL1402 doesn't show enough interconnections to the DAC...
4. Is the LS04 fast enough for the BCLK-signal?
Any comments are appreciated!
Greetings
KMT
Hi, meta!
Glad You're interested in my ADAT-DAC circuit.
Many problems disappeared til now, and an evaluation circuit is working. As opamps, I use the LM833, but I guess a good
low-noise opamp OP... by Burr-Brown would do the same.
Attached You can see a schematic of the test PCB. Supply voltage for the AL1402 and CS4228 is +5VDC, for the LM833 is +/- 15VDC. The connection to the AT90S8515 is to initialize the CS4228. Unfortunately, the use of a microcontroller seems to be the only way to set the interface format, mute control etc. of the CS4228.
A thing I'm going to change in the next versions is that the analog and digital power supplies are (yet) tied together. To improve SNR etc. I'm going to separate them.
The CLK generation doesn't seem to be the problem when receiving ADAT, because the AL1402 generates all required CLK signals for driving the CS4228. The only problem was the polarity of the BitCLK signal, which I had to change to drive the CS4228.
BTW, I've got a question, too: whats the intended use for the De-emphasis in the DACs? Should I turn it on always or is there a "hidden status bit" in the ADAT data stream which tells me to turn it on or off?!? I noticed that the sound is a little less clear with the De-emphasis turned on. An effect I'd like to avoid.
Greetings from Germany!
KMT
Glad You're interested in my ADAT-DAC circuit.
Many problems disappeared til now, and an evaluation circuit is working. As opamps, I use the LM833, but I guess a good
low-noise opamp OP... by Burr-Brown would do the same.
Attached You can see a schematic of the test PCB. Supply voltage for the AL1402 and CS4228 is +5VDC, for the LM833 is +/- 15VDC. The connection to the AT90S8515 is to initialize the CS4228. Unfortunately, the use of a microcontroller seems to be the only way to set the interface format, mute control etc. of the CS4228.
A thing I'm going to change in the next versions is that the analog and digital power supplies are (yet) tied together. To improve SNR etc. I'm going to separate them.
The CLK generation doesn't seem to be the problem when receiving ADAT, because the AL1402 generates all required CLK signals for driving the CS4228. The only problem was the polarity of the BitCLK signal, which I had to change to drive the CS4228.
BTW, I've got a question, too: whats the intended use for the De-emphasis in the DACs? Should I turn it on always or is there a "hidden status bit" in the ADAT data stream which tells me to turn it on or off?!? I noticed that the sound is a little less clear with the De-emphasis turned on. An effect I'd like to avoid.
Greetings from Germany!
KMT
Attachments
Hi ktm,
How is your circuit working?
May I ask where did you buy the Alesis chip in Germany?
I am thinking of multi-channel DAC to use with the computer soundcards ADAT output. Have you seen by chance an implementation of this?
The DAC will be simple Non oversamling using the TDA 1543... will try to go as simple as possible.
regards,
Sebastian
How is your circuit working?
May I ask where did you buy the Alesis chip in Germany?
I am thinking of multi-channel DAC to use with the computer soundcards ADAT output. Have you seen by chance an implementation of this?
The DAC will be simple Non oversamling using the TDA 1543... will try to go as simple as possible.
regards,
Sebastian
Hi, Sebastian!
I changed the used DAC to the AK4356 by AKM, and it's working pretty fine. It sounds very good, very low noise and the AK4356 is much easier to handle than the CS4228. It has symmetrical outputs, and I connected it to the proposed filter circuit by AKM (only a few resistors, caps and a NE5532A). The sound is really pretty fine, "jenseits von Gut und Böse". The only problem is to get the AK4356. I had to order them in USA...
Greetz from Hamburg
Kai Markus
I changed the used DAC to the AK4356 by AKM, and it's working pretty fine. It sounds very good, very low noise and the AK4356 is much easier to handle than the CS4228. It has symmetrical outputs, and I connected it to the proposed filter circuit by AKM (only a few resistors, caps and a NE5532A). The sound is really pretty fine, "jenseits von Gut und Böse". The only problem is to get the AK4356. I had to order them in USA...
Greetz from Hamburg
Kai Markus
@kmt
Why do you have the resistors on the DVCO, WDLK etc. connections? Did you care about seperate ground planes for digital and analog ground? And did you do any measurements about SNR? Is there an improvement from TORX 173 to 178?
I for myself am developing the same application: Adat -> 8 analog using AL 1402 and at the moment Cirrus CS 4382. In the next iteration I will switch to AKM AK 4358.
Furthermore, do you have any source for the AL1402 for private use in germany?
regards,
Christian
Why do you have the resistors on the DVCO, WDLK etc. connections? Did you care about seperate ground planes for digital and analog ground? And did you do any measurements about SNR? Is there an improvement from TORX 173 to 178?
I for myself am developing the same application: Adat -> 8 analog using AL 1402 and at the moment Cirrus CS 4382. In the next iteration I will switch to AKM AK 4358.
Furthermore, do you have any source for the AL1402 for private use in germany?
regards,
Christian
Skwit,
I am gathering the documentation to do what you are doing. I also wish to interface a AL1402 to a CS4382.
Do you have any success stories or schematics of what you have done so far as to keep me from going down any dead end roads before I start soldering?
I would much appreciate it.
Ron
I am gathering the documentation to do what you are doing. I also wish to interface a AL1402 to a CS4382.
Do you have any success stories or schematics of what you have done so far as to keep me from going down any dead end roads before I start soldering?
I would much appreciate it.
Ron
Yes, I have success stories, the device works. The CS4358 was replaced by an AKM 4358, the AKM was cheaper and did not get as hot as the Cirrus. If you use an AL1402 you have to invert the BCLK-Line.
I tried several methods of grounding, I ended with two ground planes, the digital parts in the digital domain, the analog parts including the DAC in the analog domain. They can be connected by a jumper. I think I measured 109dB(A) SNR, with which I'm quite happy.
I can't give away schematics or layout, I made this device for a company, I think they won't be happy if I give it away.
Where did you get the AL1402?
Christian
Hi to everyone!
It has been a long time since I built up my surround distribution amplifier, but I try to remember the design rules I applied that time...
The schematics I posted are obsolete. As I said in my posts before, I replaced the DAC with a AK4356, a multichannel DAC with balanced outputs. It works pretty fine. Attached You can see the "new" schematics of the DAC. The only thing connected to the output stage of the DAC is an active LPF as described in the AK4356 data sheet.
As You see in the schematics, I completely separated the analogue and digital power rails in order to improve SNR. I must confess, I never really tested the overall SNR of the amplifier I only did a "rehearsal" at my kolloquium at the end of my diploma work. The amplifier was connected to my DIGI002R soundcard via ADAT and XLR, and at the other end, connected to a very expensive surround sound studio monitor system by "Musikelektronik Geithain". There was no "audible" noise caused by the amplifier
I got my AL1402 in UK, by www.profusionplc.com
As You said, it's sometimes very hard to get such devices in Germany. The wholesale dealers here aren't interested in selling 1-10 pieces of a device. I always purchase special parts in the UK or in the USA. It takes a little bit time, but I made the experience that dealers are much friendlier to DIY users... That's what I call "service"!
Greetz from Hamburg
KMT
It has been a long time since I built up my surround distribution amplifier, but I try to remember the design rules I applied that time...
The schematics I posted are obsolete. As I said in my posts before, I replaced the DAC with a AK4356, a multichannel DAC with balanced outputs. It works pretty fine. Attached You can see the "new" schematics of the DAC. The only thing connected to the output stage of the DAC is an active LPF as described in the AK4356 data sheet.
As You see in the schematics, I completely separated the analogue and digital power rails in order to improve SNR. I must confess, I never really tested the overall SNR of the amplifier
I only did a "rehearsal" at my kolloquium at the end of my diploma work. The amplifier was connected to my DIGI002R soundcard via ADAT and XLR, and at the other end, connected to a very expensive surround sound studio monitor system by "Musikelektronik Geithain". There was no "audible" noise caused by the amplifier I got my AL1402 in UK, by www.profusionplc.com
As You said, it's sometimes very hard to get such devices in Germany. The wholesale dealers here aren't interested in selling 1-10 pieces of a device. I always purchase special parts in the UK or in the USA. It takes a little bit time, but I made the experience that dealers are much friendlier to DIY users... That's what I call "service"!
Greetz from Hamburg
KMT
Wasn't able to attach the schematics. Either the size in pixels
is too big or the file is too big. Perhaps the board could be
improved. Why do the pictures have a restriction in pixels?!?
However, here's the link to the schematic:
http://de.geocities.com/tonschnoer/DAC.gif
Greetz
KMT
is too big or the file is too big. Perhaps the board could be
improved. Why do the pictures have a restriction in pixels?!?
However, here's the link to the schematic:
http://de.geocities.com/tonschnoer/DAC.gif
Greetz
KMT
I'm getting my adat recviever boards here.
http://www.prodigy-pro.com/forum/viewtopic.php?t=9018
:Thanks for the info gentlemen!
Ron
http://www.prodigy-pro.com/forum/viewtopic.php?t=9018
:Thanks for the info gentlemen!
Ron
I am interested in a diy ADAT input based on TORX and optorec AL1402 (a TTL chip) for using with PCM1794A dac (the PCM1794 is 5V input tolerant).
PCM1794 would be set in Left justified and Al1402 too (format 01 and 24bit).
The only difference I see (between PCM1794 and AL1402 graph in left justified) is that bitclock is in opposite phase (PCM1794 falling at beginning of LRCK whereas AL1402 rising).
Is it going to be a problem ? What should I do to prevent it from being a problem ?
What is the purpose of 74LS04 (or similar) device on the bitclock line between AL1402 and dac chip in the previous schematics and messages ?
Is a similar device needed in my case (it is not shown on optorec datasheet) ?
Any help for a newbie ?
Thank you VERY MUCH, Eric
PCM1794 dac data :
http://focus.ti.com/lit/ds/symlink/pcm1794a.pdf
Al1402 optorec data :
www.wavefrontsemi.com/DataSheetsFol...frontAL1402.pdf
PCM1794 would be set in Left justified and Al1402 too (format 01 and 24bit).
The only difference I see (between PCM1794 and AL1402 graph in left justified) is that bitclock is in opposite phase (PCM1794 falling at beginning of LRCK whereas AL1402 rising).
Is it going to be a problem ? What should I do to prevent it from being a problem ?
What is the purpose of 74LS04 (or similar) device on the bitclock line between AL1402 and dac chip in the previous schematics and messages ?
Is a similar device needed in my case (it is not shown on optorec datasheet) ?
Any help for a newbie ?
Thank you VERY MUCH, Eric
PCM1794 dac data :
http://focus.ti.com/lit/ds/symlink/pcm1794a.pdf
Al1402 optorec data :
www.wavefrontsemi.com/DataSheetsFol...frontAL1402.pdf
Hi, ejaouen!
As far as I remember (it's a while ago I did this project) the
74HC04 was needed exactly to change the phase of the BCLK signal
so it fits the CS4223 clock signal input requirements. In some
cases it could be required, depending on the type of DAC you are
using. I'm not quite shure if the DAC you're gonna use will require
a LS04, but when I'm reading your description, I would guess so.
Hope this helps!
Bien des choses de Hambourg
kmt
As far as I remember (it's a while ago I did this project) the
74HC04 was needed exactly to change the phase of the BCLK signal
so it fits the CS4223 clock signal input requirements. In some
cases it could be required, depending on the type of DAC you are
using. I'm not quite shure if the DAC you're gonna use will require
a LS04, but when I'm reading your description, I would guess so.
Hope this helps!
Bien des choses de Hambourg
kmt
Thank you VERY MUCH kmt.
It seems like you would be right : rightjustified mode for the PCM1794 and AL1402G graphs look the same on every detail to my newbie eyes.
What was the reason you did not use rightjustified (standart) mode with the pcm1606 :
Is it that the pcm1606 would limit itself to 16 bit only in this mode whereas it is capable of 24bit in left justified ?
Thank you again, Eric
It seems like you would be right : rightjustified mode for the PCM1794 and AL1402G graphs look the same on every detail to my newbie eyes.
What was the reason you did not use rightjustified (standart) mode with the pcm1606 :
Is it that the pcm1606 would limit itself to 16 bit only in this mode whereas it is capable of 24bit in left justified ?
Thank you again, Eric
You're welcome !
And You're right, I chose the left justified mode because
of the limited resolution when using the right justified mode.
The PCM1606 isn't as flexible as the DACs from AKM I used
concerning the input formats.
If You're still free using another DAC I'd like to recommend
to You the DACs of Asahi Kasei (AKM). Have a look at
www.akm.com
They're not easy to purchase (I had to order them in USA),
but they're easy to handle and they provide excellent audio
quality as well as balanced outputs. Digidesign uses such DACs
in their sound cards for ProTools. I used the AK4356 because
my circuits were designed to work in a 5.1 studio environment with
active speakers with balanced inputs connected to the DAC.
Hope this helps
Greetz from Hambourg
Kai Markus
!And You're right, I chose the left justified mode because
of the limited resolution when using the right justified mode.
The PCM1606 isn't as flexible as the DACs from AKM I used
concerning the input formats.
If You're still free using another DAC I'd like to recommend
to You the DACs of Asahi Kasei (AKM). Have a look at
www.akm.com
They're not easy to purchase (I had to order them in USA),
but they're easy to handle and they provide excellent audio
quality as well as balanced outputs. Digidesign uses such DACs
in their sound cards for ProTools. I used the AK4356 because
my circuits were designed to work in a 5.1 studio environment with
active speakers with balanced inputs connected to the DAC.
Hope this helps
Greetz from Hambourg
Kai Markus
I realize this is now an old thread, but if you happen to read this kmt:
This is exactly what I need: I'm looking for a way to get high quality 5.1 balanced line output from an ADAT signal. The boxes on the market right now are just too expensive for me. 3 questions:
1. This is somewhat of a newbie question: did you take the single-ended output of the AKM-recommended LPF and then used a typical inverting/noninverting pair of opamps to generate your analog balanced output? Or is there such a thing as a balanced LPF circuit maintaining the balanced lines all the way to the output?
2. I've also considered retro-fitting an existing ADAT box that uses the AL1402 with better converters. If I used buffers do you think I could simply tap the signals I need from the existing AL1402 and build an accessory board with the new converter(s)? It's likely I'd use 3 converters because some of the stereo AKM's are even better than the multi-channel ones.
3. Have you looked into the S/MUX ADAT specification? It allows you to use two ADAT channels to transmit data at 24-bit 96 kHz (data is multiplexed across channels). But from what Wavefront data docs I've found, it looks like it would require additional circuitry to reassemble the 48 kHz data to 96 kHz for the DACs. This more out of curiosity as 44.1/48 is enough for my needs.
Thanks for any tips and for posting this in the first place!
This is exactly what I need: I'm looking for a way to get high quality 5.1 balanced line output from an ADAT signal. The boxes on the market right now are just too expensive for me. 3 questions:
1. This is somewhat of a newbie question: did you take the single-ended output of the AKM-recommended LPF and then used a typical inverting/noninverting pair of opamps to generate your analog balanced output? Or is there such a thing as a balanced LPF circuit maintaining the balanced lines all the way to the output?
2. I've also considered retro-fitting an existing ADAT box that uses the AL1402 with better converters. If I used buffers do you think I could simply tap the signals I need from the existing AL1402 and build an accessory board with the new converter(s)? It's likely I'd use 3 converters because some of the stereo AKM's are even better than the multi-channel ones.
3. Have you looked into the S/MUX ADAT specification? It allows you to use two ADAT channels to transmit data at 24-bit 96 kHz (data is multiplexed across channels). But from what Wavefront data docs I've found, it looks like it would require additional circuitry to reassemble the 48 kHz data to 96 kHz for the DACs. This more out of curiosity as 44.1/48 is enough for my needs.
Thanks for any tips and for posting this in the first place!
Hi, undertone!
It's really a long time ago, but I try to remember
1. As you said, I used the single-ended output balanced LPF circuit
as it is described (as far as i can remember) in the AK4356 datasheet.
In order to simplify the circuit design, I connected this single-ended
output to a DRV134 to get a (electronically) balanced output signal.
2. IMHO this should work. Adding a buffer to the CLK signals
could be a good thing. I'm not quite shure which load the
outputs of the AL1402 can deal with.
3. I know of some ADC boards using this option (e.g. the
ADC board of the RME Octamic), but I'm not shure if the
AL1402 work like this. My issue 2004 was to build a surround
distribution amplifier in a studio environment, and 92 kHz hasn't
yet been very popular in 2003/04.
Hope I could help a little bit!
Greetz from Hamburg
Kai Markus
It's really a long time ago, but I try to remember
1. As you said, I used the single-ended output balanced LPF circuit
as it is described (as far as i can remember) in the AK4356 datasheet.
In order to simplify the circuit design, I connected this single-ended
output to a DRV134 to get a (electronically) balanced output signal.
2. IMHO this should work. Adding a buffer to the CLK signals
could be a good thing. I'm not quite shure which load the
outputs of the AL1402 can deal with.
3. I know of some ADC boards using this option (e.g. the
ADC board of the RME Octamic), but I'm not shure if the
AL1402 work like this. My issue 2004 was to build a surround
distribution amplifier in a studio environment, and 92 kHz hasn't
yet been very popular in 2003/04.
Hope I could help a little bit!
Greetz from Hamburg
Kai Markus
Thanks for the update, kmt!
BTW, there is a schematic for a balanced LPF circuit in the AKM4394 (and other) data sheet that preserves the balanced line all the way to the output. But tight component matching is of course required for best performance so using the balanced driver is in fact simpler, more economical, and probably just as good.
BTW, there is a schematic for a balanced LPF circuit in the AKM4394 (and other) data sheet that preserves the balanced line all the way to the output. But tight component matching is of course required for best performance so using the balanced driver is in fact simpler, more economical, and probably just as good.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.