I recently got some recievers, DAC chips and filters with the intention of builing my own external DAC to play around with. My first problem is understanding the data sheets. I don't understand how you know which sampling frequency to choose and how everything operates.
The datasheet talks about how to select a clock frequency based on fs, how do I know? If I use a reciever capable of recieving up to 24/96, does it automatically pick the input speed and everything else falls into place if set up correctly?
Is there a primer somewhere that talks about what all of the things the datasheets talk about mean? They all seem to assume a certain basic knowledge that I don't appear to have.
Thanks for any help!
The datasheet talks about how to select a clock frequency based on fs, how do I know? If I use a reciever capable of recieving up to 24/96, does it automatically pick the input speed and everything else falls into place if set up correctly?
Is there a primer somewhere that talks about what all of the things the datasheets talk about mean? They all seem to assume a certain basic knowledge that I don't appear to have.
Thanks for any help!
That basic digital audio knowledge can be learnt by doing an electronics course or by doing it the hard way; trying and doing wrong followed by repairing and testing 
. The latter method costs more time and money and is pretty inefficient but more focussed on the area you want to explore.
Both methods include a lot of reading and require a basic but thorough understanding of how electronics work. Most manufacturers of digital audio chips have other documents besides their datasheets like application notes etc. that can be a source of valuable information.This site can be of help too. If you wander through the Digital section you'll find enough links on the matter. You'll get to know the Search function as a bonus
My advise is to save the documents/datasheets ( some manufacturers only offer their datasheets as long as the part is in production !!! ) you want in folders on your harddrive for future reference. I generally convert them to pdf for readability. After one harddisk crash you'll appreciate the backup you made to CDR as I found out the hard way ...
. The latter method costs more time and money and is pretty inefficient but more focussed on the area you want to explore. Both methods include a lot of reading and require a basic but thorough understanding of how electronics work. Most manufacturers of digital audio chips have other documents besides their datasheets like application notes etc. that can be a source of valuable information.This site can be of help too. If you wander through the Digital section you'll find enough links on the matter. You'll get to know the Search function as a bonus
My advise is to save the documents/datasheets ( some manufacturers only offer their datasheets as long as the part is in production !!! ) you want in folders on your harddrive for future reference. I generally convert them to pdf for readability. After one harddisk crash you'll appreciate the backup you made to CDR as I found out the hard way ...
Schaef,
I've been thinking in a similar manner recently, as I decided to attempt my own MP3 player.
Of coarse, to the everydayer, that's sounds impossible. But, if you look around on the net, there are already a lot of sites about building your own MP3 player. For instance -
YAMPP - Yet Another MP3 Player
I recently downloaded an article from Circuit Celler by Jan Syzmanski titled 'Build Your Own MP3 Player'.
Although it involves perhaps a more than average level of digital knowledge for the programming side and so forth, it does include some useful descriptions about memory usage that even I can follow.
It's definitly worth the $1.50 to download it.
A method that I have found highly beneficial is the simple rereading one.
Read and read until your mind start to ache, then do something else for a few days. When you come back to look at it the second or third time, it makes a lot more sense, even if you had absolutely no idea what any of it meant the first time. It's almost as though your mind is subconciously building the dictionary of words, terms and relationships that you'll need to know, without evening knowing their true meanings.
I've been thinking in a similar manner recently, as I decided to attempt my own MP3 player.
Of coarse, to the everydayer, that's sounds impossible. But, if you look around on the net, there are already a lot of sites about building your own MP3 player. For instance -
YAMPP - Yet Another MP3 Player
I recently downloaded an article from Circuit Celler by Jan Syzmanski titled 'Build Your Own MP3 Player'.
Although it involves perhaps a more than average level of digital knowledge for the programming side and so forth, it does include some useful descriptions about memory usage that even I can follow.
It's definitly worth the $1.50 to download it.
A method that I have found highly beneficial is the simple rereading one.
Read and read until your mind start to ache, then do something else for a few days. When you come back to look at it the second or third time, it makes a lot more sense, even if you had absolutely no idea what any of it meant the first time. It's almost as though your mind is subconciously building the dictionary of words, terms and relationships that you'll need to know, without evening knowing their true meanings.
Thanks for replies!
Thank you all for your replies! I guess I should clarify a couple of things, first I've been reading the digital forum for quite some time, so I thought I had an idea of what goes on. Digital sampling makes sense to me and I understand from a computer perspective. (I'm a programmer in the real world) My problem is that the basics behind these are a little unclear and maybe I'm worrying about something I don't have to. I've been looking at schematics of the evaluation boards and other DIY DACs I could find and trying to understand them.
So, let's take a quick step back, and start with a couple of basic questions then.
1) given a reciever chip (say DIR1701 from TI) that's good up to 96kHz. Do I have to tell the chip what data rate to expect coming in or does it auto detect the rate?
2) Once the data has gotten from the reciever inside, what do the different fs settings do in the DAC? Is there a way to select one MCLK and accept 44.1, 48 and 96? (I know one MCLK can be used for 48 and 96, but what about 44.1?)
So, I think my biggest concern is with choosing a MCLK rate that allows the greatest flexibility and understanding how to choose it.
Thanks for any additional help!
Thank you all for your replies! I guess I should clarify a couple of things, first I've been reading the digital forum for quite some time, so I thought I had an idea of what goes on. Digital sampling makes sense to me and I understand from a computer perspective. (I'm a programmer in the real world) My problem is that the basics behind these are a little unclear and maybe I'm worrying about something I don't have to. I've been looking at schematics of the evaluation boards and other DIY DACs I could find and trying to understand them.
So, let's take a quick step back, and start with a couple of basic questions then.
1) given a reciever chip (say DIR1701 from TI) that's good up to 96kHz. Do I have to tell the chip what data rate to expect coming in or does it auto detect the rate?
2) Once the data has gotten from the reciever inside, what do the different fs settings do in the DAC? Is there a way to select one MCLK and accept 44.1, 48 and 96? (I know one MCLK can be used for 48 and 96, but what about 44.1?)
So, I think my biggest concern is with choosing a MCLK rate that allows the greatest flexibility and understanding how to choose it.
Thanks for any additional help!
It's also useful to have a look at the data on the various DAC, digital receiver, and ASRC chips from people like Analog Devices and Crystal (Cirrus). The evaluation boards usually are complete DAC modules and the data sheets have full schematics and usually a lot of other helpful information.
I've just bought the AD1896 ASRC evaluation board from Analog Devices. CS8414 rx, AD1896 ASRC and AD1852 DACs (As well as co-ax and opto SPDIF receivers with a co-ax SPDIF driver on the board). Basically a high-end digital rx/ASRC/DAC/digital tx board at less than 100 UK pounds.
Data on the AD web site under part number EVAL-AD1896EB
I've just bought the AD1896 ASRC evaluation board from Analog Devices. CS8414 rx, AD1896 ASRC and AD1852 DACs (As well as co-ax and opto SPDIF receivers with a co-ax SPDIF driver on the board). Basically a high-end digital rx/ASRC/DAC/digital tx board at less than 100 UK pounds.
Data on the AD web site under part number EVAL-AD1896EB
The 1852 is the DAC that AD have put on the evaluation board. Looking at the specification for that DAC, it looks as if it will be quite good enough for my ears!
I only got the evaluation board yesterday and haven't had a chance to try it out yet. (I need to build or buy a really nice case for it and build a suitable +/- 12V PSU to power the board.)
I only got the evaluation board yesterday and haven't had a chance to try it out yet. (I need to build or buy a really nice case for it and build a suitable +/- 12V PSU to power the board.)
I'm probably going totally off track now, but my excuse is that I'm as new to this and only trying to help...
Doesn't the DAC take it's processing rate from the clock channel? I mean, for instance, in the MP3 player schematics I'm looking at, the clock is a global part.
It feeds to the microprocessor and the DAC.
When the DAC's DREG, data request, channel signals so, the microprocessor starts reading the data out of the memory and sending it out, syncronised with the positive or negative edge of the clock cycle.
The DAC then uses the same clock channel to sense when it should be receiving new packets for unpacking and decoding.
In the DAC arrangement I was considering, the clock is an external 12MHz crystal. As I say, the clock's output then went to both the microprocessor and the DAC.
The PDF of the DAC made sure to say that I should avoid using third harmonic clocks, as these would cause timing failures in the system.
If you're not going to use an independant memory and processing system, and just clip the DAC onto your desktop, surely you will only need to connect the clock of your DAC to the clock of your desktop to syncronise the data feed correctly? This is where I check out, since I have no idea which pins the clock is sent out on, a search on google should bring up some piccies though.
I don't think that this particular DAC could automatically sync it's self.
As I say, I'm not professing to even be close to some of you guys when it comes to digital, so I'm ready to be wrong. I just thought this might help.
Doesn't the DAC take it's processing rate from the clock channel? I mean, for instance, in the MP3 player schematics I'm looking at, the clock is a global part.
It feeds to the microprocessor and the DAC.
When the DAC's DREG, data request, channel signals so, the microprocessor starts reading the data out of the memory and sending it out, syncronised with the positive or negative edge of the clock cycle.
The DAC then uses the same clock channel to sense when it should be receiving new packets for unpacking and decoding.
In the DAC arrangement I was considering, the clock is an external 12MHz crystal. As I say, the clock's output then went to both the microprocessor and the DAC.
The PDF of the DAC made sure to say that I should avoid using third harmonic clocks, as these would cause timing failures in the system.
If you're not going to use an independant memory and processing system, and just clip the DAC onto your desktop, surely you will only need to connect the clock of your DAC to the clock of your desktop to syncronise the data feed correctly? This is where I check out, since I have no idea which pins the clock is sent out on, a search on google should bring up some piccies though.
I don't think that this particular DAC could automatically sync it's self.
As I say, I'm not professing to even be close to some of you guys when it comes to digital, so I'm ready to be wrong. I just thought this might help.
dip16dac said:
Okay, since you mention using the PCM1738, we'll use that to try to clarify the clocking question I have. On page 11 of the PDF, figure 1 lists the SCK values for the different sampling frequencies. Its this table that's confusing me. If I read the table correctly, I need two different clocks, depending on whether the data coming in is either 44.1kHz, or a multiple of 48kHz.
Okay, I just looked at the DIR1701 again, is this clock generated by the reciever chip? (It looks like that's output from there) If so, then that seems to fix a lot of my problems/questions. Which means SCKO is not used in this case?
Which brings up a new question, to put a filter between the two, I'd use the SCKO from the reciever to the filter, and then the SCKO from the filter to the DAC, correct? With only one crystal to drive the reciever chip? (Meaning no other crystals for the other two chips?)
Sorry for the (probably) basic questions, but things are straightening out in my mind now, so thanks for all the help!!!
P.S. - dip16dac, do you have a schematic of your dac that I can look at? Or maybe even a board layout to help me understand all of this?
The PCM1738 has a system clock detection circuit that automatically senses the 256 fs rate that I set the DIR1701 to output, which is 256 times the S/PDIF rate that the sound card outputs, which is one of the rates that I can have winamp output from either 32kHz, 44.1kHz, 48kHz files or the output plugin ssrc upsampling rates of 88.2kHz and 96kHz. I did not use the SCKO of the PCM1738. A filter would use the DIR1701 as an external clock and then it would provide clock to the PCM1738 so only the one crystal is used.
There are wiring diagrams and layout of my sockets on my web page link to more photos. I can jumper the different settings on it. I also looked at the user manual slau069a.pdf for the PCM1738 eval board which uses the DIR1703. I had most of the same questions before I put one together but it worked out easier than I first thought.
There are wiring diagrams and layout of my sockets on my web page link to more photos. I can jumper the different settings on it. I also looked at the user manual slau069a.pdf for the PCM1738 eval board which uses the DIR1703. I had most of the same questions before I put one together but it worked out easier than I first thought.
What is it that you'd like to use the decoder for Schaef? Just music?
If so, you might be interested in theVS1001K by VLSI. It's a decoder but it has a lot of built integrated components, like a stereo headphone amplifier, bass extension response skew and a digital volume control. The idea is that you can create a literal drop in decoding system, attach the headphones and be ready to start.
You mentioned being a programmer, I have absolutely no idea how to write C code for decoders really. If you guys could suggest a good place to read specifically about this kind of work, I'd really appreciate it.
I'm fine on computers and have messed around with HTML and Java, I just never really took any interest in C as I've not had a use for it, as of yet.
YAMPP has firmware available to start and control the VS1001K based players. I guess it might be interesting for you to look over Schaef, as you'll probably understand it a lot better than I!
On a kind of related subject, do either of you know of any advance waveform editors that I could feed out into the DAC to form a kind of signal generator? I'm not looking for a perfect signal generator, but this exact setup as it so happens.
If so, you might be interested in theVS1001K by VLSI. It's a decoder but it has a lot of built integrated components, like a stereo headphone amplifier, bass extension response skew and a digital volume control. The idea is that you can create a literal drop in decoding system, attach the headphones and be ready to start.
You mentioned being a programmer, I have absolutely no idea how to write C code for decoders really. If you guys could suggest a good place to read specifically about this kind of work, I'd really appreciate it.
I'm fine on computers and have messed around with HTML and Java, I just never really took any interest in C as I've not had a use for it, as of yet.
YAMPP has firmware available to start and control the VS1001K based players. I guess it might be interesting for you to look over Schaef, as you'll probably understand it a lot better than I!
On a kind of related subject, do either of you know of any advance waveform editors that I could feed out into the DAC to form a kind of signal generator? I'm not looking for a perfect signal generator, but this exact setup as it so happens.
eeka chu said:
I'm looking at building an external DAC for a CD or DVD player to play back CDs or maybe 24/96 DVDs (not DVD-A). Basically, I'm just looking to play around, and want to understand this. I'm not building an MP3 player, and right now, have very little interest in building one. Maybe further down the road, but right now I just want to put something together that I can hook a player up to and hear better sound.
My thought is to make a modular unit with the PS on one board, the reciever on another, the DACs on a third, and if I want, a digital filter on a fourth. This'll allow a lot of flexibility for experimenting and seeing what sounds best. I've also conemtplated putting the analog stages on their own as well.
I only mentioned I'm a programmer, to say that I can usually understand these things, but it just wasn't connecting in my head. However I think I know what I want to do, and will try things eventually, I just wish I didn't have to create boards to experiment...
P.S. - I've tried looking at a couple of the evaluation boards, but the ones I've seen have a "magic" PGA between the chips doing something that I don't understand.
rfbrw said:
Never mind, I went back and looked more closely at some of the other boards. It seems I was looking at boards for chips that needed software programming. Looking at the hardware ones cleared a lot of things up for me. Thanks for all the help now! I think I understand things much better now!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.