B ware
I fear you underestimate the number of real slackers out there. I hear all this real cool sounding name dropping with part numbers, filter types, and acronyms; but still can't figure out what you are trying to do. How about a block diagram of the system, and specs for digital signal format desired and DAC type. You guys are discussing the suspension tuning and nobody has mentioned if we are talking about a Greyhound bus or a 2 seat sports car. You know I have had some experience with the DTI Pro and Digital Lens and people still can hear the addition of my digital interface devices used with these technological marvels. I have taken parts OUT of the digital interface in a Wadia DAC (an FPGA wet dream) and very noticably improved the sound. Jocko Homo will tell you other digital audio horror stories like mine. Tossing complexity at audio products is masochism of the worst kind. Has Zen Master Pass taught us nothing?
Start personal rant here
This is section is optional and only advised for someone who can read all the way through one of Dice's or Greys more lenthy post.
I got sucked into a project like this at the end of my last telecom gig. They were designing a test system for high speed cable harnesses and man it was going to modular and flexible as hell. It was to have Talk battery or AC powered switching supplies and was pure high tech sex. Initial design plans get under way FPGA and processor were choices made. Mr. Slick Project Engineer walks to my office and says that they are so busy that they are going to let me help since I am not busy. They are going to throw me a crumb so I won't feel left out, think I. "Fine, where is the product spec." I say... "Well we will write it as we go since are behind as it is". Okay "what is it exactly you want to test about the cable harness?" say I. Hand waving speech about signal integrity, insertion loss, crosstalk intermittent terminations, bidirectional buffers, mom, apple pie, patrotism, and the American way. Oh and the circuit has to be cheap and not take up much board space and they need it in less than a week. "What is the software going to look like for the test ports"?" Well we don't have a guy to write to software yet but we think it should do so and so". Well good natured dummy me goes and designs a very simple rectification integrator circuit that actually works waits two days for some cable assemblies to test with. Mr. Slick project engineer comes in my cube demanding to know if it works. At this point I pointed him to lab and said "it is all set up knock your self out". PCBs got designed. Thousands of bucks worth of parts got ordered and received. Protoboards built. Hundreds of man hours paid for. And the whole project flushed down the john when they disbanded our cost reduction/profit improvement group. Oh yes....our return for expenditure was about ten to one as opposed to ten percent for the other business groups. The rational? Cost reduction is the responsibility of each business group! According to the head of *******el USA. Oh yea.... I found out Mr. Slick with the same job grade and no degree (I have a BSEE) was making $17,000 bucks a year more than I was. This guy was picking my brain so often that I would go down the hall the other way if I saw him first.
End rant here
Anybody see the similarities here? Now, I know nobody is asking anything from me (including my opinion) and I have great respect for anyone that knows what FIR,IIR,BGA,CIA,FBI,BSA,DAR,BLT,PDQ and TLA (Three Letter Acronyms) are. The moral of the story is don't throw technology at a problem until you know what you are trying to fix! A Sunday school teacher asked a sweet little girl in class "what do we have to be forgiven for our sins?" She thought a few seconds and said "first you have to sin."
Now I know I am going to be tarred and feathered But I have a question....
I have heard of hardware, software, firmware, shareware, vaporware, and even almost/there (behind schedule project). Hell I even got a CDROM describing the set up and features for my $900 office chair at work, that I suppose was chairware.
I just want to know if maybe we are the unknowing victims of the fist appearance of DIYbuzzwordware? I will take my lumps and give a real apology if you guys prove me wrong by say... March. Until then I reserve the right to nod and giggle quietly.
Your not quite repented sinner,
Fred
I fear you underestimate the number of real slackers out there. I hear all this real cool sounding name dropping with part numbers, filter types, and acronyms; but still can't figure out what you are trying to do. How about a block diagram of the system, and specs for digital signal format desired and DAC type. You guys are discussing the suspension tuning and nobody has mentioned if we are talking about a Greyhound bus or a 2 seat sports car. You know I have had some experience with the DTI Pro and Digital Lens and people still can hear the addition of my digital interface devices used with these technological marvels. I have taken parts OUT of the digital interface in a Wadia DAC (an FPGA wet dream) and very noticably improved the sound. Jocko Homo will tell you other digital audio horror stories like mine. Tossing complexity at audio products is masochism of the worst kind. Has Zen Master Pass taught us nothing?
Start personal rant here
This is section is optional and only advised for someone who can read all the way through one of Dice's or Greys more lenthy post.I got sucked into a project like this at the end of my last telecom gig. They were designing a test system for high speed cable harnesses and man it was going to modular and flexible as hell. It was to have Talk battery or AC powered switching supplies and was pure high tech sex. Initial design plans get under way FPGA and processor were choices made. Mr. Slick Project Engineer walks to my office and says that they are so busy that they are going to let me help since I am not busy. They are going to throw me a crumb so I won't feel left out, think I. "Fine, where is the product spec." I say... "Well we will write it as we go since are behind as it is". Okay "what is it exactly you want to test about the cable harness?" say I. Hand waving speech about signal integrity, insertion loss, crosstalk intermittent terminations, bidirectional buffers, mom, apple pie, patrotism, and the American way. Oh and the circuit has to be cheap and not take up much board space and they need it in less than a week. "What is the software going to look like for the test ports"?" Well we don't have a guy to write to software yet but we think it should do so and so". Well good natured dummy me goes and designs a very simple rectification integrator circuit that actually works waits two days for some cable assemblies to test with. Mr. Slick project engineer comes in my cube demanding to know if it works. At this point I pointed him to lab and said "it is all set up knock your self out". PCBs got designed. Thousands of bucks worth of parts got ordered and received. Protoboards built. Hundreds of man hours paid for. And the whole project flushed down the john when they disbanded our cost reduction/profit improvement group. Oh yes....our return for expenditure was about ten to one as opposed to ten percent for the other business groups. The rational? Cost reduction is the responsibility of each business group! According to the head of *******el USA. Oh yea.... I found out Mr. Slick with the same job grade and no degree (I have a BSEE) was making $17,000 bucks a year more than I was. This guy was picking my brain so often that I would go down the hall the other way if I saw him first.
End rant here
Anybody see the similarities here? Now, I know nobody is asking anything from me (including my opinion) and I have great respect for anyone that knows what FIR,IIR,BGA,CIA,FBI,BSA,DAR,BLT,PDQ and TLA (Three Letter Acronyms) are. The moral of the story is don't throw technology at a problem until you know what you are trying to fix! A Sunday school teacher asked a sweet little girl in class "what do we have to be forgiven for our sins?" She thought a few seconds and said "first you have to sin."
Now I know I am going to be tarred and feathered But I have a question....
I have heard of hardware, software, firmware, shareware, vaporware, and even almost/there (behind schedule project). Hell I even got a CDROM describing the set up and features for my $900 office chair at work, that I suppose was chairware.
I just want to know if maybe we are the unknowing victims of the fist appearance of DIYbuzzwordware? I will take my lumps and give a real apology if you guys prove me wrong by say... March. Until then I reserve the right to nod and giggle quietly.
Your not quite repented sinner,
Fred
Re: B ware
Harry, sorry Fred, you are talking some sense now. I have also some difficulties to "haja klyket". many three letter words... still they have got some idea that built in filters aren't any good. I have no clue what's the problem. Yes, what are they trying to achieve really? Maybe it's wiser to improve the speakers.
Fred Dieckmann said:
Harry, sorry Fred, you are talking some sense now. I have also some difficulties to "haja klyket". many three letter words... still they have got some idea that built in filters aren't any good. I have no clue what's the problem. Yes, what are they trying to achieve really? Maybe it's wiser to improve the speakers.
I have no patience for Fred's infinite screeds, but my original question still holds. What does the original poster intend to achieve in terms of better sonics than manufactured components.
The only thing more tiresome than overlong posts is turning an audio project into a software project. If I wanted a computer, I'd call Dell.
The only thing more tiresome than overlong posts is turning an audio project into a software project. If I wanted a computer, I'd call Dell.
I have no patience for Fred's infinite screeds
Hmmmm..... and I usually am cited for the brevity of my post.
Forgive my literary offence please. I excuse you from the obligation to read my post. I will try to use more TLAs in the furture. Thanks for caring enough to complain though.
UYB,
Fred
P.S."If I wanted a computer, I'd call Dell." Which leaves us to wonder how you are reading this now........
If you're going to San Francisco
Be sure to wear some flowers in your hair
If you're going to San Francisco
You're gonna meet some gentle people there
Hmmmm..... and I usually am cited for the brevity of my post.
Forgive my literary offence please. I excuse you from the obligation to read my post. I will try to use more TLAs in the furture. Thanks for caring enough to complain though.
UYB,
Fred
P.S."If I wanted a computer, I'd call Dell." Which leaves us to wonder how you are reading this now........
If you're going to San Francisco
Be sure to wear some flowers in your hair
If you're going to San Francisco
You're gonna meet some gentle people there
brevity of posts...
Hmm. Something for me to consider...
Sorry to hijack your thread, alvaius. I know you meant this to be mainly a DAC discussion...
Anyway, is the effort of a DSP system worth it? I guess for me, it's all about the music in the end, and perhaps also about the technical challenge. A general purpose DSP offers the possibility to do things better than the stock digital audio ICs. Also, the usual DIY DAC is pretty boring to me now (except for the analogue circuitry, which always seems to pose new mental challenges). For me, DSP will do things like loudspeaker crossovers, channel mixing and upsampling, digital delays, multi-channel bass management, etc. Alvaius wants to use DSP to replace the oversampling filter chips, such that he can use his own oversampling algorithms. There are lots of possibilities, but none of this stuff can be done any other way.
Fred: I have actually invested a lot of time and money in this project already, so I do have a good incentive to finish it. I'm also not bent on bringing in others for help. I could easily complete this on my own, no problemo. As I mentioned, this is has been a very long term project, and it will be Feb. or March before I can get <i>started</i> on it once again!!! I don't expect any real harware to materialize for maybe 8mo, depending how hard I work at it, and how much other stuff takes precedence.
jwb: The FPGAs actually saved a lot of complexity, and will allow for some reconfiguring if things don't work as expected the first time. Before their addition, there was a lot of glue logic and level translators etc, and the board functionality was very limited in comparison. I will try and put together a block diagram soon.
I don't exactly want a programming project either, which is why I'm <i>not</i> using a computer to do all the processing. The unfortunate part is, if you want to do any audio processing digitally (at least anything which you have direct control over), you're going to have to write some code at some point. I figure programming the embedded stuff should generate considerably fewer headaches than a PC... and once the software is written, it will be pretty static. The resulting stand-alone box can sit on the shelf with my other stereo gear and number-crunch silently away, without crashing or otherwise being a nuisance.
Hmm. Something for me to consider...
Sorry to hijack your thread, alvaius. I know you meant this to be mainly a DAC discussion...
Anyway, is the effort of a DSP system worth it? I guess for me, it's all about the music in the end, and perhaps also about the technical challenge. A general purpose DSP offers the possibility to do things better than the stock digital audio ICs. Also, the usual DIY DAC is pretty boring to me now (except for the analogue circuitry, which always seems to pose new mental challenges). For me, DSP will do things like loudspeaker crossovers, channel mixing and upsampling, digital delays, multi-channel bass management, etc. Alvaius wants to use DSP to replace the oversampling filter chips, such that he can use his own oversampling algorithms. There are lots of possibilities, but none of this stuff can be done any other way.
Fred: I have actually invested a lot of time and money in this project already, so I do have a good incentive to finish it. I'm also not bent on bringing in others for help. I could easily complete this on my own, no problemo. As I mentioned, this is has been a very long term project, and it will be Feb. or March before I can get <i>started</i> on it once again!!! I don't expect any real harware to materialize for maybe 8mo, depending how hard I work at it, and how much other stuff takes precedence.
jwb: The FPGAs actually saved a lot of complexity, and will allow for some reconfiguring if things don't work as expected the first time. Before their addition, there was a lot of glue logic and level translators etc, and the board functionality was very limited in comparison. I will try and put together a block diagram soon.
I don't exactly want a programming project either, which is why I'm <i>not</i> using a computer to do all the processing. The unfortunate part is, if you want to do any audio processing digitally (at least anything which you have direct control over), you're going to have to write some code at some point. I figure programming the embedded stuff should generate considerably fewer headaches than a PC... and once the software is written, it will be pretty static. The resulting stand-alone box can sit on the shelf with my other stereo gear and number-crunch silently away, without crashing or otherwise being a nuisance.
I conceed then
So you are talking about a do it YOURSELF project. Go man go!
There are plenty of interesting topics on DAC design for us mere mortals. I could practicaly write a book on the AES and SPDIF interface which I have NEVER scene done right in a commecial DAC or transport . Asychonous reclocking, voltage regulation optimized for digital circuits, signal integrity, groundind schemes, PLL design, RFI suppression..... we are talking a long list of things to design and tweak.
Fred
So you are talking about a do it YOURSELF project. Go man go!
There are plenty of interesting topics on DAC design for us mere mortals. I could practicaly write a book on the AES and SPDIF interface which I have NEVER scene done right in a commecial DAC or transport . Asychonous reclocking, voltage regulation optimized for digital circuits, signal integrity, groundind schemes, PLL design, RFI suppression..... we are talking a long list of things to design and tweak.
Fred
Sorry, was on business travel, so have not been able to post.
To HifiZen, thank you for your offer of your development. I actually wan't to avoid the use of chips such as the AD1896. Essentially, it is an oversampling digital filter. Hence, it adds something into the equation I can not control. While I like the idea of processing at one single frequency a lot, I think I would prefer to assume the input is 44.1, 48, or some multiple of those two and work from there.
Yes, the idea of a 21161 in BGA is not a very nice thought. To that end, what I am thinking of doing is using an EASYKIT-LITE for the 21161 from ADI as the processing board, and then building everything else on a daughter card (or mother card) if you prefer. I am going to want to reclock the IIS anyway to remove jitter. I could consider a Motorola 563xx chip, as you can get free assembly tools, but I like the 32 bit\floating point of ADI as well as the better systems capabilities. I also like the MCF5249 from Motorola as the perfect digital audio system chip, within the limits of horsepower.
I am lucky in that I have access to free software.
To those who have once again mentioned off the shelf digital filters, they all appear to be linear phase. I don't want that.
To those who question what I want to accomplish. Well, I want to accomplish a specific digital processing chain that I don't feel I can achieve with off the shelf digital filters, sample rate convertors, or with the use of Sigma-Delta DACS. Heh, it could be a complete flop, but aren't most DIY projects the first time around. I am intrigued by the results of NON-OS DACS, overall, I do not think they sound better than a good OS DAC, but there are qualities of their playback on certain types of music that are intriguing. This system should be able to recreate any benefits of a NON-OS DAC, while giving all the benefits of an OS DAC.
To those who think this is beyond the DIY scope, not all DIYers are at the same level. For me, and at least some others, this is not a daunting project, beyond the obvious time investment. For those that are not at this level yet, I hope it will be a good chance to learn first hand from our mistakes, tribulations, and success. And in DIY fashion, if you want to build your own copy, you can.
What will result is a flexible platform that will offer the potential for a "purest" approach to digital audio playback. The nice thing about software is that it is easy to change and hence lots of potential for customization exists, and of course still lots of room to play with on the output end.
Alvaius
To HifiZen, thank you for your offer of your development. I actually wan't to avoid the use of chips such as the AD1896. Essentially, it is an oversampling digital filter. Hence, it adds something into the equation I can not control. While I like the idea of processing at one single frequency a lot, I think I would prefer to assume the input is 44.1, 48, or some multiple of those two and work from there.
Yes, the idea of a 21161 in BGA is not a very nice thought. To that end, what I am thinking of doing is using an EASYKIT-LITE for the 21161 from ADI as the processing board, and then building everything else on a daughter card (or mother card) if you prefer. I am going to want to reclock the IIS anyway to remove jitter. I could consider a Motorola 563xx chip, as you can get free assembly tools, but I like the 32 bit\floating point of ADI as well as the better systems capabilities. I also like the MCF5249 from Motorola as the perfect digital audio system chip, within the limits of horsepower.
I am lucky in that I have access to free software.
To those who have once again mentioned off the shelf digital filters, they all appear to be linear phase. I don't want that.
To those who question what I want to accomplish. Well, I want to accomplish a specific digital processing chain that I don't feel I can achieve with off the shelf digital filters, sample rate convertors, or with the use of Sigma-Delta DACS. Heh, it could be a complete flop, but aren't most DIY projects the first time around. I am intrigued by the results of NON-OS DACS, overall, I do not think they sound better than a good OS DAC, but there are qualities of their playback on certain types of music that are intriguing. This system should be able to recreate any benefits of a NON-OS DAC, while giving all the benefits of an OS DAC.
To those who think this is beyond the DIY scope, not all DIYers are at the same level. For me, and at least some others, this is not a daunting project, beyond the obvious time investment. For those that are not at this level yet, I hope it will be a good chance to learn first hand from our mistakes, tribulations, and success. And in DIY fashion, if you want to build your own copy, you can.
What will result is a flexible platform that will offer the potential for a "purest" approach to digital audio playback. The nice thing about software is that it is easy to change and hence lots of potential for customization exists, and of course still lots of room to play with on the output end.
Alvaius
PCBs
<font face=verdana size=1>
alvaius - glad youre back, hope your trip was pleasant. Although some people had questions as to your intentions with this project, i thought it was quite obvious from the first post. also, please read what i have to say to hifizen.
<b><font color=#ffffff>Begin Rant</font>
harry/fred (alters?) - i didn't mind reading your personal rant, but damn, re-read the thread a couple times before you post at least, man... don't have to be such pessimist all the time and if you don't know how to pronounce DSP it doesn't mean the (main) people authoring the thread and the project don't.
Sorry ...I meant "program" not "pronounce."
If you're one of those who, like me, want to learn more about this type of project, read, contribute and learn. If you're already a DSP programmer and you're interested, then read, and contribute. If you're a DSP programmer and you don't want to program a complex, masochistic DSP on a project that clearly stated that it will use a DSP, then go onto another thread.
harry/fred, your negativism and critical sarcasms are by no means constructive to the discussion - or the project. for people who are reading the thread or participating in it to learn something, your post was niether constructive, nor entertaining. this rant could go on and on, but I digress. Therefor...
<font color=#ffffff>End Rant</font></b><font face=verdana size=1>
hifizen - living in the city of angels... I am in a position to find quite good prices on labor intensive (expensive) tasks such as PCB fabrication. If you and/or the group thinks that it might be worthwhile, I will locate a PCB fabricator here in Bangkok and get some generalized pricing, and later a real quote - once we are further into the layout. I have never had a board fabricated for me here, but as an example my neighborhood CNC shop charges less than $25/hr for setup & milling, which is half what I used to give in the US (Cali.). So you can see for PCB fabrication, there are a lot of variables that could make the price considerably cheaper here than in North America.
I will probably be seeking out a board house soon for my own projects anyway, so I will check around for who has the best quality and who has the cheapest prices. Especially for multi-layer. I'm not sure if their are many places here that can do BGA mounting and X-ray inspection, I will check though, as mounting BGAs in the US can be a few hundred dollars (though I'm not sure about Canadian prices).
Anyway, cheers all, and goodnight Harry ... I mean Fred
Dan
<font face=verdana size=1>
alvaius - glad youre back, hope your trip was pleasant. Although some people had questions as to your intentions with this project, i thought it was quite obvious from the first post. also, please read what i have to say to hifizen.
<b><font color=#ffffff>Begin Rant</font>
harry/fred (alters?) - i didn't mind reading your personal rant, but damn, re-read the thread a couple times before you post at least, man... don't have to be such pessimist all the time and if you don't know how to pronounce DSP it doesn't mean the (main) people authoring the thread and the project don't.
Sorry ...I meant "program" not "pronounce."
If you're one of those who, like me, want to learn more about this type of project, read, contribute and learn. If you're already a DSP programmer and you're interested, then read, and contribute. If you're a DSP programmer and you don't want to program a complex, masochistic DSP on a project that clearly stated that it will use a DSP, then go onto another thread.
harry/fred, your negativism and critical sarcasms are by no means constructive to the discussion - or the project. for people who are reading the thread or participating in it to learn something, your post was niether constructive, nor entertaining. this rant could go on and on, but I digress. Therefor...
<font color=#ffffff>End Rant</font></b><font face=verdana size=1>
hifizen - living in the city of angels... I am in a position to find quite good prices on labor intensive (expensive) tasks such as PCB fabrication. If you and/or the group thinks that it might be worthwhile, I will locate a PCB fabricator here in Bangkok and get some generalized pricing, and later a real quote - once we are further into the layout. I have never had a board fabricated for me here, but as an example my neighborhood CNC shop charges less than $25/hr for setup & milling, which is half what I used to give in the US (Cali.). So you can see for PCB fabrication, there are a lot of variables that could make the price considerably cheaper here than in North America.
I will probably be seeking out a board house soon for my own projects anyway, so I will check around for who has the best quality and who has the cheapest prices. Especially for multi-layer. I'm not sure if their are many places here that can do BGA mounting and X-ray inspection, I will check though, as mounting BGAs in the US can be a few hundred dollars (though I'm not sure about Canadian prices).
Anyway, cheers all, and goodnight Harry ... I mean Fred
Dan
Size of letters
Hi madalo,
Please use normal size letters. I have a fairly big monitor (21 Inch) but these hurt my eyes!
I don't think Fred was pessimistic, only realistic. Before I install a DSP in my DAC a lot of things have to happen.
And all that processing will not hurt the digital signal? How about the clock feeding those Digital Engines? I even hear the difference between stopped clock technique and using the inverted LATCH technique taken from my Asynchronous Reclocker.[Both methods used to get separation into left and right audio channels in my NON-OS DAC]
Subtilities count in analog but also in digital I learned.
Hi madalo,
Please use normal size letters. I have a fairly big monitor (21 Inch) but these hurt my eyes!
I don't think Fred was pessimistic, only realistic. Before I install a DSP in my DAC a lot of things have to happen.
And all that processing will not hurt the digital signal? How about the clock feeding those Digital Engines? I even hear the difference between stopped clock technique and using the inverted LATCH technique taken from my Asynchronous Reclocker.[Both methods used to get separation into left and right audio channels in my NON-OS DAC]
Subtilities count in analog but also in digital I learned.
re-post; normal letters; minus rant
>>> snip
alvaius - glad youre back, hope your trip was pleasant. Although some people had questions as to your intentions with this project, i thought it was quite obvious from the first post. also, please read what i have to say to hifizen.
>>> snip
hifizen - living in the city of angels... I am in a position to find quite good prices on labor intensive (expensive) tasks such as PCB fabrication. If you and/or the group thinks that it might be worthwhile, I will locate a PCB fabricator here in Bangkok and get some generalized pricing, and later a real quote - once we are further into the layout. I have never had a board fabricated for me here, but as an example my neighborhood CNC shop charges less than $25/hr for setup & milling, which is half what I used to give in the US (Cali.). So you can see for PCB fabrication, there are a lot of variables that could make the price considerably cheaper here than in North America.
I will probably be seeking out a board house soon for my own projects anyway, so I will check around for who has the best quality and who has the cheapest prices. Especially for multi-layer. I'm not sure if their are many places here that can do BGA mounting and X-ray inspection, I will check though, as mounting BGAs in the US can be a few hundred dollars (though I'm not sure about Canadian prices).
>>> snip
>>> snip
alvaius - glad youre back, hope your trip was pleasant. Although some people had questions as to your intentions with this project, i thought it was quite obvious from the first post. also, please read what i have to say to hifizen.
>>> snip
hifizen - living in the city of angels... I am in a position to find quite good prices on labor intensive (expensive) tasks such as PCB fabrication. If you and/or the group thinks that it might be worthwhile, I will locate a PCB fabricator here in Bangkok and get some generalized pricing, and later a real quote - once we are further into the layout. I have never had a board fabricated for me here, but as an example my neighborhood CNC shop charges less than $25/hr for setup & milling, which is half what I used to give in the US (Cali.). So you can see for PCB fabrication, there are a lot of variables that could make the price considerably cheaper here than in North America.
I will probably be seeking out a board house soon for my own projects anyway, so I will check around for who has the best quality and who has the cheapest prices. Especially for multi-layer. I'm not sure if their are many places here that can do BGA mounting and X-ray inspection, I will check though, as mounting BGAs in the US can be a few hundred dollars (though I'm not sure about Canadian prices).
>>> snip
No doubt this a big (huge) project. I also would have thought a development kit is a good idea for anyone starting out in this. US$299 for a 21065 seems pretty reasonable considering it includes Visual DSP ++ although that is limited to use only with the evaluation board.
I've used the free Xilinx software with CPLDs and it seems to work very well. I'm using version 4.2i with service Pack 3. Verison 5.x didn't run on my machine for some reason.
Here's a link to the feature matrix of the xilinx software
http://www.xilinx.com/ise/devsys_feature_guide.pdf
The free one has most of the features.
I've used the free Xilinx software with CPLDs and it seems to work very well. I'm using version 4.2i with service Pack 3. Verison 5.x didn't run on my machine for some reason.
Here's a link to the feature matrix of the xilinx software
http://www.xilinx.com/ise/devsys_feature_guide.pdf
The free one has most of the features.
OK, I've neglected to follow this thread the last few days... let me catch up here.
alvaius:
The AD1896 are only a footprint on the board, and are looped through the FPGA. Hence, they're entirely optional. If you don't need them, just don't install them - it won't affect anything else. In fact, it will free up some FPGA IO which can then be reassigned to something else (connector footprints are also included on those IOs). The reason I added dedicated footprints for the 1896s was to to simplify the DSP code when working with multiple sample rates and multiple sources, by allowing conversion of everything to a single, synchronized sample rate, locked to the master clock. For my application, this was a given requirement. But, now that I think of it, I may just remove the 1896 footprints to further reduce the overall size and cost. We'll see how much space they consume in the end.
For any additional devices you want to use, the board is set up so that it is very simple to design a daughtercard and plug it in. Configure the FPGA to route your audio signals as desired, and you're off. The FPGA will take care of any data format and other interface problems. Thus, it's pretty well universal in this respect.
I thought of using the EZ-KIT, but it is expensive (building my own will be in the same ballpark cost), and much more restrictive in terms of flexibility of the final application. I think I can build a smaller form factor, more flexible and easier to use board which will be much more suitable for purist DIY requirements and is set up specifically for this application (eg, it has everything needed to drop it into a nice case and operate as a stand-alone unit without a PC connected). Another primary advantage will be in the quality of the audio bus clocking and interface to other boards. Also, the eval copy of VDSP++ which comes with the EZ-KIT can be obtained for free anyway (full-functionality version, 30-day limitation).
Oh yes - something i want to try, and will require further investigation, is the possibility of using <i>wavelet</i> techniques! I don't know enough about them at this point to say if they'll be of use, but this could be promising, so I'm excited about that...
Elso:
My board is designed to provide several clocking options. Basically, the footprint on the board will be for a full-size canned oscillator. But, you don't have to use a canned oscillator here, you can plug in a discrete oscillator you build on a little daughterboard, or use something like an ECL or LVDS receiver on a little daughterboard to receive the master clock from somewhere else (the most likely approach for me), such as a source-slaved PLL or a master clock in the DAC. Lots of possibilities here to tweak the clocking to suit your preference. Data going out to the DACs will be formatted and sent from the FPGA, which is directly slaved to this clock, and so if you're careful to implement a good clocking scheme, the data coming on to the DAC boards will already be very closely synchronized with the master DAC clock. Re-latching in close proximity to the DAC chips could be used to remove that last few dozen ps of timing uncertainties if you feel it is necessary to go that far.
alvaius:
The AD1896 are only a footprint on the board, and are looped through the FPGA. Hence, they're entirely optional. If you don't need them, just don't install them - it won't affect anything else. In fact, it will free up some FPGA IO which can then be reassigned to something else (connector footprints are also included on those IOs). The reason I added dedicated footprints for the 1896s was to to simplify the DSP code when working with multiple sample rates and multiple sources, by allowing conversion of everything to a single, synchronized sample rate, locked to the master clock. For my application, this was a given requirement. But, now that I think of it, I may just remove the 1896 footprints to further reduce the overall size and cost. We'll see how much space they consume in the end.
For any additional devices you want to use, the board is set up so that it is very simple to design a daughtercard and plug it in. Configure the FPGA to route your audio signals as desired, and you're off. The FPGA will take care of any data format and other interface problems. Thus, it's pretty well universal in this respect.
I thought of using the EZ-KIT, but it is expensive (building my own will be in the same ballpark cost), and much more restrictive in terms of flexibility of the final application. I think I can build a smaller form factor, more flexible and easier to use board which will be much more suitable for purist DIY requirements and is set up specifically for this application (eg, it has everything needed to drop it into a nice case and operate as a stand-alone unit without a PC connected). Another primary advantage will be in the quality of the audio bus clocking and interface to other boards. Also, the eval copy of VDSP++ which comes with the EZ-KIT can be obtained for free anyway (full-functionality version, 30-day limitation).
Oh yes - something i want to try, and will require further investigation, is the possibility of using <i>wavelet</i> techniques! I don't know enough about them at this point to say if they'll be of use, but this could be promising, so I'm excited about that...
Elso:
My board is designed to provide several clocking options. Basically, the footprint on the board will be for a full-size canned oscillator. But, you don't have to use a canned oscillator here, you can plug in a discrete oscillator you build on a little daughterboard, or use something like an ECL or LVDS receiver on a little daughterboard to receive the master clock from somewhere else (the most likely approach for me), such as a source-slaved PLL or a master clock in the DAC. Lots of possibilities here to tweak the clocking to suit your preference. Data going out to the DACs will be formatted and sent from the FPGA, which is directly slaved to this clock, and so if you're careful to implement a good clocking scheme, the data coming on to the DAC boards will already be very closely synchronized with the master DAC clock. Re-latching in close proximity to the DAC chips could be used to remove that last few dozen ps of timing uncertainties if you feel it is necessary to go that far.
Dave,
since you posted while I was composing... I'm hoping to reduce the cost down to be comparable with the EZ-KIT (see my previous comments), and I'm hopeful I can come close. The original budget was very large, but included a very <i>big</i> 4-layer board ($$$), as well as all of the SPDIF receivers, DACs and analogue circuitry. With all of that removed, the board should be much smaller and the cost will be much more reasonable. I'm hoping to come up with some more specific details soon, and I think I should have time to do this over the weekend.
since you posted while I was composing... I'm hoping to reduce the cost down to be comparable with the EZ-KIT (see my previous comments), and I'm hopeful I can come close. The original budget was very large, but included a very <i>big</i> 4-layer board ($$$), as well as all of the SPDIF receivers, DACs and analogue circuitry. With all of that removed, the board should be much smaller and the cost will be much more reasonable. I'm hoping to come up with some more specific details soon, and I think I should have time to do this over the weekend.
Hello,
If I remember the first intention – the goal was the DAC. Essentially, this is not about the DAC, this is about the filter. So, my question is – what filter? Is any possible to do something better than already exists on the market? Yes, there is not a lot of option among the stock filters, but, what is the reason for that? What is possible to do in digital domain at all? If you look at that function sin(x)+sin(x)/3, how would you define phase to get something without ringing, or minimal ringing? And if you go for Bessel in digital domain, will it be better than analog Bessel? Idea of the flexible platform is great, but does anyone have idea what to do with it?
At the other hand, as I know, main thing with new “upsampling” filters is that they have soften roll-off than digital filter used to have. That is all, and that is perceived as improvement. (I never had a chance to listen such players.) Is that everything that digital can develop after all these years? If yes – what do we essentially need here is the Bessel type filter done in digital domain. If not – most possibly we need to be smarter than others. And not in the terms of making such flexible platform, but in making some very special, never seen before, type of the oversampling or/i.e. digital low pass filter (everyone is free to understand these as same or different things).
One question more:
If one of those dCS’s filters is Bessel, from where does it came ringing? If it is not any, why dCS, which obviously plays with filters does not implement it?
Huh, not enough optimism to do anything right? Maybe, but my intention is not to deter anyone from this. Not at all! May not look as a contribution, but I can not contribute here the other way. Not familiar with DSP. Sorry. But will look what you guys do with this. Good luck!
Pedja
If I remember the first intention – the goal was the DAC. Essentially, this is not about the DAC, this is about the filter. So, my question is – what filter? Is any possible to do something better than already exists on the market? Yes, there is not a lot of option among the stock filters, but, what is the reason for that? What is possible to do in digital domain at all? If you look at that function sin(x)+sin(x)/3, how would you define phase to get something without ringing, or minimal ringing? And if you go for Bessel in digital domain, will it be better than analog Bessel? Idea of the flexible platform is great, but does anyone have idea what to do with it?
At the other hand, as I know, main thing with new “upsampling” filters is that they have soften roll-off than digital filter used to have. That is all, and that is perceived as improvement. (I never had a chance to listen such players.) Is that everything that digital can develop after all these years? If yes – what do we essentially need here is the Bessel type filter done in digital domain. If not – most possibly we need to be smarter than others. And not in the terms of making such flexible platform, but in making some very special, never seen before, type of the oversampling or/i.e. digital low pass filter (everyone is free to understand these as same or different things).
One question more:
If one of those dCS’s filters is Bessel, from where does it came ringing? If it is not any, why dCS, which obviously plays with filters does not implement it?
Huh, not enough optimism to do anything right? Maybe, but my intention is not to deter anyone from this. Not at all! May not look as a contribution, but I can not contribute here the other way. Not familiar with DSP. Sorry. But will look what you guys do with this. Good luck!
Pedja
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.