[Francis_Vaughan]

Sadly the reality of designing and prototyping the board is pretty much as has been outlined. The cost of entry into the four layer board arena is enough to exclude efforts like this. As far as I can see group buys are still prohibited on this forum (despite guarantees that a clarification would be forthcoming some months ago.) Even then, it is quite a few people taking a pretty deep breath to stump up the cash to run the prototyping. With enough enthusiasm and will it could be done. But I think the whole design forum needs to prove itself a little more. Eventually I might hope we will see such a thing.

Be careful about the justification for four layers. One thing we don't need is a power plane. This isn't a high speed digital design. It has a curious mixture of needs, and it has much more the needs of an RF design than a digital one. We don't need to supply lots of digital switching, rather we need to very carefully control the switching energy that is present, which leads us to lots of bypassing, and lots of low pass filtering on the power. Sure, four layers would make life easier, but providing a power plane would not be one of the uses of a layer.

 

[Clutz]

FYI: A running total of the putative parts costs so far (note, these parts were priced from Digikey.ca, not digikey.com, so the prices are in Canadian Dollars, and have been converted back to US assuming an exchange rate of $0.85. Parts costs using the PCM1794 DAC, excluding all the other important bits like ferrites, resistors, capacitors, diodes, in $US (which was where my $150 to $200 came from) $78, using the PCM1798 were is $68. A good clock like the Tent XO is about $30 (something I will probably spend money on, but woudln't count this into the budget)- I don't know how much an Elso Kwak clock would cost though.



DIR1703$5.31 15.31
PCM2707$8.44 18.44
OPA1632$4.89 14.89
REG101$2.95 38.85
AD1896$23.27123.27
PCM1794$25.42 125.42
PCM1798$13.44 00
CS8416$7.97 17.97
LM317$2.00 24
LM337$2.00 24
Multiplexer ??? ??? ???



CDN Sub$78.40
US Sub$66.64


The biggest concern I have with a four layer board is 'debugging' it. If we have ten people who were willing to participate in the PCB (just assuming for the moment), then I don't think that even $50 is necessarily outrageous for the PCB. My guess is that even a two layer PCB is going to cost about $20 to $30 for the number of boards we're going to be making, the extra $30 itself doesn't make me shrink. If it's a four layer PCB and we have a problem with the board,our abilities to debug and/or fix it are substantially reduced relative to the case of a two layer PCB, right? Anyway, I have no goals to turn this into a group buy. Even if I wanted to do that, I just don't have the money up front to buy a whole bunch of boards to sell to others later on.

I don't really understand the ins and outs of the importance of a ground plane- that's really something that is lost on me from reading the various threads on DACs.

Anyway, that's just my thoughts on the matter.

 

[00940]

Quote:

Originally Posted by dsavitsk I think there is a disconnect here. The only way you are going to get people on this or any other forum to put down that kind of money for a DAC, it had better be prototyped and it had better work way better than the other projects out there that can be built for about $100. This means that you are going to have to sell a ton of boards, you are going to have to eat the cost of prototyping, you are going to have to get it pretty much right on the first try to avoid several rounds of prototyping, and you are going to have to beat Yeo's board. Any one of these seems difficult, but doing all is probably an unreasonable thing to ask. Go with a 2 layer board, send it to batch PCB to make a handful, and be done with it. This is a fun learning project, not a best DAC ever made project. I think (and correct me if I am wrong) when Clutz says $150-200 he means that's what he wants to spend, not that he wants to put down $1000 to develop a DAC whose intellectual property rights and sunk costs he is going to give away so he can sell $35 to $55 DAC boards.

I broadly agree. While the final cost of a 4 layers board is reasonable (at eurocircuit, you can get eurocard 4 layers for around 25€ by 20), it's the prototyping which gets expensive fast.

The goal is to make the most of the budget. This being said, I think the project will appeal to different diyers than Yeo's boards. Aggressive oversampling/upsampling vs nos, last generation dac vs old philips TDA154*, ASRC vs reclocking, passive I/V vs opamps. That's like the opposite philosophy.

 

[Francis_Vaughan]

OK, I've being doing a bit of reading. The TUSB3200 supports Isochronous transfers in Asynchronous mode. This means it is possible to slave the PC data source to the DAC clock. So, after all the worry, it may be that this is actually the best answer.

Not home free. The TUSB3200 is a much more complex device than a PCM2704. The internal microcontroller must be programmed. But there is an SDK. (Available for free on the TI website.) I have not had a chance to look at it yet - but one hopes there is most of the code needed ready to roll - including asynchronous mode.

Also - one needs to know if the USB sound driver on the PC end will support asynchronous mode.

I have also been looking at the possibility of opto-isolation of the USB. It is a pain, but maybe achievable within the constraints of this project.

 

[00940]

Francis : drivers side should be fine under windows. Read this reply by Thomaspf : http://www6.head-fi.org/forums/showp...46&postcount=3

I know John Swenson on audioasylum is currently working on a usb asynchronous interface based on the tas1020b

 

[pipp]

Quote:

Originally Posted by mikechai What is the total cost to build? Any pictures? Any comparison with commercial DACs or sound cards?

As for price it all depends of parts chosen, as in all DIY projects. Im sure it can be buildt with fairly good components for say 50-100USD using the usb bus for power. However I would not recommend that. When trying that here there where some bakground noize that came of the usb bus power.

Im werry carefull when it comes to rating stuff I have buildt, because when you build something you're not always impartial. With that said, I have sold my ECD1

Im not going to say that the DAC is bether then the ECD1 thats for others to decide for themself. However I prefere the USB dac over the ECD1.

edit: Oh and pictures, I will post some when I have it all cased up and it looks all neat and nice

 

[ble0t]

Here's my interpretation of the schematic thus far...

I added a Neutrik XLR connector for the voltage rails and an LM317/LM337 section. In light of the recent posts concering PCM2706, I'll probably move the bypass caps in a bit closer as well. You can see the DIR1703 and the multiplexer on the lower right.

 

[00940]

Who here has a commercial version of Eagle ? We will need it for the final board. I suggest the best way to proceed is to have a few different versions of each operating blocks, fusion those version (which could be an interesting lesson in pcb layout) and then send the final version to someone with a commercial version of eagle to put all the blocks on the final pcb. We need some method, otherwise we are going to go crazy fast

 

[Clutz]

I thought it was steinchen?

 

[SnoopyRocks]

I've been outvoted so I'll drop the 4 layer request and spare everyone my responses. Back to work.

The schematic has grown quite and keeping track of the different versions is not as easy as it needs to be. It would be nice to have a regularly updated "official" version so that we're all on the same page. I think this is pretty important if different layout blocks done individually are going to match up and be merged later as 00940 suggests. Opinions?

 

[00940]

Should we set up a wiki ?

 

[dsavitsk]

Quote:

Originally Posted by 00940 Should we set up a wiki ?

How about a CVS system?

 

[Clutz]

Quote:

Originally Posted by dsavitsk How about a CVS system?

I don't mind hosting the files on twtch. I've never set up wiki or cvs, but I could create an account for someone. It's not the fastest machine in the world, an old OS X 10.3.9 (which can be upgraded to 10.4, if need be) G4 733 w/ 768MB RAM. I occasionally use it to run Mathematica, Matlab, R or C simulations on- but not that often.

 

[ble0t]

I have the Standard edition of EAGLE, so I can do a Eurocard size layout since I think that is what we'll need.

 

[00940]

allright then. Is there anyone actually willing to layout, beside Bleot and I ? If not, we can work together and then submit a common design to criticism here.