It sound pretty damn good for the price - chips (Burr-Brown pcm2702 and an opa2353) are free samples thanks to Texas Instruments. The other components are common ones (harder to get audiophile ones in my part of the world), just the audio path is matched at 1% left vs. right. Total cost for me... about 15 euros (ok the pcb is a cheap non-professional one but it works).
Compared to the readily made professional quality dem-pcm2702 board which is priced at $149... do the math...
Briefly tested on a vincent amp with dynaudio contour speakers it sounded right compared to a vincent cdplayer.
It does only 16bit stereo with 32/44.1/48khz sampling rate.
One thing though, for better quality it should use an external psu with separate 3.3V (digital part) and 5V (analog part) regulators. This version derives both from usb power for the sake of simplicity.
This version is an output only card; for analogue input and spdif input/output refer to the pcm2903 chip.
Compared to the readily made professional quality dem-pcm2702 board which is priced at $149... do the math...
Briefly tested on a vincent amp with dynaudio contour speakers it sounded right compared to a vincent cdplayer.
It does only 16bit stereo with 32/44.1/48khz sampling rate.
One thing though, for better quality it should use an external psu with separate 3.3V (digital part) and 5V (analog part) regulators. This version derives both from usb power for the sake of simplicity.
This version is an output only card; for analogue input and spdif input/output refer to the pcm2903 chip.
You could build somthing like this:
http://www.overclockers.com.au/~mwp/gallery/?l=My DIY/USB-DAC-HeadphoneAMP
All built using TI sample parts.
Best headphone amp ive ever built, and thats even with its switching headphone amp supplies.
http://www.overclockers.com.au/~mwp/gallery/?l=My DIY/USB-DAC-HeadphoneAMP
All built using TI sample parts.
Best headphone amp ive ever built, and thats even with its switching headphone amp supplies.
All info is on TI's website.
For the DEM-PCM2702 board have a look at http://focus.ti.com/docs/toolsw/folders/print/dem-pcm2702.html
The schematics is the one there without the external power connector and with transistors replaced with local accessible parts (they are not in the signal path).
Because the pcb in the datasheet is at a horrible resolution it helped as an excuse to learn a little eagle and design my own pcb. It's not optimal as I do not have enough pcb design knowlegde. Double sided, bottom layer has only a few traces, rest being the ground plane. Eagle source or pdf available on request.
Chips should be soldered with a hot air station, rest goes with a normal temperature-controlled soldering station with a fine tip. PCB should be done using the photo method, can be done at home if you did such a thing before.
Board recognized by windows xp (as "usb speakers") without need of drivers. Supports volume control through windows mixer. Just plug in into usb port and it becomes the default sound card.
I'm still shocked it works
For the DEM-PCM2702 board have a look at http://focus.ti.com/docs/toolsw/folders/print/dem-pcm2702.html
The schematics is the one there without the external power connector and with transistors replaced with local accessible parts (they are not in the signal path).
Because the pcb in the datasheet is at a horrible resolution it helped as an excuse to learn a little eagle and design my own pcb. It's not optimal as I do not have enough pcb design knowlegde. Double sided, bottom layer has only a few traces, rest being the ground plane. Eagle source or pdf available on request.
Chips should be soldered with a hot air station, rest goes with a normal temperature-controlled soldering station with a fine tip. PCB should be done using the photo method, can be done at home if you did such a thing before.
Board recognized by windows xp (as "usb speakers") without need of drivers. Supports volume control through windows mixer. Just plug in into usb port and it becomes the default sound card.
I'm still shocked it works
danb1974
Could you please comment a little on the sound quality. It is only a 16 bit converter...
When I saw the type number (2702) I was extatic as it reminded me of the very best BB DAC (pcm1702 and 1704) , but I was disappointed.
Integrated digital filter. Integrated post analog filter and OPamp.
Looks to me as a poor performer
Koldby
Could you please comment a little on the sound quality. It is only a 16 bit converter...
When I saw the type number (2702) I was extatic as it reminded me of the very best BB DAC (pcm1702 and 1704) , but I was disappointed.
Integrated digital filter. Integrated post analog filter and OPamp.
Looks to me as a poor performer
Koldby
It sounded comparable to a vincent cdplayer (tested on vincent line, dynaudio contour speakers). This is good enough for me given the price (around 20 bucks). If you heard one and didn't like the sound, no problem. If you didn't hear one please refrain from "this is how I think it would sound" type comments.
Jitter is minimized by the patented adaptive pll algorithm (called spact) which ensures (a) stable clock synchronized to the rate data is coming from usb and (b) very fast lock-up time.
If you read the story of the engineer that worked on it you'll see it's not just a dac patched into an usb receiver with a BB sticker on it to make it sell. And it's not trivial to make a dac with an usb input.
Jitter is minimized by the patented adaptive pll algorithm (called spact) which ensures (a) stable clock synchronized to the rate data is coming from usb and (b) very fast lock-up time.
If you read the story of the engineer that worked on it you'll see it's not just a dac patched into an usb receiver with a BB sticker on it to make it sell. And it's not trivial to make a dac with an usb input.
Why not use the PCM2706... it connetcs to the USB port so it doesn't have to deal with SPDIF conversion... also it has both analog and IS2 output... needs no external drivers, (well you do need up to date windlows)... I am currently bussy assembling mine, and will report back when its done.... if I don't make smoke...
danb1974 said:
Dane, do you have any idea what jitter values are to be expected?
Don't trust marketing. Just because they brag about it, or because it's hard to design it doesn't mean it is good... It has to deliver good sound as well.
And by using the internal DAC or the SP/DIF output, that doesn't do it in my case. But that's because I know what to expect...
No idea about jitter values. Datasheet only mentions things like thd+n which is typical 0.002% and max. 0.005%. Which is close to theoretical minimum for 16bit/44.1khz.
In my case jitter is the least problem I worry about, partly because there is a long way to at least decent speakers and amp, partly because I'm not sure anybody can point out jitter efects from speaker/room coloration or source quality (and no, pops and clicks are not jitter but a badly scratched cd).
I did not know what to expect from my pcm board, I build it out of curiosity, but I'm really pleased with the results.
BTW, can you give me an example of a system good enough that jitter efects (what do they sound like?) become clearly audible and can be recognized in a blind test?
In my case jitter is the least problem I worry about, partly because there is a long way to at least decent speakers and amp, partly because I'm not sure anybody can point out jitter efects from speaker/room coloration or source quality (and no, pops and clicks are not jitter but a badly scratched cd).
I did not know what to expect from my pcm board, I build it out of curiosity, but I'm really pleased with the results.
BTW, can you give me an example of a system good enough that jitter efects (what do they sound like?) become clearly audible and can be recognized in a blind test?
On just a fairly decent system , you should be able to hear jitter effects. Of course high values of jitter reduction is easier to hear.
How it sounds is perhaps a little dependend on the kind of jitter, but better focus and reduction in smearing of sounds is typical effects of reduction of jitter.
If you listen to a DAC/Drive combo capable of using both S/Pif and I2S you can easily hear the difference.
The two most striking demonstration of the bad effects of jitter was both performed with a NAGRA D digital tape recorder and a seperate DAC.
One in winter CES by Gene Pope where he change between standard AES/EBU and clock regeneration from that and a distributed clock from an outboard clock.
Another at home at Mr. Pope where I had modified the NAGRA D so it could use the proffesionel S-Dif format (Three wires with L-data, R-data and clock). Huge difference!!
You are right that it is not enough to look at data sheets to know how it sounds, but on the other hand - experinece tells me that DAC chips with integrated digital filter, analog filter and op-amps NEVER have sounded as good as f.ex. PMD100 filter PCM1704 and AD797 op amp -NEVER!
Koldby
How it sounds is perhaps a little dependend on the kind of jitter, but better focus and reduction in smearing of sounds is typical effects of reduction of jitter.
If you listen to a DAC/Drive combo capable of using both S/Pif and I2S you can easily hear the difference.
The two most striking demonstration of the bad effects of jitter was both performed with a NAGRA D digital tape recorder and a seperate DAC.
One in winter CES by Gene Pope where he change between standard AES/EBU and clock regeneration from that and a distributed clock from an outboard clock.
Another at home at Mr. Pope where I had modified the NAGRA D so it could use the proffesionel S-Dif format (Three wires with L-data, R-data and clock). Huge difference!!
You are right that it is not enough to look at data sheets to know how it sounds, but on the other hand - experinece tells me that DAC chips with integrated digital filter, analog filter and op-amps NEVER have sounded as good as f.ex. PMD100 filter PCM1704 and AD797 op amp -NEVER!
Koldby
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.