Audio Workshop D3 Kit

[lawbadman]

Hey guys
Has anyone tried the D3 kit from AudioWorkshop.
It looks very interesting and its only for US$60.
I want to get your opinions before I go and Buy it.

So what do you guys think of it?

http://www.audioworkshop.com.hk/index_en.htm

Thanks
Lawrence

 

[labjr]

I was looking at that kit too. Cool little project. It doesn't appear that too many people here have built them though.

I wanted to know how to self-power the receiver chip and change the crystal to a low jitter clock circuit.

I wonder not how it compares to onboard computer sound but how well it sounds next to a good CD player or transport DAC combo. Because, in the future, I would like all my music to be on computer.

 

[ChuckT]

I have built it and it is very very good.
I don't think you can go wrong with that price.
I had trouble trying to wire the pcm2706 in self-power mode, in the end, I just replace the USB power with a TL317 5V reference from the 7.5V (original was 6.8V) feed to 1543 dac.

 

[labjr]

Hi Chuck,

I thought in self power mode it needed a 3.3 Volt supply. at least that's what I think the data sheet specified. Seems strange since the USB pwers it with 5 volts correct ?

Also, did you modify the clock circuit at all?

THX

 

[ChuckT]

Yes, you are right, I did use a 3.3V but I couldn't get it to work, so I just replace the usb 5v.
I also replace the crystal with an external ocillator.

 

[labjr]

Which oscillator did you use.? Did you try different ones?

Did the oscillator make a difference in the sound?

From reading an old article about PCM2702 I was under the impression that in PCM 27XX series there is a PLL circuit in the chip which takes care of clock problems, No?

THX

 

[labjr]

Chuck,

I have one more question. Would you say the sound quality is good as a good CD player or just better than a computer sound card.

I don't want to build a project just to exceed the sound quality of my computer. That's pretty easy to do with any old CD player.

I want to do some serious listening of music of my CDs which I will move to my hard drive.

 

[ChuckT]

In my opinion, the sound from this dac is better than a lot of cd players.
I think this is due to TDA1543 dac and I think PCM2706 did a pretty good job on the I2S out too.
The ocillator made some difference, but replacing the usb 5v made a bigger difference.

 

[ronardi]

I'm going to build this kit too.
I got a question here. If I have a 12Mhz TCXO in hand, can I directly use it on the board to substitute the stock crystal?

 

[labjr]

I guess you could. I think Chuck said the difference probably more subtle than self-powering the PCM2706 though.

 

[ChuckT]

Please try it if you have an ocillator on hand, feed the ocillator some clean power and the difference will be bigger.
You'll need to pull out the 2 silver mica caps and solder.
Replace the decoupling Elna cap to Oscon cap too.
Change the coupling cap for the 1543 to a film cap.

 

[ronardi]

Finally, my AW-D3 kit arrived this Friday, and I have finished the soldering work already. However, poor me... when I connect it to my headphone amp, music did come out, but in low volumn while noise (hiss) was so loud. I believe it should be something wrong in the soldering process.

Despite the small sound volume and big hiss, there is another problem that the D3 has two mode by opening or shorting J2, it worked when I first finished it but when I resolder all joints on the PCB to prevent cold soldering, the J2 doesn't work since then. The light of the SPDIF kept lit and anolog output kept active. I should have short something as the two modes worked before!! plz help as I've been driven mad....

 

[labjr]

I got this info from TI engineering support about self-powering the pcm2706. I hope it's accurate.


The Vbus pin of device must be connected to 3.3V. The Vbus of USB interface should be connected to HOST (Pin3)-pin. Also, analog 3.3V power supply should be prepared, to provide to Vccx of device.

 

[John Swenson]

I haven't actually built this kit, but I have built several 2706 DACs from scratch and have helped several people with the construction and modification of this kit.

The guts of the 2706 runs at 3.3V, but it has an internal 3.3V regulator. What TI refers to as self power mode is bypassing the internal 3.3V regulator and using your own. You do not have to do that in order to not use the USB bus power. There is nothing wrong with using an external 5V supply and feeding that into the internal 3.3V regulator. Using your own 3.3V reg is a bit more complicated to do and not really necassary.

To use an external supply all you have to do is take out the inductor that goes between the USB connector and the 2706 and feed in a good clean 5V supply. The supply on the board is not bad, although I would use a separate regulator for the 2706. I personally like the Panasonic AN8005. The chip is wired up to already use the internal 3.3V reg.

If you really want to use your own external 3.3V reg you have to change the sense of the PSEL pin (I don't have the spec sheet in front of me and I don't remember which way it goes, but I THINK I remember it being gnd for an external 3.3V supply). With this kit its much easir to just feed it 5V and let the internal reg do the work.

On the clock domain the clock used DOES efffect the sound (from the description of the chip it shouldn't but it does) . The Cryptek 12 MHz oscillator from Mouser is a popular one to use for 2706 designs. Its 5x7mm SMT package so its a bit fun to solder in. It runs off 3.3V, you can easily use the 3.3V from the 2706 to power it, or use your own regulator. (again I like the AN8035 for this).

After these two changes the sound is quite good, but can be significantly improved! I have designd an output stage that works very well with this kit that radically improves the sound, but its fairly complicated and takes some significantl voltages (at least 70 volts!) I've got the schematic around somewhere, if anybody is interested I'll post it here.

John S.

 

[lawbadman]

Oh yes John, most certainly interested. Lets see the schematic.
Do you know of any other mods to improve the sound?



Lawrence

 

[John Swenson]

An externally hosted image should be here but it was not working when we last tested it.

This is an output stage that works really well with the kit. Its based on the CCS circuits by Gary Pimm, with a few value modifications. The schematic is for one channel, you need two of these! The primary advantage to this over what comes with the kit is the impedance the DAC output sees. Since it is a current output it wants to see a very low impedance, the simple resistor I/V used in the kit loads it with WAY too high an impedance for really good results. This stage loads the DAC with 10 ohms instead of 1.3K, that makes a huge difference in the sound.

The schematic shows a buffer stage, you don't really need it if you have a fairly high impedance preamp input and you have fairly short interconnects. If the input impedance of the preamp is 10K or less it might be a good idea to use the buffer. It doesn't harm anything to use the buffer (its a VERY good sounding buffer) and it can improve the sound in most situations, but in some the improvement is going to be slight, in other systems it will be a big improvement.

To use this stage the ouptut pins of the DAC go directly to the stage input, NOTHING else on the kit board should be connected to that signal, thus any resistors or caps on the output of the DAC should be removed.

This circuit assumes you are running the 1543 at 8Volts or so, not the usual 5 volts. After the stage is constructed its hooked up to its power supply but NOT the DAC and the pot is adjusted until the voltage on the input pin is 3.5V. Then the DAC outputs are connected, they should be at 3.5V. If running with a 5V supply on the 1543, you can lower the voltage on the output pin of the DAC to 2.5-3V, but it won't sound quite as good as using 8V supply and 3.5V on the output pin.

The circuit can be built either by hand on perfboard or you can get PC boards from gary Pimm. He sells little boards that are for each CCS circuit (the self bias CCS board). If you don't use the buffer you need two, if you do use the buffer you need 4.

The stage needs 70V DC to run, this works well from a 50V transformer, you CAN use a bridge rectifier and cap (make sure they are rated for at LEAST 100V) but I prefer a choke supply with bridge rectifier, cap, choke and cap. This is more complicated and expensive (I just use fairly cheap Hammond choke for this) but it does sound better. I don't have the supply schematic up yet.

BTW in order to get the stage working they way I want it there is 30mA going through each section (60mA for stereo, double that if using the buffer), so the power supply needs to handle that AND the resistors have to be a high enough wattage to handle it as well. R20 has about 40V across it so its going to be dissapating 1.2 watts so a little 1/4 watt resistor will not do here! The 5 watt Kiwamees work very well here, as do a parallel or series combination of the ohmite big carbon comps. R1 on the buffer circuit also has the same issues but not quite as bad, its only .6 watts or so.

The type of resistor used for R20 is very critical to the sound, but make sure it can handle the power. At various times I've had 4 resistors in parallel to get the right power handling when trying a resistor type that was not available in big wattage ratings.

All the parts for this are available from Mouser and cost about $8 per channel. The boards from Gary cost $16 so its $24 per channel if you go that route.

As with any good circuit the output caps are very important, play around with the type and value, with this stage the sound is so good that the caps will really make a difference. I've had one person use transformers on the outputs instead of caps, this worked very well.



I hope this doesn't sound too difficult to build, its very much worth the sonic improvement.

John S.

 

[jkeny]

Well Guys,

I know this thread has been dormant for a while but here goes.

Did anybody else get this USB DAC kit? I built one recently & found the sound excellent from headphone output (PCM2706 internal DAC) - can't get I2S to TDA1543 external DAC working - think the chip might be faulty - checked my board & soldering - no problems.

Ronardi, did you fix your problem re J2 jumper & I2S?

Looking forward to resolving the I2S issue.

Has anybody built John S output stage - how does it sound?