Nice initial review of first listen to the dac Lucas...definitely would give it time for parts to burn in before really judging it though...but from what your saying...your already very happy with it
With regards the 431shunts...i also think that if you can connect the Salas shunts directly(bypassing the 431shunts) that it would sound better. Providing the connections from Salas shunt to dac board was kept very short.
So im looking forward to your comparision, with 431's and without...after burn in of course.
The way i understood Thorstens comment was that, He was not familier with the Salas shunt and so suggested the above. However those of us who have used the Salas shunt know it is damn good at what it does, so as long as the connections are kept short i believe there will be an improvement. I guess the only way to find out is by trying it.
I will wait to hear what you think of the differences Lucas and then may try it myself too.
Alon
That's exactly how I read it too - Thorsten wasn't familiar with the Salas shunts, nor their renown, and so assumed them to be fairly standard stuff. Those of us with experience of these shunts know that they are top notch. Even for double DAC, there is no intermodulation, because both loads are identical in current draw per millisecond, so secondary shunts are needless. I've just been discussing it with Salas on his shunt reg. page if you want to read what has been said about this. I always felt to do this as a hunch if nothing else, and now I am quite sure of it.
Regarding my rambling review, yes it's amazing the issues it had when first switched on, and how just two hours later things had opened up. I've never really noticed burn-in to this extent on anything else I've built - it really changes over time, including some distortions disappearing. I wonder which components in particular change over time and heat. My guess is transistors, like the DAC chip itself.
5m USB cables don't work - they carry power (of sorts) but computer doesn't register Teradak with 5m cable.
The USB spec calls for the host to be able to supply 500 mA and for the cable to be able to carry it. (IIRC, the 5-metre limit is down to issues with propagation delay.)
I note that the Teralink takes ~350 mA. It might just be drawing more than your Macbook can deliver over a lengthy cable. It happens. (If OTOH the Teralink is separately powered, you can safely ignore the above . . .)
Whatever, with DD's USB-to-I2S module, I got a very significant improvement by using this device to connect the USB link:
ADuM4160 USB Isolator board, populated « Circuits@Home
For 'scope traces, see Step 5 here:
ADuM4160 USB Isolator Assembly Guide « Circuits@Home
The effort to "tune" a USB port certainly matters with audio which is typically a "real-time" process. Though my experience is with a PC, Mac users report significant improvements from various tuning procedures.
It seems a shame to sell the farm to build a top-of-the-range DAC and not, sooner or later, feed it the best possible signal.
The USB spec calls for the host to be able to supply 500 mA and for the cable to be able to carry it. (IIRC, the 5-metre limit is down to issues with propagation delay.)
I note that the Teralink takes ~350 mA. It might just be drawing more than your Macbook can deliver over a lengthy cable. It happens. (If OTOH the Teralink is separately powered, you can safely ignore the above . . .)
Whatever, with DD's USB-to-I2S module, I got a very significant improvement by using this device to connect the USB link:
ADuM4160 USB Isolator board, populated « Circuits@Home
For 'scope traces, see Step 5 here:
ADuM4160 USB Isolator Assembly Guide « Circuits@Home
The effort to "tune" a USB port certainly matters with audio which is typically a "real-time" process. Though my experience is with a PC, Mac users report significant improvements from various tuning procedures.
It seems a shame to sell the farm to build a top-of-the-range DAC and not, sooner or later, feed it the best possible signal.
USB Cable
Lucas,
I use an active repeater Cable, which is 5m long plus 1m normal USB cable as an extention. No problem to recognize the Teralink X2, when powered by USB!!! Just tried it and it worked.
The brand is DIGITUS: DIGITUS USB 2.0 repeater cable | Digitus
Price: around 7 Euro.
Btw, I doubt, that Hama products are made in Germany. Maybe I am wrong, but they probably buy their stuff, where all the world is buying currently....
Cheers Ernst
Lucas,
I use an active repeater Cable, which is 5m long plus 1m normal USB cable as an extention. No problem to recognize the Teralink X2, when powered by USB!!! Just tried it and it worked.
The brand is DIGITUS: DIGITUS USB 2.0 repeater cable | Digitus
Price: around 7 Euro.
Btw, I doubt, that Hama products are made in Germany. Maybe I am wrong, but they probably buy their stuff, where all the world is buying currently....
Cheers Ernst
Hi Ryelands,
very interesting advice!
So this device could be powered by the USB host, if i powered my X2 with an external PS, right?
So this device could be powered by the USB host, if i powered my X2 with an external PS, right?
No - best to use the 5-volt line to the X2 (it only draws ~10mA) - or its own PSU though I doubt that's necessary. I can't do it now but I'll try to post later a pic of one connected to a DD module.
The idea is that the host PC is isolated by the chip's transformers on both the Gnd and +5v.
* I haven't tested it at >48 KHz sampling rates and so cannot promise it will work at 96 KHz though the specs suggest it should. But at 44.1/48, it works very well indeed and markedly improves the sound.
No - best to use the 5-volt line to the X2 (it only draws ~10mA) - or its own PSU though I doubt that's necessary. I can't do it now but I'll try to post later a pic of one connected to a DD module.
The idea is that the host PC is isolated by the chip's transformers on both the Gnd and +5v.
* I haven't tested it at >48 KHz sampling rates and so cannot promise it will work at 96 KHz though the specs suggest it should. But at 44.1/48, it works very well indeed and markedly improves the sound.
Salas is, of course, the best person to consult about this, but he tends to advise for a much smaller, preferably polypropylene cap for the CCS position (nearest input) and a larger electrolytic in the other position. In the DCB1 power supply, for example, which is a Salas low voltage shunt regulator for a DC 10v load, the CCs cap can be any value between 0.22uF and 100uF, and the second is usually 100uF. The values will change the sound significantly. 470uF is very large for this application.
Try Rubycon ZLH 220uF 35v
RUBYCON|35ZLH220M8X11.5|CAPACITOR, 220UF, 35V | Farnell United Kingdom
I haven't used them, but they have been highly recommended to me for this very application.
RUBYCON|35ZLH220M8X11.5|CAPACITOR, 220UF, 35V | Farnell United Kingdom
I haven't used them, but they have been highly recommended to me for this very application.
It is not Wima, but another MKP - 1uF, on the CCS position, as per Salas instructions/suggestion, in the DCB1, not as you said at the output of the circuit (which would be truly wrong), and not in this DAC circuit either.
I have been a good boy with this DAC, done as I was told in the BOM, and used 220uF Nichicon Fine Gold
They sound fine to me. I am not one for swapping quality components like this for other quality components unless there is a good reason - I would normally be more inclined to change values. There may be some advantage, however, to putting a much smaller value (like 1uF to 10uF) CCS (first) position film cap in place of the CCS 220uF electrolytic...just as Salas suggested with the DCB1, and you observed I have done.
The DAC is sounding very good (still only 1x DAC chip) - more photos soon when 2x DAC chips!
I have been a good boy with this DAC, done as I was told in the BOM, and used 220uF Nichicon Fine Gold
They sound fine to me. I am not one for swapping quality components like this for other quality components unless there is a good reason - I would normally be more inclined to change values. There may be some advantage, however, to putting a much smaller value (like 1uF to 10uF) CCS (first) position film cap in place of the CCS 220uF electrolytic...just as Salas suggested with the DCB1, and you observed I have done.The DAC is sounding very good (still only 1x DAC chip) - more photos soon when 2x DAC chips!
No.
The X2 +5V Reg. is connected additionally to the +15V line of the +/- PS-Module.
It is not Wima, but another MKP - 1uF, on the CCS position, as per Salas instructions/suggestion, in the DCB1, not as you said at the output of the circuit (which would be truly wrong), and not in this DAC circuit either.
I have been a good boy with this DAC, done as I was told in the BOM, and used 220uF Nichicon Fine Gold
The DAC is sounding very good (still only 1x DAC chip) - more photos soon when 2x DAC chips!
I see that clarifies. Should have better paid attention what I was looking at (DCB1, not DAC).
By the way, EC-designs has identified that his ultra-simple output stage is not optimal either. He is now using a bit more complex topology.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.