The pi does not really need a clock though right? Since it just passes I2S? And does not have to convert one type of "signal" to another which is what the WaveIO has to do. (Well ofcourse it does...since info comes in through the network card)
Anyway its task (Pi) is much simpler than the WaveIO's...is what I'm trying to get across...or ask...
Last edited:
There may be a language gap here
Having just read what you say it looks as though you have rather misjudged it. WE empathise with DDDAC and we share a very common view and goal. The Doede designs have been and are excellent. He has contributed so much, and it has paved the way for many DIYers to have very good DACs. And hopefully he will continue.
It is the major brands that make costly DACs and other hifi that is a major driver for DIY where we can make similar for 1/10 or less of the branded retail finished article. They give extra embellishments that are often superfluous.
Many DAC's are built with two output op amp stages.
The process of multistacking DACs is a proven excellent practice. Even BBC use it. But there will eventually be a chip that does it all and that is progress . We all applaud new developments and we can take on board any new emerging technology as DIYers with the help of the more knowledgable DIYers trying it first.
Without Doede, Lampizator, Gravesen many diyers would be lost.
So if the Wave IO already has a decent clock how come it needs a re clocker?
Also, has anyone tried to improve the Wave psu?
The on board regs look too small to do much about.........
An I2S connected pi sounds better than through the wave IO. I have ditched the Wave IO a while ago for that reason.
If I try to analyse why the pi on I2S sounds better the only thing I can come up with is that it does not use the USB transfer protocol.
The clock of the pi is not so good but I suspect the USB protocol and its coding/decoding will introduce a higher jitter level than the pi clock produces......
@Bas,
Every computer needs a clock! You cannot have the different processes running around in the processer without a planning! This would create total chaos 😀
The I2S signal must be clocked, so jitter is an issue here.
Last edited:
2SK208 CCS Update
Today I bypassed the cascode portion of the 2SK208"O" JFET CCS circuit on Pin 20. The sound is richer and more "musically exciting" compared with the cascode version which is very pleasing but somewhat polite sounding by comparison. I prefer the single JFET CCS sound.
I cannot say it definitely sounds better than the 2SK246Y because I am now running my DDDAC in my Florida system and I have added the reclocker circuit which makes comparison difficult, especially from memory.
As far as I'm concerned, both the 2SK246Y and 2SK208"O" can be used as a CCS upgrade over the 6K resistor on pin 20. Nige2000 found the 2SK208"R" to be the best CCS JFET he has heard on pin 20. All these JFETs have a low temperature coefficient with the 2SK208"O" changing current no more than 1% from 25C to 40C+.
If there is a two transistor CCS solution that is supposed to sound better I have no idea what to try next. Besides, the DDDAC sound is now so "right" I don't what improvement to look for.
If anyone is designing a new version of the DDDAC board there should be no hesitation to making an alternate provision for TO-92/SOIC-23 mounting pads with a SMD resistor and pot to be able to adjust the Vsource resistance to achieve 0.40 ma from pin 20.
Today I bypassed the cascode portion of the 2SK208"O" JFET CCS circuit on Pin 20. The sound is richer and more "musically exciting" compared with the cascode version which is very pleasing but somewhat polite sounding by comparison. I prefer the single JFET CCS sound.
I cannot say it definitely sounds better than the 2SK246Y because I am now running my DDDAC in my Florida system and I have added the reclocker circuit which makes comparison difficult, especially from memory.
As far as I'm concerned, both the 2SK246Y and 2SK208"O" can be used as a CCS upgrade over the 6K resistor on pin 20. Nige2000 found the 2SK208"R" to be the best CCS JFET he has heard on pin 20. All these JFETs have a low temperature coefficient with the 2SK208"O" changing current no more than 1% from 25C to 40C+.
If there is a two transistor CCS solution that is supposed to sound better I have no idea what to try next. Besides, the DDDAC sound is now so "right" I don't what improvement to look for.
If anyone is designing a new version of the DDDAC board there should be no hesitation to making an alternate provision for TO-92/SOIC-23 mounting pads with a SMD resistor and pot to be able to adjust the Vsource resistance to achieve 0.40 ma from pin 20.
Does it? I may have it completely wrong but I understood the advantage of asynchronous/ isochronous USB mode was that the data is only clocked by the receiver device. So the Wave has a clock then we are re-clocking with this mod???
After a long time without modding my beloved DDDAC, I've finally upgraded the four I/V resistor.
I had the bog standard Dales in place, with intermittent harshness and edginess, and veiled highs.
Not mentioning a significant offset I wouldn't suspect to mess the sound that much.
I've put Texas instrument's TX2575 133 Ohm 0.1% 30 minutes ago.
It's a serious buttkick, the music seems clother to me, unveiled, with good layering, VERY clean. The noise level got down with the offset, and so did my frustration.
One more sweet and no-brainer upgrade.
I had the bog standard Dales in place, with intermittent harshness and edginess, and veiled highs.
Not mentioning a significant offset I wouldn't suspect to mess the sound that much.
I've put Texas instrument's TX2575 133 Ohm 0.1% 30 minutes ago.
It's a serious buttkick, the music seems clother to me, unveiled, with good layering, VERY clean. The noise level got down with the offset, and so did my frustration.
One more sweet and no-brainer upgrade.
For the less technically inclined - you have removed the ZTX 450 and the LED? Just going back to the simple CCS? Wish I could speak "schematics"! Can only see them when they are drawn.
Thanks very much for your efforts!
Hi Doede,
Regarding the problem Stefan "Supersurfer" had not being able to get your DAC to work with the Philips CD-PRO drive because it has non-standard I2S word length,
Does that mean that the "newer" CD-PRO2 LF (VAU 1254/31) I2S out won't work either with your 1794DAC ???
Thanks,
Peter
Regarding the problem Stefan "Supersurfer" had not being able to get your DAC to work with the Philips CD-PRO drive because it has non-standard I2S word length,
Does that mean that the "newer" CD-PRO2 LF (VAU 1254/31) I2S out won't work either with your 1794DAC ???
Thanks,
Peter
you could use a schottky barrier in series, each diode adds about 0.2V. Avoid reverse biased elements like zeners. LEDs are fine.
Has anyone tried finding the supply voltage limit of the digital side of the 1794?
One wonders if it would allow five volts as the analogue side survives with eight?
Would eliminate a power supply for users of the BLUE mainboard (USB only) if one could power all of the digital devices from the same supply.
I am willing to try if someone thinks it might work.
One wonders if it would allow five volts as the analogue side survives with eight?
Would eliminate a power supply for users of the BLUE mainboard (USB only) if one could power all of the digital devices from the same supply.
I am willing to try if someone thinks it might work.