Even plain Elektro-optik link to normal DACs is better than anything I have tried so far:
AMNESIS VFET plus Elektro-optik link. - YouTube
Cheers,
M.
AMNESIS VFET plus Elektro-optik link. - YouTube
Cheers,
M.
Makes me think that people having academic knowledge and distributes bad or good notes but not too much in the diy should come not only with critic notes like a professor but also contribute "academicly".
TNT, may you mind to write a sticky about jitter, filters, oversamplings, speed of circuits and clock and how they impacts the sound of the dac to be able to hear the difference with ears and to know the culprits by listening ?
Should be usefull both for PCM and DS dac chips, NOS and no-NOS .
I think such good professor people like you are able to write such sticky, I am not. Maybe we focus too much on some things/concepts without be able for most of us , modest enthusiasts, not to make the difference at ears ?
This Digital Line section disearves such a sticky imho ?
TNT, may you mind to write a sticky about jitter, filters, oversamplings, speed of circuits and clock and how they impacts the sound of the dac to be able to hear the difference with ears and to know the culprits by listening ?
Should be usefull both for PCM and DS dac chips, NOS and no-NOS .
I think such good professor people like you are able to write such sticky, I am not. Maybe we focus too much on some things/concepts without be able for most of us , modest enthusiasts, not to make the difference at ears ?
This Digital Line section disearves such a sticky imho ?
From a theoretical perspective this isn't true. A "virused" jitter signal can be brought back to a digital file as absent of jitter. It can then be reintroduced later in time into another DAC network with some "new" form of jitter introduced. in other words it can be eliminated by returning it back to a digital file. Hence if a digital signal can be brought back to the equivalent of a digital file it can never be game over.
Although it can be argued that the jitter can remain in the system in forever time (as decaying by some form) the level of such jitter at some point can be considered irrelevant, and whereupon other artifacts can take over as the more dominant noise source (anything not signal). This is not to suggest that jitter can't be a major contributor to the quality of reproduction.
Exactly what I said in my previous posts, if the jitter in the DAC is source dependent it does mean that there is something wrong in the interposed devices that don't do correctly the job (remove the jitter coming from the source).
Here you can find some measurements of a FIFO buffer device
The Well Tempered Master Clock - Building a low phase noise/jitter crystal oscillator
The jitter is only partially removed so a part of the jitter coming from the source is reflected to the output affecting the timing of the DAC.
Not even a very good master clock helps to remove totally the jitter of the source, so the DAC is source dependent.
But this does not mean that the jitter could not be totally removed, I have already explained the ritght way to do the job in post #7686
Building the ultimate NOS DAC using TDA1541A
Here you can find some measurements of a FIFO buffer device
The Well Tempered Master Clock - Building a low phase noise/jitter crystal oscillator
The jitter is only partially removed so a part of the jitter coming from the source is reflected to the output affecting the timing of the DAC.
Not even a very good master clock helps to remove totally the jitter of the source, so the DAC is source dependent.
But this does not mean that the jitter could not be totally removed, I have already explained the ritght way to do the job in post #7686
Building the ultimate NOS DAC using TDA1541A
There is a lively discussion about ECDESIGNS products on AS forum:
ECdesigns - Page 39 - DAC - Digital to Analog Conversion - Audiophile Style
Thank you
Matt
Hi I took aa quick look at the pages from the link.
.
It is about optical isolation.
I think that it is possible to achieve good isolation of audio digital interface.
- First isolate visual unit (dysplay, TV etc) using typically HDMI cable. With optical HDMI like this one. Amazon.com: FIBBR Crystal Fiber Optic HDMI Cable High Speed 18Gbps- 4K 60hz Hdmi Cable HDCP 2.2, HDR 10 and Dolby Vision, 4:4:4, HDMI 2.0 Cable for UHD TV, Blue-ray Player, PC, Xbox, PS4 PS4Pro, Projector, 3.3ft: Industrial & Scientific
or this
Amazon.com: HDMI Fiber Optic Cable HDMI 2.0 18Gbps Active Optical Cable (AOC) / UHD 4K 60FPS 4:4:4 (5ft (1.5m)): Electronics
Maybe better because it is clear from the picture that metal cans of the 2 connectors are NOT conneted. So the groundings from HDMI is not connected to groundings metal can at the USB.
.
Next Adopt digital bus from the usb interface to optical driver and output. There are some Thunderbolt cables with differential driver-to-fibreoptics and opposite.
Thunderbolt™ Optical Cables | Corning
I think that it is possible to transfer 4 lines digital bus with this type of connection without making special cables?
.
it is only a need to design digtal circuit from single ended signal's to differential standard?
.
Or use some fibre optics conns (Google: Avago) and cables for every digital audio line like MSB did.
.
Tip for very good USB standard cable Phoenix contact. (They using firewire cable...)
Security Check
.
cheers
.
It is about optical isolation.
I think that it is possible to achieve good isolation of audio digital interface.
- First isolate visual unit (dysplay, TV etc) using typically HDMI cable. With optical HDMI like this one. Amazon.com: FIBBR Crystal Fiber Optic HDMI Cable High Speed 18Gbps- 4K 60hz Hdmi Cable HDCP 2.2, HDR 10 and Dolby Vision, 4:4:4, HDMI 2.0 Cable for UHD TV, Blue-ray Player, PC, Xbox, PS4 PS4Pro, Projector, 3.3ft: Industrial & Scientific
or this
Amazon.com: HDMI Fiber Optic Cable HDMI 2.0 18Gbps Active Optical Cable (AOC) / UHD 4K 60FPS 4:4:4 (5ft (1.5m)): Electronics
Maybe better because it is clear from the picture that metal cans of the 2 connectors are NOT conneted. So the groundings from HDMI is not connected to groundings metal can at the USB.
.
Next Adopt digital bus from the usb interface to optical driver and output. There are some Thunderbolt cables with differential driver-to-fibreoptics and opposite.
Thunderbolt™ Optical Cables | Corning
I think that it is possible to transfer 4 lines digital bus with this type of connection without making special cables?
.
it is only a need to design digtal circuit from single ended signal's to differential standard?
.
Or use some fibre optics conns (Google: Avago) and cables for every digital audio line like MSB did.
.
Tip for very good USB standard cable Phoenix contact. (They using firewire cable...)
Security Check
.
cheers
Ecdesigns,
I just read this post from a while back. I have played with "current drive" and the taming of impedance peaks with mechanical methods (flow resistance) on the drivers. I did not continue the investigations due to external reasons, I think it could be made to work very well and there can be multi driver/way systems. Have you tried this with your speakers? (ThorstenL wrote ages ago about this and it grabbed my attention since then).
S
Power DAC is a great concept
While the proof will be in the pudding, the PowerDAC concept is compelling.
Once we get a traditional DAC to its peak performance (clock, power, I/V output buffer) every optimization of the sound of a system tends to be eliminating subtractive elements of the electronics. IE removing signal degradation due to coupling caps, resistors, signal transformers etc. Having seen the improvements just by replacing or eliminating these elements it is hard to imagine how good a system could be if all of this was simply eliminated.
I guess the rest would be to pair it with speakers that work well with it in the room.
While the proof will be in the pudding, the PowerDAC concept is compelling.
Once we get a traditional DAC to its peak performance (clock, power, I/V output buffer) every optimization of the sound of a system tends to be eliminating subtractive elements of the electronics. IE removing signal degradation due to coupling caps, resistors, signal transformers etc. Having seen the improvements just by replacing or eliminating these elements it is hard to imagine how good a system could be if all of this was simply eliminated.
I guess the rest would be to pair it with speakers that work well with it in the room.
Another option for those intrigued about Elektro-optik link.
Use an USB-to-SPDIF/Toslink converter, like the one in the picture (inexpensive and good), remove the Toslink Transmitter and wire the Toslink output to the SPDIF connector. In this particular module, which I am very fond of, the coaxial output is transformer coupled so I removed the Tx and made sure that the ground of the connector was properly grounded.
Edit to be more clear:
1) Remove the Optical Transmitter.
2) Remove the SPDIF coupling transformer.
3) Connect "optical signal" to "coaxial signal" and "ground to coaxial ground".
Of course you still have to install the Red LED/Toslink adaptor to one end of the signal cable to be connected to your DAC's optical input. This is to have totally clean and high Q sound from a PC source.
Of course, your software will then be the limiting step to great sound...
I hope you like it,
M.
PS: of course, you can also keep the Toslink Transmitter and wire the Coaxial connector in parallel, to be able to compare between cable formats...
Use an USB-to-SPDIF/Toslink converter, like the one in the picture (inexpensive and good), remove the Toslink Transmitter and wire the Toslink output to the SPDIF connector. In this particular module, which I am very fond of, the coaxial output is transformer coupled so I removed the Tx and made sure that the ground of the connector was properly grounded.
Edit to be more clear:
1) Remove the Optical Transmitter.
2) Remove the SPDIF coupling transformer.
3) Connect "optical signal" to "coaxial signal" and "ground to coaxial ground".
Of course you still have to install the Red LED/Toslink adaptor to one end of the signal cable to be connected to your DAC's optical input. This is to have totally clean and high Q sound from a PC source.
Of course, your software will then be the limiting step to great sound...
I hope you like it,
M.
PS: of course, you can also keep the Toslink Transmitter and wire the Coaxial connector in parallel, to be able to compare between cable formats...
Attachments
Last edited:
Hi,
This is one example:
Hi-Fi CM6631A 192KHZ to Coaxial Optical SPDIF Convertor DAC Board 24bit USB 2.0 | eBay
Greetings,
M.
This is one example:
Hi-Fi CM6631A 192KHZ to Coaxial Optical SPDIF Convertor DAC Board 24bit USB 2.0 | eBay
Greetings,
M.
Hi matthiasw,
I have been optimising the Fractal D/A converter driver boards. PCBs for the next test version have been ordered.
There are plans to offer a bridge version (double output power) for driving less sensitive speakers.
Focus right now is on the development of a novel DAPI (Digital Audio Parallel Interface) receiver for the PowerDAC and headphones PowerDAC.
USB only DAPI receiver, proof of concept version, is already up and running. The USB / Toslink / Coax DAPI receiver is under development.
I have been optimising the Fractal D/A converter driver boards. PCBs for the next test version have been ordered.
There are plans to offer a bridge version (double output power) for driving less sensitive speakers.
Focus right now is on the development of a novel DAPI (Digital Audio Parallel Interface) receiver for the PowerDAC and headphones PowerDAC.
USB only DAPI receiver, proof of concept version, is already up and running. The USB / Toslink / Coax DAPI receiver is under development.