[lag0a]

Hi.

Since digital transmission travel on lines on a PCB board originally then would using a PCB board instead of a coaxial digital cable be better? For example, where the RCA connectors are at the source and DAC, get rid of it and just solder the signal wire and ground wire to the source with a 1 meter PCB board run to your DAC then solder the signal wire and ground wire to the SPDIF receiver in the DAC?

It might not be practical but in light of 75 ohm, cable wire metal, cable geometry, reflections, pulse transformers, attentuators and so on. Would this be better?

[randytsuch]

Impedance control applies to PCB's as well as coax, something I don't see mentioned very often.

I'm not sure, but I kind of doubt that most manufacturers do impendance control on PCB's, for audio.

So, running the signal on a PCB doesn't eliminate the need for impedance control, at least not that I know of.

At work, we have Ethernet and USB signals that require impedance control, so you need to pay special attention to the routing of this signals, and work with the PWB manufacturer to get the impedance you want.

Randy

[lag0a]

I see. I didn't know impedance control applies to PCB boards too. I am just wondering if there is another method of audio transmission using as less parts as possible that is better than what we have today in rca, toslink, aes, hdmi, usb, and so forth. Also the method of audio transmission through silver, gold, copper cables seems flawed.

I am just wondering instead of using connectors and cable to transmit the signal would a pcb board sound better between the source and dac?

[Regalma]

In my business we design instruments for measuring high frequencies. You can get cables that do a good job of handling 40 GHz. You can't do that with any PCB, even with the best and most expensive microwave board material. Boards like that cost many time more than boards you'll find in even the most audio equipment.

If you were to go to ceramic substrates you could. But that is really expensive.

But it is all kind of silly. 40 GHz is orders of magnitude more difficult to work with than anything in the audio world.   

[Speedskater]

Plus analog audio interconnects don't act as well behaved impedance controlled transmission line circuits.
In a well behaved transmission line circuit, the output circuit, the cable and the input circuit all must have the same impedance at the frequency in question.
The typical analog output circuit has a very low impedance.
The typical analog input circuit has a high impedance.
And the analog interconnect characteristic impedance will vary a lot in the audio band.

Digital cables are a whole different ball-game.

[lag0a]

What would be the correct way to align the impedance of the output circuit, cable, and input circuit so it would be an exact match even when it is connected to the input circuit as in the DAC's spdif receiver?

Is it possible to provide an exact impedance match without affecting any audio frequency passing along the circuit?

[audioengr]

Impedance control applies to PCB's as well as coax, something I don't see mentioned very often.

I'm not sure, but I kind of doubt that most manufacturers do impendance control on PCB's, for audio.






So, running the signal on a PCB doesn't eliminate the need for impedance control, at least not that I know of.

At work, we have Ethernet and USB signals that require impedance control, so you need to pay special attention to the routing of this signals, and work with the PWB manufacturer to get the impedance you want.

Randy

We at Empirical Audio certainly use impedance controlled traces on our digital boards, but we a probably the exception.  Never seen anyone else in high end using these.

As for making a cable from FR4 PC board.  Bad idea.  In fact even FR4 is a piss poor medium for digital transmission.  Teflon boards or Polyamide is better.  The dielectric constant (4.7 typ) is too high, absorption is too high.  It is the industry standard FR4, so we are stuck with it.

Steve N.
Empirical Audio

[audioengr]

What would be the correct way to align the impedance of the output circuit, cable, and input circuit so it would be an exact match even when it is connected to the input circuit as in the DAC's spdif receiver?

Is it possible to provide an exact impedance match without affecting any audio frequency passing along the circuit?

It is possible, but you must use BNC and terminate to the board with coax. 
then the trace width on the board is sized according to the distance to the ground plane.  These things are never perfect, but you can get a preetty good match.  I still recommend using a 1.5m cable length to minimize the effect of the reflections.

Steve N.

[Speedskater]

What would be the correct way to align the impedance of the output circuit, cable, and input circuit so it would be an exact match even when it is connected to the input circuit as in the DAC's spdif receiver?

Is it possible to provide an exact impedance match without affecting any audio frequency passing along the circuit?

The first time through, I missed that this is about digital SPDIF signals.
The digital signal band does not change with audio frequency.
The SPDIF system is very robust and can deal with reasonable mis-matches.
But longer digital cables should be matched to the nominal 75 Ohm characteristic impedance.

[HT cOz]

If you all had the option to build all the audio equipment in one box would you do that?

[lag0a]

I have thought of that. My Twisted Pear Dual Mono Opus DAC is on 24/7 and draws less than 1 watt. It sounds better being on 24/7. If you were to include the source, dac, and headphone amp onto one PCB board it would definately draw more watts and it would be a waste of electricity to have it on 24/7 unless you have like many mini off and on switches on each section (source, dac, headphone amp) parts of the PCB board powered separately even though it is on one PCB board.