Well, first, why should these sources be isolated? If they are noise generating, that is something that should be addressed. I am personally a fan of galvanic isolation (pulse transformer) for S/PDIF inputs. If you do that, then there is absolutely no reason to switch grounds.
Aside from that, the job of a ground is to provide the data line with a common reference point. When signals are floating, there could be several volts of difference on the grounds between two components. Once the grounds are connected, this difference becomes 0V. Switching grounds on live sources can lead to spikes (and disruption) when switching. In analog circuits, you get pops. I just don't see the point in it.
Just use a little piece of perf board, two pulse transformers, two 75R resistors for termination. Share the grounds after the transformers, then go to the switch with the signal lines.
Aside from that, the job of a ground is to provide the data line with a common reference point. When signals are floating, there could be several volts of difference on the grounds between two components. Once the grounds are connected, this difference becomes 0V. Switching grounds on live sources can lead to spikes (and disruption) when switching. In analog circuits, you get pops. I just don't see the point in it.
Just use a little piece of perf board, two pulse transformers, two 75R resistors for termination. Share the grounds after the transformers, then go to the switch with the signal lines.
Hi Brian
I used to use pulse transformers on all my inputs, until I bought a CD transport with a pulse transformer on its output. Now a coaxial, screened 75 R impedance cable- conventionally used for S/PDIF connections- can and does pickup hum, and- presumably- any rf fields knocking about. Any ideas for a balanced 75R cable design? I could use whatever suitable balanced connector I liked at the Buffalo end (currently a 75 R BNC), but am stuck with a phono socket out from the CD transport
Thanks
Paul
I used to use pulse transformers on all my inputs, until I bought a CD transport with a pulse transformer on its output. Now a coaxial, screened 75 R impedance cable- conventionally used for S/PDIF connections- can and does pickup hum, and- presumably- any rf fields knocking about. Any ideas for a balanced 75R cable design? I could use whatever suitable balanced connector I liked at the Buffalo end (currently a 75 R BNC), but am stuck with a phono socket out from the CD transport
Thanks
Paul
I tend to use whatever cable I have hanging around. My S/PDIF lines are usually not very long. I have some decent (heavy duty) 1ft Monster cable RCAs I bought in 1988 that I typically use. I also use the crappy RCAs you get free with components. I might be the wrong guy to ask about such things 
Either you have a bad cable, or one of the devices in your house is causing a ground loop. Most likely you have a grounding problem somewhere. Provided the BII is built correctly of course.
While cables are often causing heated debates, my take on it is that you need to create good connections. The connectors on decent cables often provide you with that better connection. I've used excellent connectors both at the BII and on my DIY cables, the resulting fit is simply incredible. The cable I use is selected for easy soldering, and in a diameter that matches well to the connectors used.
P.S. you could try toslink if everything else fails.
Hi
Think people have missed my point here.
At the moment my transport outputs via a pulse transformer. As there isn't a transformer on Buffalo's input, the 75R coax cable I'm using works perfectly. But if I put a pulse transformer on the dac's input too, the cable can pickup hum. Not surprising, as the conductors in the cable- both screen and inner- are now floating and are prone to hum pickup due to he unsymmetrical construction and an ungrounded screen. To avoid this we'd need to use a balanced cable of symmetrical construction- Belden etc make 78R impedance screened twisted pair cable which should work well.
My point about connectors was that, at the CD transport at least, I'm stuck with a coaxial design, rather than one designed for use with a balanced cable. I agree whole heartedly that connector quality is far more important than cable quality in general. Though- and I hesitate to say it here!- I've thought I've head greater differences betwen "digital cables" than "audio" ones
I was just pointing out that "too many" pulse transformers might in fact be detrimental
Paul
Think people have missed my point here.
At the moment my transport outputs via a pulse transformer. As there isn't a transformer on Buffalo's input, the 75R coax cable I'm using works perfectly. But if I put a pulse transformer on the dac's input too, the cable can pickup hum. Not surprising, as the conductors in the cable- both screen and inner- are now floating and are prone to hum pickup due to he unsymmetrical construction and an ungrounded screen. To avoid this we'd need to use a balanced cable of symmetrical construction- Belden etc make 78R impedance screened twisted pair cable which should work well.
My point about connectors was that, at the CD transport at least, I'm stuck with a coaxial design, rather than one designed for use with a balanced cable. I agree whole heartedly that connector quality is far more important than cable quality in general. Though- and I hesitate to say it here!- I've thought I've head greater differences betwen "digital cables" than "audio" ones
I was just pointing out that "too many" pulse transformers might in fact be detrimental
Paul
you're right then, using transformers is a right way to go. I meant just using a switch. your's better
On another matter, for short distante and digital transmissions, you don't need to care TOO much of the cable.
I use Toslink from my computer (from which I play all the music, and movies too) as a way to break ground to the DAC. I managed to hear with headphones some background noise with the USB connection attached. Even if only was selected only the Toslink input to sound on the DAC.
I thought I read somewhere there will be an adapter for fiber optic cables? It has a lot of promise. I hope the physical connectors and protocol don't keep changing every year because the industry will be reluctant to adopt it as a new standard.
Hey Russ
I have heard several people saying that it's best that the SPDIF goes directly into the Sabre chip (hence not using any receiver), the MUX will add a further processing stage, isn't it?
Best
Pietro
Matching the DACs
Dear All,
I've a small problem: I have a buffalo II DAC, configured in dual mono. The problem is that one of the DACs have the output that is slightly higher than the other. I am sure that is the DAC, as I tried to swap them. The difference in 0.3 dBu. This cause the soundstage to be slightly moved towards the left, where the signal is higher.
What can I do ? Shall I just trim it with an output resistor, or can I do something smarter ? I am controlling the DACs with a microprocessor.
Best Regards,
Davide
Dear All,
I've a small problem: I have a buffalo II DAC, configured in dual mono. The problem is that one of the DACs have the output that is slightly higher than the other. I am sure that is the DAC, as I tried to swap them. The difference in 0.3 dBu. This cause the soundstage to be slightly moved towards the left, where the signal is higher.
What can I do ? Shall I just trim it with an output resistor, or can I do something smarter ? I am controlling the DACs with a microprocessor.
Best Regards,
Davide
Of course you could do it via DAC registers. Just set the master trim on the high side DAC to trim out the excess. But you could easily do it in the analog realm too.
Are you using AVCC modules?
If so because you are dual mono and each AVCC module can be very slightly different. This difference should be pretty small. You may just want to put something like a 200K trimmer configured as a rheostat across C2. You could do this on each AVCC or just the one with the higher output. Then you can tweak the output voltages to be "exactly" equal.
BTW most transducers have output differences much larger than that.
Last edited:
One other really easy trick if the DACs are close to each other and share a common ground would be to run a short wire between AVCC modules at the reference voltage side of C2. Now both AVCC modules have the same reference voltage. I am not sure I would ever do it myself, but hey it would work.
I am not sure I would ever do it myself, but hey it would work.
Less tongue in cheek now, honestly that is exactly what the master trim registers are for. You will always have some variation in output channels even in stereo with the same AVCC voltage because of part tolerances and even differences in the DAC itself. The point I was driving home is that its a pretty simple thing to tweak in a number of different ways. It really just depends on what you feel like doing.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.