|
|
Post by MartinT on Oct 26, 2017 11:45:22 GMT
I have tried various digital audio links in the past and am now using a Beresford optical cable between Pi and SEG. It's the best optical cable I've used.
However, now that I'm running the Digi+ Pro card which seeks to reduce jitter to very low levels, I'm interested in whether a good co-ax cable can improve on the inherent jitter in optical cables. So, what recommendations do you have for an inexpensive digital audio cable that sounds good?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 26, 2017 12:01:17 GMT
I used coax for years with the various Beresford DAC's I have had, recently did a comparison with optical and found it to be quite a bit better sounding than coax with the Caiman II into any device.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 26, 2017 12:02:25 GMT
If you want to use a Coax cable, Stan makes one which is very good sounding reasonably priced, in fact, you can borrow mine if you want to, as I am not using it at the moment.
|
|
|
|
Post by MartinT on Oct 26, 2017 12:35:57 GMT
Thanks, Paul. I found a proper RG-59/U 75 Ohm cable in my box of bits that I forgot I'd had. I'm trying it now and am unconvinced that it's any better than the optical. Of course, mode 3 (data recovered clock processing) in the SEG may be obscuring any differences between the two.
|
|
|
|
Post by ChrisB on Oct 26, 2017 13:53:47 GMT
Coax is consistantly better than optical into my Accuphase DAC with every permutation of transport and cable I have tried.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 26, 2017 15:05:11 GMT
Coax is just something that Audio guys have to use  |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 26, 2017 16:20:15 GMT
The age old which digital signal transfer device is best.
Purely depends on the implementation, and as you point out the use of which recovery method, screening and digital ground power is used.
The ideal data transfer medium (imho) would be I2S over ultra short length (under 140mm) mini coax.
The issue I have always found is the conversion from electrical to light and back again at the TX and RX ends this always introduces issues.
I have a couple of digital Tx/Rx boards kicking about that have the Wolfson Micronics WM8805 if you wish to have go at this Martin
|
|
|
|
Post by MartinT on Oct 26, 2017 17:11:09 GMT
Thanks, Tony, but I'm just looking for the best link from my Digi+ Pro ultra low jitter add-on card for the Pi and the Caiman SEG. I can use optical or co-ax (with phonos). So far, the Beresford optical cable has it. It sounds marginally better than any co-ax or other optical I've tried.
|
|
|
|
Post by Clive on Oct 26, 2017 18:09:48 GMT
Martin, it's terrible when people don't answer the question you've asked....so that's what I'm about to do!
I've not tried the Pro but I have the Digi+ I'm now using the Allo DigiOne which has galv isolation, reclocking and a lot of dc power work too. It's quite a revelation, it's coax only (as toslink is too jittery). The step up with the DigiOne was so profound vs the Digi+; my guess is that the step up from the Digi+ to the Pro would not be so significant.
Just my experience...
|
|
|
|
Post by MartinT on Oct 26, 2017 18:29:01 GMT
Hi Clive
Now that is very interesting, I've heard only a little about the Allo DigiOne. My interest is piqued, though. I take it that it's compatible with Volumio? I'll have a look.
What co-ax cable are you using, as a matter of interest?
|
|
|
|
Post by Clive on Oct 26, 2017 18:51:13 GMT
Hi Martin,
Yes the DigiOne is supported by Volumio and Moode. I personally find Volumio a tad easier to use.
As for Coax I'm using a old Trichord coax cable with BNCs, it's always performed well. It uses a rigid coax...I've no idea what the wire is.
|
|
|
|
Post by Clive on Oct 26, 2017 19:25:24 GMT
Oh btw there's a battery UPS available from the UK on ebay. It connects to the GPIO pins, it's supposed to be really good.
|
|
|
|
Post by naim1425 on Oct 26, 2017 20:23:22 GMT
|
|
|
|
Post by MartinT on Oct 27, 2017 20:44:41 GMT
I've now discovered that the jumper on the Digi+ Pro board offers full galvanic isolation for the co-ax connection so I'm going to have to make the comparison again. A job for tomorrow.
|
|
|
|
Post by Stratmangler on Oct 28, 2017 9:21:44 GMT
It uses a rigid coax...I've no idea what the wire is. It sounds like it's an appropriate comms one, and if it's BNC both ends it will have crimped ends. Soldering stuff together affects the impedance. I use one of these markgrant.co.uk/digital-cables/138-belden-1694a-digital-coax-black-phono-to-phono.html#/24-length-1_5_metresThe length is as short as you should be going for coaxial SPDIF installation. Longer isn't a potential problem, but shorter can introduce signal reflections, which in turn can affect the original communications signal (note that I did not say music signal - at this point it doesn't exist). |
|
|
|
Post by MartinT on Oct 28, 2017 10:31:36 GMT
From previous research, it should be a minimum of 1.2m in length. Is that right, Stratmangler ? |
|
|
|
Post by Stratmangler on Oct 28, 2017 10:40:31 GMT
1.5m is generally accepted as the correct length.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 28, 2017 11:53:12 GMT
|
|
|
|
Post by MartinT on Oct 28, 2017 12:09:19 GMT
1.5m is generally accepted as the correct length. Thanks. |
|
|
|
Post by MartinT on Oct 30, 2017 18:09:53 GMT
Ok, performance differential solved. I did wonder why optical sounded so much better than co-ax and now I know. It was my fault as I hadn't appreciated that the SEG saves its Mode setting per input. So I was comparing optical in mode 3 with co-ax in mode 0. Now that I've set the co-ax input to mode 3 all is well and the two sound virtually identical to me. Mode 3 is Stan's most sophisticated setting, called data clock recovery. John when we were listening to your server via the SEG's USB we were listening in mode 0! |
|
