View Full Version : Daft question
Ok guys, this is probably a simple question for you, but I'm trying out new digital cable and it has 'Ground end' marked on it at one end of the cable.
Does this end go at the source (CDP) or at the DAC end? :)
Oh, and what's "Ground" got to do with it? :scratch:
Marco.
Spectral Morn
11-11-2009, 17:56
Ok guys, this is probably a simple question for you, but I'm trying out new digital cable and it has 'Ground end' marked on it at one end of the cable.
Does this end go at the source (CDP) or at the DAC end? :)
Oh, and what's "Ground" got to do with it? :scratch:
Marco.
As a guess I would say DAC. Audio Note cables are grounded at one end (at least the old ones were) and you hook that end up to the Pre or where the signal is flowing to.
Regards D S D L
Hi Neil,
It's a digital cable connected between a CDP and DAC, so if you're saying that the ground end runs from the source, then surely that end should be at the CDP, not the DAC? :scratch:
I don't know why they didn't just mark it 'Source end' instead of 'Ground end' - then there would be no confusion.
Marco.
Spectral Morn
11-11-2009, 18:11
Hi Neil,
It's a digital cable connected between a CDP and DAC, so if you're saying that the ground end runs from the source, then surely that end should be at the CDP, not the DAC? :scratch:
I don't know why they didn't just mark it 'Source end' instead of 'Ground end' - then there would be no confusion.
Marco.
No the signal flows from the transport CDP to---------------the DAC.
Regards D S D L
Yes, usually the ground is at the dac, I think.
Rare Bird
11-11-2009, 18:13
Ok guys, this is probably a simple question for you, but I'm trying out new digital cable and it has 'Ground end' marked on it at one end of the cable.
Does this end go at the source (CDP) or at the DAC end? :)
Oh, and what's "Ground" got to do with it? :scratch:
Marco.
Marco:
it's a sheilded cable with a drain wire running the full length just like the belden mains cable which has a foil with drain wire running down it, the drains wired into the eath at one end usually the mains plug side only..
If its just a "drain", then i'd connect it to the CD player. Try it both ways and see if there's a difference. I suspect there won't be.
Rare Bird
11-11-2009, 18:17
Shocking thing is it's satalite cable no hifi
No the signal flows from the transport CDP to---------------the DAC.
Regards D S D L
LOL - yes, I know that. That's wot I meant when I said:
It's a digital cable connected between a CDP and DAC, so if you're saying that the ground end runs from the source, then surely that end should be at the CDP, not the DAC?
...Thinking that the 'Ground end' should be at the source (CDP) end. To be honest, I'm still none the wiser!
I believe in the effects of cable directionality (as experience tells me that it makes a difference), but normally there are arrows to follow to make life easy...
What I'm really looking for is someone to say categorically "Connect the end marked 'Ground' to the source", or vice versa, and explain the reason for it. But no-one so far has been that definitive.
Marco.
Spectral Morn
11-11-2009, 18:40
Fair point...I can't be that definitive...sorry. Ring VDH and ask, I assume its the VDH cable you bought recently...look on the net.
Regards D S D L
Do DAC's have a ground screw/lug though? I've seen amps with them.
The Vinyl Adventure
11-11-2009, 19:07
shuv one end up your bum and the end marked ground into the soil..
:lolsign:
The Vinyl Adventure
11-11-2009, 19:18
i think that this should acuratly demonstate my thinking behind my theory on this one!
http://i728.photobucket.com/albums/ww282/hamish_gill/leccynokillmarco.jpg
What I'm really looking for is someone to say categorically "Connect the end marked 'Ground' to the source", or vice versa, and explain the reason for it. But no-one so far has been that definitive.
I think it's precisely because there is no definitive answer that it says ground end rather than source end. try it both ways; if one way sounds better, that's the right way!
i think that this should acuratly demonstate my thinking behind my theory on this one!
http://i728.photobucket.com/albums/ww282/hamish_gill/leccynokillmarco.jpg
Hamish,
Yer aff yer heid, ya daftie, but I lurves it! :lol: :lolsign:
Shane,
Thanks - I see what you mean, and that makes sense :)
Neil,
Yes it's this recently bought VDH cable here, the 'Digi-coupler':
http://cgi.ebay.nl/NEW-VAN-DEN-HUL-DIGICOUPLER-SPDIF-DIGITAL-INTERLINK_W0QQitemZ370259346430QQihZ024QQcategoryZ 8322QQcmdZViewItemQQ_trksidZp4340.m444QQ_trkparmsZ algo%3DCRX%26its%3DC%252BS%26itu%3DSI%252BUA%252BL M%252BLA%26otn%3D10%26ps%3D63
...and bloody brilliant it is, too - probably the best digital coaxial cable I've heard (yes, better than my current Mark Grant 1694A and anything else much more expensive that I've used before it).
Frankly, if it turns out to be as good as it seems now (sometimes first impressions are misleading), it's an absolute bargain in 'high-end' terms.
Full review to come once I've had a proper listen.
Marco.
Spectral Morn
11-11-2009, 19:55
i think that this should acuratly demonstate my thinking behind my theory on this one!
http://i728.photobucket.com/albums/ww282/hamish_gill/leccynokillmarco.jpg
Brilliant Hamish absolutely brilliant :lol::lol::lol::lol::lol::lol::lol::lol::lolsign:
You should do more of these from time to time when the topic allows your imagination to rise to the occasion.
Regards D S D L
Neil,
Yes it's this recently bought VDH cable here, the 'Digi-coupler':
http://cgi.ebay.nl/NEW-VAN-DEN-HUL-DIGICOUPLER-SPDIF-DIGITAL-INTERLINK_W0QQitemZ370259346430QQihZ024QQcategoryZ 8322QQcmdZViewItemQQ_trksidZp4340.m444QQ_trkparmsZ algo%3DCRX%26its%3DC%252BS%26itu%3DSI%252BUA%252BL M%252BLA%26otn%3D10%26ps%3D63
It's written that :
About ground end stickers:
Each van den Hul cable is marked at one side where the screen is connected to the ground. This way no signal transfer occurs through the shield. The shield functions as a screen only...Be sure to connect this side of the cable to the signal source (e.g. digital drive or disc/CD/DVD player)
Rare Bird
11-11-2009, 20:22
Just plug it in, it's all bollerks anyway
The Vinyl Adventure
11-11-2009, 20:25
Brilliant Hamish absolutely brilliant :lol::lol::lol::lol::lol::lol::lol::lol::lolsign:
You should do more of these from time to time when the topic allows your imagination to rise to the occasion.
Regards D S D L
Neil, for you i shall try! ... im not sure why, but i have this strange urge to pick on marco ?? ... hmm ... im not promising anything but i will keep an eye out for future subjects for my artistic "skills"
Spectral Morn
11-11-2009, 20:31
Neil, for you i shall try! ... im not sure why, but i have this strange urge to pick on marco ?? ... hmm ... im not promising anything but i will keep an eye out for future subjects for my artistic "skills"
As Bill and Ted would say "Excellent"...."Party on Dude":)
Regards D S D L
It's written that :
About ground end stickers:
Each van den Hul cable is marked at one side where the screen is connected to the ground. This way no signal transfer occurs through the shield. The shield functions as a screen only...Be sure to connect this side of the cable to the signal source (e.g. digital drive or disc/CD/DVD player)
Ah, now that's exactly what I needed to know! Cheers, Dimitri :)
The problem is, there's too much bloody stuff written on the Ebay advert, so I missed that, and there's no such information supplied with the cable itself.
Marco.
Rare Bird
11-11-2009, 23:29
can i ask what make this cable is please
Here it is, Andre: http://cgi.ebay.nl/NEW-VAN-DEN-HUL-DIGICOUPLER-SPDIF-DIGITAL-INTERLINK_W0QQitemZ370259346430QQihZ024QQcategoryZ 8322QQcmdZViewItemQQ_trksidZp4340.m444QQ_trkparmsZ algo%3DCRX%26its%3DC%252BS%26itu%3DSI%252BUA%252BL M%252BLA%26otn%3D10%26ps%3D63
I've been listening to music through it for most of the evening, and it's a very quiet, clear, detailed sounding cable, with an extremely natural way of reproducing voices and instruments.
It's not ultra-cheap, granted, but it's fairly affordable and way, way better than some digital cables I've heard at ten times its price!
What's annoying is that it doesn't seem to be available in the UK - I had to import mine from Belgium.
What's even more annoying is that VDH's products are so inconsistent, because some of the interconnects I've heard are crap (overly smooth and lush sounding), but this Digicoupler, and 'The Wind' speaker cables I already use, are truly excellent. You have to get to know the quirks of their product range.
Marco.
Here it is, Andre: http://cgi.ebay.nl/NEW-VAN-DEN-HUL-DIGICOUPLER-SPDIF-DIGITAL-INTERLINK_W0QQitemZ370259346430QQihZ024QQcategoryZ 8322QQcmdZViewItemQQ_trksidZp4340.m444QQ_trkparmsZ algo%3DCRX%26its%3DC%252BS%26itu%3DSI%252BUA%252BL M%252BLA%26otn%3D10%26ps%3D63
I've been listening to music through it for most of the evening, and it's a very quiet, clear, detailed sounding cable, with an extremely natural way of reproducing voices and instruments.
It's not ultra-cheap, granted, but it's fairly affordable and way, way better than some digital cables I've heard at ten times its price!
What's annoying is that it doesn't seem to be available in the UK - I had to import mine from Belgium.
What's even more annoying is that VDH's products are so inconsistent, because some of the interconnects I've heard are crap (overly smooth and lush sounding), but this Digicoupler, and 'The Wind' speaker cables I already use, are truly excellent. You have to get to know the quirks of product range.
Marco.
That works out to £1 per centimetre! Hope it's worth it.
Have you found out which way round to connect it? Can you hear a difference?
Regards
I'd say that it was well worth it, Barry :)
Yes, I've tried the cable both ways, and music reproduced definitely sounds better (more cohesive and natural) with the 'ground end' connected at the source.
I had a feeling that this would be an excellent digital cable, and I was right. It synergises extremely well with the Mark Grant G1000HD interconnects, and it's quite a bit better than Mark's own (already superb) Belden-based 1694A digital coaxial cable.
Marco.
That's fine Marco. Just thought I'd chip in with my comment as I think you know my views about expensive cables!
By the way, what do you think about my idea of members of JJs cable listening session listening to the four cables again in their own systems? (Assuming JJ is agreeable to this.)
Regards
Yup, I'm well up for that, Barry :)
Marco.
Rare Bird
12-11-2009, 00:56
Marco you probably don't wanna hear this but there's a manufacturer that sells satellite cable of exactly the same construction with the same drain wire, it's cheap if i remember, we have some in our satellite system grounded inside.Exactly the same thing..
That's cool, matey. Maybe you could get hold of some more and make me up a cable to try against the VDH? :)
Marco.
anthonyTD
12-11-2009, 10:27
yep,
ground should be terminated to first piece of source equipment, simples! [i cant do the squeek]:lolsign:
A...
Cheers big ears! Steve says he'll make you squeak later :eyebrows:
Marco.
anthonyTD
12-11-2009, 10:43
:lol::lol::lol::lol::lol::lol::lol::lol::lol::lol: :lol::lol::lol::lol::lol::lol::lol::lolsign:
A...
Steve says he'll make you squeak later :eyebrows:
Marco.
Squeak when you're squoken to!
yep,
ground should be terminated to first piece of source equipment, simples! [i cant do the squeek]:lolsign:
A...
That would indeed be the conventional way of doing it. However, not always! ;)
My own CDP, for example, is double insulated and has one of those figure 8 type power cords. Connecting the 'ground' end to the source in this instance means the only path back to earth is via the signal return and to whatever earth is available at the 'sink' end!
In this instance I've alway found it better to reverse the connection.
If both items are double insulated, then God knows!... You may well get better results with an unscreened cable!
Cheers..
anthonyTD
12-11-2009, 17:26
That would indeed be the conventional way of doing it. However, not always! ;)
My own CDP, for example, is double insulated and has one of those figure 8 type power cords. Connecting the 'ground' end to the source in this instance means the only path back to earth is via the signal return and to whatever earth is available at the 'sink' end!
In this instance I've alway found it better to reverse the connection.
If both items are double insulated, then God knows!... You may well get better results with an unscreened cable!
Cheers..
hi mike,
in this instance what you say makes sense as you have no alternative!:)
A...
Marco - have you tried one of Mike's specials? I can't remember, and that thread was sooooooo big!
Nope, the bugger said he would send me one but oi never gots it... :(
He's gettin' on a bit now so sometimes forgets these things! :eyebrows:
Mikey,
Good points! All you can really do I guess is listen to the cable both ways and pick the way which sounds the best (if you can hear a difference, that is).
I've tried that and it definitely sounds better with the 'ground end' at the source (CDP).
Marco.
I've tried that and it definitely sounds better with the 'ground end' at the source (CDP).
It should do... yours as a proper erf at der sauce end, like. :)
I'll sort out a cable for you to try if/when I eventually receive some more WBT's... this postal strike lark has been wreaking havoc. I've got 'back orders' to fill first though! :(
btw, if someone likes to test digital coaxes, I have a U-Byte (http://www.analogresearch-technology.net/ubyte.html) which is a very interesting cable. Can be easily sent. ;)
Ian Walker
12-11-2009, 22:41
:lol::lol::lol::lol::lol::lol::lol::lol::lol::lol: :lol::lol::lol::lol::lol::lol::lol::lolsign:
A...
You reminisin there Glenda:lol:
Rare Bird
13-11-2009, 00:40
That's cool, matey. Maybe you could get hold of some more and make me up a cable to try against the VDH? :)
Marco.
Why bother won't sound any different!! :lol:
I suspect it would, as I've yet to hear two cables that sound identical. I've no idea which one would sound best, though!
Marco.
Steve Toy
16-11-2009, 22:05
Ok guys, this is probably a simple question for you, but I'm trying out new digital cable and it has 'Ground end' marked on it at one end of the cable.
Does this end go at the source (CDP) or at the DAC end? :)
Oh, and what's "Ground" got to do with it? :scratch:
Most bizzarre. Same results at mine. It says run ground end nearer the source. Critical listening here says do the opposite. Do as per instructions and voices sound disembodied and basslines play out of sync with the rest of the mix. Do the opposite and everything snaps into temporal focus.
It's all in the mind. :D
I've just paid for my cable - £92
Ian Walker
16-11-2009, 22:07
I've just paid for my cable - £92
You are defo this weeks Double Daftee
Most bizzarre. Same results at mine. It says run ground end nearer the source. Critical listening here says do the opposite. Do as per instructions and voices sound disembodied and basslines play out of sync with the rest of the mix. Do the opposite and everything snaps into temporal focus.
Indeed. That's exactly what we heard, and it was the same in my system, too! :)
When people suggested that the end marked 'ground' should go to the source (as makes sense) I initially thought that it sounded better this way, but somehow I just wasn't getting into the music. Familiar albums I know well and love were rendered as 'uninteresting'.
Then I thought 'bugger it', tried it the other way, and the sound was instantly better in the way Steve describes. I'd also add that it seemed to make my system sound 'out of phase' with the ground end connected to the source.
Perhaps VDH marked the wrong end of the cable or the 'ground' in my system and Steve's is not at the source end...
No matter! What we both agreed on though is that the VDH is fundamentally better than the Mark Grant 1694A digital cable, which is already excellent. I'll go into more detail about this when I do a separate review later. What I will say though is that I reckon it's probably one of the best digital cables available at any kind of an affordable price.
It would certainly be interesting hearing Mike's cable against it :cool:
Marco.
I've just paid for my cable - £92
You are defo this weeks Double Daftee
Nope - there's actually quite a significant difference between both cables, muchacho :)
Can't test it at yours though because you're an 'optical boy'...
Have you got yer Mark Grant interconnects yet? :smoking:
Marco.
Guys you need to listen to a 6m 75ohm cable one day... ;)
6M? That's nearly three times the length of my room!! :lol:
The Digicoupler is 75ohm, and strictly so ;)
Marco.
6M? That's nearly three times the length of my room!! :lol:
The Digicoupler is 75ohm, and strictly so ;)
Marco.You have to know that, it doesn't exist an RCA cable "strictly" 75ohm. It is reserved for BNC connexions only. The best cables are at 75ohms +-0.5ohm. All cable manufacturers know that, whatever they may advertise.
The 6m is a way of adressing signal reflexions due to impedance differences : http://www.positive-feedback.com/Issue14/spdif.htm
The number of reflexions depend on the length of the cable and the impedance difference. As we cannot minimize the difference (because we don't know the impedance error of the S/PDIF interfaces on each side), we increase the cable length so that the reflexions don't mess up with the signal and can be discarded by the dac's circuits.
You have to know that, it doesn't exist an RCA cable "strictly" 75ohm. It is reserved for BNC connexions only.
Ah, that's interesting... Still, I'd never use a 6m cable - it's just too impractical ;)
Besides, I'm very happy with the results I'm getting from the VDH :)
Marco.
Still, I'd never use a 6m cable - it's just too impractical ;)True. When you have "open" shelves, it's hard to hide...
Anyway. Most modern dacs treat well transmission jitter now. After all, jitter is probably an over-rated concept. :)
Ian Walker
16-11-2009, 23:18
Can't test it at yours though because you're an 'optical boy'...
Has Steve even tried an optical cable with his new dac?,i bet it would be just as good.
Rare Bird
16-11-2009, 23:24
Are we not getting a tad carried away with Digital cables? On simular lines, i got myself a Kimber USB lead, totally no difference in performance from a cheap printer USB, just looks more pretty
Has Steve even tried an optical cable with his new dac?,i bet it would be just as good.
He would've done if he'd had one... ;)
I'm not saying necessarily that coaxial is better than optical (or vice versa), but rather that the VDH coaxial cable is considerably better than the Mark Grant coaxial cable :)
Dimitri,
What I find strange is that if BNC is so crucial then why don't more DACs and CDPs have BNC digital outputs, instead of phonos? :scratch:
Steve's Bel Canto CDP has a BNC connector for the digital out, but in my experience, it's far from being common. Phonos are what I see mostly.
Marco.
Dimitri,
What I find strange is that if BNC is so crucial then why don't more DACs and CDPs have BNC digital outputs, instead of phonos? :scratch:
Steve's Bel Canto CDP has a BNC connector for the digital out, but in my experience, it's far from being common. Phonos are what I see mostly.
Marco.
Sheer convenience. Actually often well made dacs have a BNC, but very few CDPs do have indeed. Oh well, perhaps it has to do with the BNC cheapo-for-satellite image.
Look, pros at studios prefer AES/EBU XLRs although the XLR connector and cables are NOT suitable (not designed) for digital audio... and unable to cope correctly with the frequencies and the bandwidth involved, so what to expect from hifi manufacturers... ?
To me, the overall impression is that digital audio is not as carefully made as analog. :( Designers are somehow... lazy and not "perfectionists".
Steve Toy
17-11-2009, 01:47
Are we not getting a tad carried away with Digital cables? On simular lines, i got myself a Kimber USB lead, totally no difference in performance from a cheap printer USB, just looks more pretty
Andre, your system setup is clearly wanting. We hear what we hear because our systems are properly setup to reveal all kinds of temporal subtleties.
You should try comparing RCA against proper 75R BNC plug and sockets for your s/pdif connections sometime, you may be surprised;)
Hi Leo,
I don't doubt that BNCs are better in the application we're discussing, although results will likely vary from CDP to CDP, or DAC to DAC, depending on how they deal with jitter, but could you (or someone else) explain the logic behind it in layman terms?
Excuse my ignorance, but why is a BNC plug 'more accurately 75 Ohms' than a phono - what makes it so?
The other thing that puzzles me (although I see what Dimitri is saying in terms of convenience, etc), is that my Sony CDP/DAC was a top-of-the-range, 'cost-no-object 'statement' piece of engineering, by Sony in an era when CD was in its heyday (late 80s/early 90s)...
Therefore, surely if BNC sockets were *so* crucial for the accurate transfer of digital data, Sony would've known that and supplied the appropriate connections on their 'statement' players? I can't see the convenience thing applying here, as they've 'maxed out' (in a big way) everything else about the design...
Or is BNC perhaps a later, more modern, concept? :scratch:
Lastly, how easy/difficult would it be to remove existing phonos and fit BNCs - is it simply just a case of replacing one with the other?
Marco.
Hi Leo,
I don't doubt that BNCs are better in the application we're discussing, although results will likely vary from CDP to CDP, or DAC to DAC, depending on how they deal with jitter, but could you (or someone else) explain the logic behind it in layman terms?
Excuse my ignorance, but why is a BNC plug 'more accurately 75 Ohms' than a phono - what makes it so?
The other thing that puzzles me (although I see what Dimitri is saying in terms of convenience, etc), is that my Sony CDP/DAC was a top-of-the-range, 'cost-no-object 'statement' piece of engineering, by Sony in an era when CD was in its heyday (late 80s)...
Therefore, surely if BNC were *so* crucial for the accurate transfer of digital data, Sony would've known that and supplied the appropriate connections on their 'statement' players? I can't see the convenience thing applying here, as they've 'maxed out' everything else about the design...
Or is BNC perhaps a later, more modern, concept? :scratch:
Lastly, how easy/difficult would it be to remove existing phonos and fit BNCs - is it simply just a case of replacing one with the other?
Marco.
I struggled to find an article about impedance (and BNC versus RCA) that's about the best I could find : http://www.bluejeanscable.com/articles/impedance.htm and also http://www.bluejeanscable.com/articles/75ohmrca.htm
In short, whet they say is :
Maintaining 75 ohm impedance on mini-DIN and HD-15 plugs is a lost cause; but is there such a thing as a true 75 ohm RCA plug? Not really; Canare's RCAP-series plugs, which we feel are the best RCA plugs available for video, are often referred to as "true 75 ohm" plugs, but that's not quite accurate. At the same time, a look at the construction of these plugs shows that they are easily the best plug on the market for a good impedance match with 75 ohm cable.
For your Sony, the explanation I can see is :
The jitter, although well known as a phenomenon in real-time clock applications since the '60s, it was almost not addressed at all when redbook arrived. The industry was so "obsessed" about convincing and migrating the recording studios that any criticism about the imperfections was seen like a kind-of-subjectivist nonsense.
Once most of the studios had invested (at about 90-95) we started publishing papers about sound issues (including jitter issues), and -finally- we ended up addressing most of the problems of the A/D/A process using re-sampling and up-sampling. I can provide some of the papers involved, if you wish.
Transmission-line effects are a well-known issue in HF signals, but they are only a part of the problem (see complete list of jitter issues here: http://www.positive-feedback.com/Issue43/jitter.htm). I'm not a positive-feedback.com fan, but Nugent provides an accurate list, that's why I provide the link.
So, it's not that the Sony engineers were "lazy" : it's that they didn't know how to address some issues in the late '80s... ;)
Lastly, having a "perfect" 75ohm cable (and plugs) is only half-a-solution : you need to measure the impedance of your input and output transmitters (which cannot be done easily) and correct any deviation from the cable's one. So, simply changing the plug is not enough sometimes.
Rare Bird
17-11-2009, 09:34
Andre, your system setup is clearly wanting. We hear what we hear because our systems are properly setup to reveal all kinds of temporal subtleties.
No i's not wanting i'm just not obsessed with meaningless things..;)
Dave Cawley
17-11-2009, 09:42
Excuse my ignorance, but why is a BNC plug 'more accurately 75 Ohms' than a phono - what makes it so? Because it is, simples! But simply putting in a BNC connector is likely to achieve nothing.
All high end stuff uses BNC, why not? But jitter is more fundamental to the electronics in use, not the connector and cable.
www.soundhifi.com/test.html shows a jitter meter of the type used in factories to test CD Players before despatch. However there are much better ways and Noel Keywood of Hi Fi World often tests in great detail. For more to read on jitter go to www.soundhifi.com/grimm.html and download the White Paper?
Regards
Dave
No i's not wanting i'm just not obsessed with meaningless things..;)Well, sound accuracy is just an accumulation of individually meaningless things, anyway... ;)
Thanks for the info, Dimitri - much appreciated :)
That certainly explains quite a lot.
Regarding the Sony's lack of BNC sockets, I suspect that what you've given is a very good explanation. However, it could also be that both the CDP and DAC are so well-designed and over-engineered (check out the photos below) that such things were of minor significance...
Sony X-777ES here: http://www.thevintageknob.org/SONY/sonyes/CDPX777ES/CDPX777ES.html#
And here:
http://www.thevintageknob.org/SONY/sonyes/CDPX777ES/CDPX777ES-more.html
And here (this is the finish I have my DAS-R1 in):
http://www.thevintageknob.org/SONY/sonyes/CDPX777ES/CDPX777ES-gold.html
DAS-R1 DAC: http://www.thevintageknob.org/SONY/sonyesprit/CDPDASR1/CDPR1DASR1.html
I guess the only way to find out is to have some BNCs fitted and test out the theory! ;)
Thing is though, the sound I'm getting at the moment with the VDH cable is so stunningly good that it's difficult to imagine how things could get much better :)
What's interesting about the DAS-R1 (if you read the article) is that at the time Sony favoured the 'Twin Link' connection, over anything else, and listening tests (when I had the matching R1 transport before it 'broke') seemed to suggest this.
From the photographs you can see just how serious a piece of kit this is and how it puts many high-end modern CDPs to shame, making them look (and sound) like mere toys in comparison! :eyebrows:
Marco.
Rare Bird
17-11-2009, 09:58
Well, sound accuracy is just an accumulation of individually meaningless things, anyway... ;)
Dimitri
There's too much bullshit going on here for my liking..Sound accuracy! This is the trouble with stereo's these days they don't sound right..
Conversing about analogue cables is one thing but i'm sorry i don't buy all this digital cables thing.On the same note, it won't make a blind bit of difference between BNC & a phono.
I guess the only way to find out is to have some BNCs fitted and test out the theory! ;)
No, I think that the simplest is to plug a 6m coax cable and test whether it maces a difference. If it makes a difference, then your system is prone to transmission jitter and you can see whether you address it and how.
If not : well, as you say it's well engineered and there's no point in doing anything at all. ;)
Hi Dave,
Because it is, simples! But simply putting in a BNC connector is likely to achieve nothing.
I see. What if a lack of BNC socket was the only thing 'holding back' an otherwise 'well-sorted digital signal', as it were?
If what you're saying is that what happens before and after the signal reaches the BNC socket is more important than the socket itself, then perhaps me fitting them to the Sony would achieve nothing? :)
Marco.
Conversing about analogue cables is one thing but i'm sorry i don't buy all this digital cables thing.On the same note, it won't make a blind bit of difference between BNC & a phono.
I have a different experience : I have found digital equipment more prone to digital cable swap than to analogue cable swap. ;)
Things are getting better though as time passes : let's not forget that CD is 30 years old. Clocks, D/A converters, S/PDIF interfaces, SRCs have all evolved. And they are cheaper.
Hi Andre,
i'm sorry i don't buy all this digital cables thing.On the same note, it won't make a blind bit of difference between BNC & a phono.
That's fine - don't buy it. However, please leave others who can clearly hear a difference to their valid observations. Cheers!
Dimitri,
I take your point about a 6m cable, but I'm afraid I won't be doing that :eyebrows:
What I'd like to test is what difference (if any) would fitting BNCs both to my VDH digital cable and Sony CDP and DAC make compared to using the exisitng phonos?
Or should I just stop arsing around with this pish and continue enjoying the music? :lol:
Marco.
Things are getting better though as time passes : let's not forget that CD is 30 years old. Clocks, D/A converters, S/PDIF interfaces, SRCs have all evolved. And they are cheaper.
Cheaper, yes, but in comparison with what I'm using, experience tells me in most cases they're most certainly not better.
Transport quality has definitely deteriorated, especially with the advent of cheap OEM plastic DVD-ROM mechanisms, and in my opinion, in terms of DACs, nothing beats a properly-implemented TDA 1541-based design! ;)
Ian's due to bring Duncan's latest D.I.Y DAC incarnation round for a listen, which is very good indeed, but I've yet to hear any DAC which I'd use instead of my Audiocom-modified Sony, not even any of Stan's excellent designs, as good as they are in their own context.
Marco.
Dave Cawley
17-11-2009, 10:15
Hi Marco
I see. What if a lack of BNC socket was the only thing 'holding back' an otherwise 'well-sorted digital signal', as it were? It won't be, trust me!
If what you're saying is that what happens before and after the signal reaches the BNC socket is more important than the socket itself, then perhaps me fitting them to the Sony would achieve nothing? That is what I said.
Gratuitously changing sockets is likely to cause more harm than good. To fit a 75ohm BNC properly would require that the termination on the socket and board, and the cable in-between was also perfectly 75 ohm, and most would find this difficult to achieve. But even if one could, the difference would not be noticeable, unless one made it worse!!
Dave
Replacing the PCB mounted digital coax connector with a BNC, allowed me to hear a worthwhile and noticeable difference in my DVD player. Of course i was then using a different cable, albeit a very cheap one.
Could have been the cable or the socket, or both.
Spending an extra five pounds on a couple of metres of better budget cable, improved things again.
Any improvement is welcome, doesn't matter where it comes from.
Hi Dave,
Gratuitously changing sockets is likely to cause more harm than good. To fit a 75ohm BNC properly would require that the termination on the socket and board, and the cable in-between was also perfectly 75 ohm, and most would find this difficult to achieve. But even if one could, the difference would not be noticeable, unless one made it worse!!
I take your point, and will probably leave things alone. However, it's interesting how your observations differ from Al's (above), who incidentally is someone else whose opinion I rate highly... :)
I guess that I should 'suck it and see', but I'm not sure I can be arsed.
Marco.
Steve Toy
17-11-2009, 11:19
Andre, that being the case you'd best not pollute these threads :p
Hi Leo,
I don't doubt that BNCs are better in the application we're discussing, although results will likely vary from CDP to CDP, or DAC to DAC, depending on how they deal with jitter, but could you (or someone else) explain the logic behind it in layman terms?
Excuse my ignorance, but why is a BNC plug 'more accurately 75 Ohms' than a phono - what makes it so?
The other thing that puzzles me (although I see what Dimitri is saying in terms of convenience, etc), is that my Sony CDP/DAC was a top-of-the-range, 'cost-no-object 'statement' piece of engineering, by Sony in an era when CD was in its heyday (late 80s/early 90s)...
Therefore, surely if BNC sockets were *so* crucial for the accurate transfer of digital data, Sony would've known that and supplied the appropriate connections on their 'statement' players? I can't see the convenience thing applying here, as they've 'maxed out' (in a big way) everything else about the design...
Or is BNC perhaps a later, more modern, concept? :scratch:
Lastly, how easy/difficult would it be to remove existing phonos and fit BNCs - is it simply just a case of replacing one with the other?
Marco.
To answer your first point Marco, the characteristic impedance of a coaxial connector or of a coaxial cable is determined by its geometry and by the dielectric constant of the insulation used. BNC connectors are a precision connector and are designed to have a specific impedance. For more details have a look at the article on audio connectors in the Library. Without going into the maths, RCA 'phono' connectors cannot have an impedance of 75 Ohm, they actually have an impedance ~ 41 Ohms; don't believe what manufacturers will tell you otherwise.
To answer your second point, the digital output on virtually every CD transport uses what is known as S/PDIF, which stands for "Sony / Philips Digital InterFace"; an arrangement jointly agreed by Sony and Philips. The specification for this interface is to use a characteristic impedance of 75 Ohms and RCA connectors. That is why Sony players are fitted with a 'phono' digital output. Does it matter that 'phono's cannot be 75 Ohms? Probably not as the 'phono' plug/socket combination is electrically 'short', so the mismatch is small.
Thus to answer your last point - changing the connector would be relatively straightforward, but unlikely to make much improvement.
However I would like to qualify my last comment with the caveat that the sampling waveforms used in some players have very fast rise and fall times - which means that one is effectively dealing with frequencies in the MHz region. At these frequencies the cable is starting to behave as a transmission line and the maintenance of the correct characteristic impedance becomes important.
If I have sent you to sleep I apologise - but you did ask!
Regards
Probably only worth trying by the diyers , those buying ready made cables and having to pay somebody to do the alterations for them is probably best off not bothering ;)
Its not a huge difference but for me I found things did improve slightly, enough so that I fit them on all dacs I now have (yes I was sad enough to compare direct) , even on the Caiman , this has two co-axial inputs with RCA sockets, I modified one for 75R BNC, its easy to compare with that dac although as already mentioned on the thread its not so straightforward changing on the equipment.
Main difference I hear is in the extreme high's which are a little more clean and clear, not enough to blow you off your seat.
like with most things you either hear an improvement or you don't, simple as that.
BTW if you try it make sure the BNC's are 75R and not the more common 50R, some people get them mixed up and can't tell the difference physically
Also don't bother with those BNC converter plugs either
Hi Dave,
I take your point, and will probably leave things alone. However, it's interesting how your observations differ from Al's (above), who incidentally is someone else whose opinion I rate highly... :)
I guess that I should 'suck it and see', but I'm not sure I can be arsed.
Marco.
In your case I wouldn't bother , it'll probably cost too much to make it worthwhile tbh, for the diyers who are bored one sunday its worth a go.
Hi Leo (and others),
I'll comment later on this thread, but with regard to the above, not when you've got friends into D.I.Y hi-fi who will gladly wield a soldering iron on your behalf and fit the bits you've bought wherever they need to go ;)
How much are (high quality) 75R BNC plugs and sockets of which you speak - do you have any links to where these are available?
Marco.
There was something elsewhere around here that goes into 'transmission lines' and 'characteristic impedance' that was posted ages ago.
If anyone's interested enough you might be able to find it lurking around, I can't remember where it is though... :scratch:
Hi Leo (and others),
I'll comment later on this thread, but with regard to the above, not when you've got friends into D.I.Y hi-fi who will gladly wield a soldering iron on your behalf and fit the bits you've bought wherever they need to go ;)
How much are (high quality) 75R BNC plugs and sockets of which you speak - do you have any links to where these are available?
Marco.
Their not expensive or anything fancy say a few quid each,their available from places like Farnell and RS, theres various shapes and sizes of plugs and sockets to suit what their being fitted to, it should be stated on description what impedance they are, as long as its 75R
How much are (high quality) 75R BNC plugs and sockets of which you speak - do you have any links to where these are available?
They're available from me, for 'shitpence'... it's the tools that cost!
Oh right... Noted ;)
Cheers!
I'll give you a shout if I decide to put my CDP, DAC and cables through the necessary 'ordeal' :eyebrows:
Marco.
Brilliant!...
Oh the thought of putting a drill through your DAC and CDP! :gig:
Primalsea
17-11-2009, 23:02
On the subject of what ground to connect .. Well not on it any more but I've come in late again.
Is anyone up for an experiment??
Connect the CDP & DAC what ever way you want then get 2 sturdy leads the same length and connect these from the signal grounds to a mains earth point. This can be a mains earth pin on a mains plug or a earth binding point on the back of a piece of equipment. The signal ground can be connected by croc clips to the interconnect grounds.
The idea is that its not really about cable direction per see but rather both pieces of equipment have their grounds and exactly the same voltage. Rather than using the interconnect signal ground to achieve this you use the other wires.
Is anyone interested in trying this, I've found that it seems to lower the noise floor. I'm wondering if anyone else might here a difference.
On modifying phono's to BNC's I have a question, more of an idea really.
If the phono's did actually cause a deviation form 75 ohm could you not just change the resistors in the SPDIF transmitter & receiver circuits to correct for it. It would be less invasive than changing sockets.
AKA... 'equipotential bonding'? :)
Hi Paul!... I still want to drill holes in Marco's gear! ;)
I thought it'd be more ma 'heed' than ma gear! :lolsign:
Marco.
I thought it'd be more ma 'heed' than ma gear! :lolsign:
Marco.
You've more than enough holes in yer heed! :ner:
S'pecially that one that keeps flappin! :lol:
The difference is though your 'heed' is full of fresh air! :D
Marco.
The difference is though your 'heed' is full of fresh air! :D
Marco.
Fresh as a daisy! :):)
Hi Mike,
There was something elsewhere around here that goes into 'transmission lines' and 'characteristic impedance' that was posted ages ago.
If anyone's interested enough you might be able to find it lurking around, I can't remember where it is though...
That would be me:
http://theartofsound.net/forum/showthread.php?t=3994&page=7&highlight=75Ohm ,
post 44 onwards.
Re-doing the calculation for effective frequencies of 5.74MHz, effect of an impedance mismatched ‘phono’ connector produces reflections which will be -54dB relative to the intended signal.
In principle true 75Ω connectors, such as a BNC, would be preferred, however this assumes that the impedance of the circuit behind the connector is also 75Ω. Maybe this is why the S/PDIF specification calls for ordinary phono connectors.
I can well imagine that some improvement might be made in changing connectors on some CD transports but not others.
Regards
Primalsea
21-11-2009, 08:54
Is no one interested in the experiment??
Gratuitously changing sockets is likely to cause more harm than good. To fit a 75ohm BNC properly would require that the termination on the socket and board, and the cable in-between was also perfectly 75 ohm, and most would find this difficult to achieve. But even if one could, the difference would not be noticeable, unless one made it worse!!
Dave
Maybe yes, maybe. It is true that if you are not experienced at this sort of work that you could in fact make things work. Over on Audio Circle, in the Lab sub-forum, there are a few threads where I demonstrate the right way and wrong way to to this. And measurements to back it up. I don't know the URL, but if you use their search feature (using the name "art"), you should be able to find them. Mind you, they are 2-3 yeas old.
So...............if you know what you are doing, it is possible to make a difference, and one that is an improvement. Although one of our dealers (in Chicago, back in the early 90s) thought that he could do better than us. He complained that it now sounded worse, and expected us to fix it for free. After he buggered it. With friends like him, who needs enemies!
Cheers,
Pat
To answer your first point Marco, the characteristic impedance of a coaxial connector or of a coaxial cable is determined by its geometry and by the dielectric constant of the insulation used. BNC connectors are a precision connector and are designed to have a specific impedance. For more details have a look at the article on audio connectors in the Library. Without going into the maths, RCA 'phono' connectors cannot have an impedance of 75 Ohm, they actually have an impedance ~ 41 Ohms; don't believe what manufacturers will tell you otherwise.
Yes.
To answer your second point, the digital output on virtually every CD transport uses what is known as S/PDIF, which stands for "Sony / Philips Digital InterFace"; an arrangement jointly agreed by Sony and Philips. The specification for this interface is to use a characteristic impedance of 75 Ohms and RCA connectors. That is why Sony players are fitted with a 'phono' digital output. Does it matter that 'phono's cannot be 75 Ohms? Probably not as the 'phono' plug/socket combination is electrically 'short', so the mismatch is small.
Thus to answer your last point - changing the connector would be relatively straightforward, but unlikely to make much improvement.
However I would like to qualify my last comment with the caveat that the sampling waveforms used in some players have very fast rise and fall times - which means that one is effectively dealing with frequencies in the MHz region. At these frequencies the cable is starting to behave as a transmission line and the maintenance of the correct characteristic impedance becomes important.
If I have sent you to sleep I apologise - but you did ask!
Regards
Ah...........a caveat.............
Well, at the frequencies that the parts of the waveform that are responsible for nice, sharp, clean rising and falling edges, it makes a big difference.
Here are some measurements that show how much of a difference using RCAs instead of BNCs make. Yes, you really do have to worry about what happens out to 50 MHz, or maybe higher. 15 MHz does not cut it, despite what some industry "experts" claim. (Mind you, they are "audio" experts, not "RF experts. The caveat.)
http://analogresearch-technology.net/geekspeak.html
Pat
Ok guys, this is probably a simple question for you, but I'm trying out new digital cable and it has 'Ground end' marked on it at one end of the cable.
Does this end go at the source (CDP) or at the DAC end? :)
Oh, and what's "Ground" got to do with it? :scratch:
Marco.
OK, since this is what started the thread, and I was not here......
What kind of cable is it? Coax? Twisted pair?
If it is a coax, then is should not matter.
If it is a twisted pair, then it has more problems that just what end goes where. A twisted pair, with a shield that is connected at one end only, is a mess, when it comes to impedance. In short, it has a different impedance, depending on what end you are looking at.
I would hope that is not what you have, but I have seen mic cable sold as a "digital" cable. Really bad idea. ("But it sounds good as a mic cable. Why won't it sound good as a digital cable?" Run from anyone who says this!)
Pat
Ah, that's interesting... Still, I'd never use a 6m cable - it's just too impractical ;)
Besides, I'm very happy with the results I'm getting from the VDH :)
Marco.
Uh............it was designed to be coiled up. Nothing impractical about that.
Pat
Dave Cawley
02-12-2009, 08:21
Hi Pat
I am very experienced! I have read your technical description on your web site and have a couple of questions:
How did you establish the 30dB return loss aim? What happens when the return loss is lower and do you have measurements to show this?
How did you establish the frequency response up to 10Mhz (was that the aim) ? What happens when the response is lower and do you have measurements to show this?
I'm an analogue person so would be very interested to see the effects.
BTW, Texas, what a great place, although bigger that England, Wales and Scotland put together? What is that nice looking lake to your East like?
Thanks
Dave
Hi Pat,
Welcome to AoS :)
Sorry to appear thick, but which company is it you represent (or perhaps Dave could let me know this)?
I will then open up a free trade account for you.
Marco.
Ok, got it now.
Hi Pat,
I may be interested in buying one of your U-byte cables and reviewing it on the forum, as it sounds very interesting. However, there's no picture of it on your website (unless I missed it)... A picture of the cable to accompany the 'techincal blurb' and description would work wonders...
Any chance of taking one and popping it on here? :)
It might seem insignificant, but I like to look at the construction/appearance of cables before buying them.
Also, one question: have you tried comparing the sonic effect of your cables in the standard length you supply and also, say, in a 0.5m or 1m version? If so, was there a difference and what did you hear?
Please note that I'm simply an audio enthusiast and music lover, not a scientist or engineer, so measurements mean little to me. It's (subjective) perceived sonic effects heard which I relate to.
I just need to be suitably 're-eductated' in respect of the length issue because experience so far tells me that the shorter a digital cable is, the better (I've done comparisons), however that is of course using cables with RCAs fitted.
Cheers :cool:
Marco.
Hi Pat
I am very experienced! I have read your technical description on your web site and have a couple of questions:
How did you establish the 30dB return loss aim? What happens when the return loss is lower and do you have measurements to show this?
We did a series of listening tests. We found that when we got into that range on both ends, as well as the cable, it was hard to tell any difference. Also, one can look at the recovered clock on the DAC end, and measure what happens as those parameters change.
How did you establish the frequency response up to 10Mhz (was that the aim) ? What happens when the response is lower and do you have measurements to show this?
10 MHz is too slow!!!!!!!!!!!!! That won't pass the fifth harmonic. That came about when we started looking at stock outputs of various transports. The ones made by giant conglomerates all were crippled, in the sense that the BW was severely limited. This was done to pass EMI criteria. Smaller firms (usually US ones) weren't interested in pleasing the regulatory agencies. They all sounded better. "Fixing" the slow ones made them sound more like the ones made for the lunatic fringe.
Low BW equals slow rise times. Fast rise times are essential to minimise the effects of jitter.
I'm an analogue person so would be very interested to see the effects.
BTW, Texas, what a great place, although bigger that England, Wales and Scotland put together? What is that nice looking lake to your East like?
There is something of interest to our east? Huh, we never paid any attention to it.
Pat
Dave Cawley
02-12-2009, 14:06
Hi Pat
Also, one can look at the recovered clock on the DAC end, and measure what happens as those parameters change.
What did you measure? and what were the results?
10 MHz is too slow!!!!!!!!!!!!! That won't pass the fifth harmonic.
Is the fifth harmonic essential, how much does that add to the jitter?
Thanks
Dave
Hi Pat,
Any chance of an answer to my questions please? :)
Oh, and a photo of your cable.
Cheers!
Marco.
Ok, got it now.
Hi Pat,
I may be interested in buying one of your U-byte cables and reviewing it on the forum, as it sounds very interesting. However, there's no picture of it on your website (unless I missed it)... A picture of the cable to accompany the 'techincal blurb' and description would work wonders...
Any chance of taking one and popping it on here? :)
Hmm.............never thought of that. The person who did that sort of thing has moved on. I suppose that I could try. No promise that it will be ok.
It might seem insignificant, but I like to look at the construction/appearance of cables before buying them.
We are not the consumers. We tend to think in other ways. If it was totally up to me, everything would look rather bog standard.
Also, one question: have you tried comparing the sonic effect of your cables in the standard length you supply and also, say, in a 0.5m or 1m version? If so, was there a difference and what did you hear?
Ok.............long story short............
We made our only DAC back in the early 90s. It took about 1/2 hour to realise that something was not right. The cable was too short! I used the first cable that I had, something in the range of 1/3 m. It sounded like dreck. So, I started pulling out every cable that we had in the shop. (Quite a collection. Not all was 75R; a lot was 50R, as that is common for the test equipment that we use.) A very enlightening experiment. Found out that the longer ones weren't as intolerable to listen to. But, still too short.
Sat down, figured out why that was so, made some longer ones...................problem fixed.
Prototype DAC built in one night. Six more months to fine-tune the input stage. Probably another six months to come to the "30 dB rule". Lots of listening, lots of measuring.
Please note that I'm simply an audio enthusiast and music lover, not a scientist or engineer, so measurements mean little to me. It's (subjective) perceived sonic effects heard which I relate to.
If you were an engineer, you would say that "bits are bits", and that I was a gormless get.
Wait a tick...............audiophiles who have never tried anything longer than 1/2 m say the same thing! Funny thing is, I have been saying this since '92. None of our detractors ever went out and bought a cheap piece of RG-59, that was several metres long. All they would have to do is to try it. Then chop off about 30 cm, and see how that sounds. Nope, that is too easy. Much easier, and entertaining, to watch some high-strung nut from Texas goes bonkers!
I just need to be suitably 're-eductated' in respect of the length issue because experience so far tells me that the shorter a digital cable is, the better (I've done comparisons), however that is of course using cables with RCAs fitted.
Somewhere.............lemme find that link..............(dramatic pause).........
http://www.audiocircle.com/index.php?topic=60136.0
That might help some. Short cables don't really make the rise time much faster, and allow all the reflections to arrive too soon.
Speaking of which.............how many reflections are too many?
Look at it this way:
If you have a return loss of only -14 dB (and a lot of you do, you just don't know it!), that means that 20% of the incident wave bounces back to the source. Assuming that it is also -14 dB (yes, they are out there!), 20% of that 20% that has bounced back gets bounced back to the receive end. That comes out to be a first reflection that is 4% of the original. Might not seem like much. It isn't. Not enough to cause errors, but enough to slightly screw up the decision point of the circuitry.
Think about how much money most of us have spent because a we found an interconnect that slightly made things sound better. Our friends think that we are daft, but they listen to ..............uh, might not be allowed to mention a company that fits the "No highs, no low, must be ****." space.
Slightly is all that there ever is. Usually so slight, that folks like me have not found a way to measure it. But we can hear it. Luckily for us, this SPDIF stuff is easy to measure. And hear.
Let's say that you can make something that is now -40 dB (wonder who that is?). You now only have 1% to bounce back. And 1% of 1% is only 0.01%. Not enough to even register.
One thing that I should mention: just because we can measure it does not mean that we can say "Yes, this will sound x% better than that other brand." Doesn't work that way. Try reversing the direction of your cable some time. Does it sound different? Well, it might! I can measure any cable, and see a difference in either direction. All that I could tell you is that you might hear a difference, but there would be absolutely no way to tell which way will be better.
Unless someone took a twisted-pair mic cable, connected the shield on only one end................then I could tell you.
But then you would probably want to kill me!
Ok, some probably do already. Just another day in the life.
Pat
Hi Pat
Also, one can look at the recovered clock on the DAC end, and measure what happens as those parameters change.
What did you measure? and what were the results?
10 MHz is too slow!!!!!!!!!!!!! That won't pass the fifth harmonic.
Is the fifth harmonic essential, how much does that add to the jitter?
Thanks
Dave
It is not that easy to describe. You have to look at the frequency content of the clock. Most of it just looks like noise. You can see an increase in a certain area of the spectrum, or a rise in the noise floor. Other ways to tell include listening to the control voltage on the PLL. You can hear a highly distorted version of what you are listening to. Sometimes, that is what we go by. There is not a simple answer to your question. Wish there was. It would make designing a lot easier.
Ditto for the fifth harmonic. You can not say exactly how much the jitter will increase if it is not there. (If it isn't, you are in trouble! Won't look anything like a square wave.) All that I can say is faster rise times yield a much cleaner signal to begin with. If you start with dreck, you can't fix it on the RX end.
Pat
However, there's no picture of it on your website (unless I missed it)... A picture of the cable to accompany the 'techincal blurb' and description would work wonders...
I can make a nice picture of it this weekend, as I have one. ;) I will post it in my gallery, if you want.
Dave Cawley
02-12-2009, 15:23
Hi Pat
I was hoping having read and understood your technical explanation on your web site, and recognising a fellow engineer, that you would be able to measure the effect as numerically as you do with return loss.
I need convincing that 30dB return loss is essential and high frequency responses are desirable. With your test gear I'm sure you could provide the measurements and hence bury your competition?
I do like the thought of measuring the VCO control voltage noise, have you FFT analysed this with different cables?
Regards
Dave
It is not easy, as to just assign a number to it. It all looks like a lot of noise. Trying to pick out something, and assign a value to it is a lost cause.
Even when it comes to jitter: a number all by itself means nothing. X pSec of jitter, without knowing the frequency of the source, and the spectral content of the jitter tells us nothing.
If you look at the way most commercial outfits spec their oscillators, you will see "X pSec, measured >1 kHz from carrier." Useless information. Even the most bog standard oscillator will look good, if you only measure above 1 kHz out from the carrier. But, if they measured it at say...............10 Hz, and the next guy still uses data above 1 kHz, some git in purchasing would say "Hey, this one is 12 dB better. We should buy it." Only to find out that it much worse at frequencies close in to the carrier.
Which is why we make our own, and spend hours trying to measure phase noise at 1 Hz offset. Hard to do, as thermal drift of ANYTHING in the measuring chain makes such measurements a pain. But we know how much quieter our stuff is than some of the highly-touted stuff out there. (Some of them are rather disappointing. Barely better than what one can buy from Mouser for $1.)
Measuring return loss, by comparison, is easy. I hook it up to a TDR, and see how much the bump is. I can then go to a network analyser, and see what it looks like in the frequency domain. Much easier than trying to make sense of a FFT plot of the recovered clock that is 40 dB noisier than the reference that generated the SPDIF waveform. With spikes sticking out all over the place. YUK!
Trust me.
Pat
Ali Tait
02-12-2009, 19:19
So,what length of cable would you recommend using Pat?
I can make a nice picture of it this weekend, as I have one. I will post it in my gallery, if you want.
Hi Dimitri,
I'd like that, thanks! - Just to give me an idea of how well constructed it is, and how 'bulky' the whole lot is when coiled up :)
Let's see some nice close-ups, please!
There's something else I'd like you to do though (unless you've done it already) is to compare the A.R.T digital cable to the Mark Grant one. The G1000HD interconnects you bought recently are coaxial in construction, and therefore double-up as digital cables...
All you have to do is use a single interconnect as a digital cable. You'll obviously have to use some other interconnects as analogue pairs whilst doing the test, but at least you'll be able to compare the A.R.T digital cable to Mark Grant's, sonically - and then of course let us know the outcome! :cool:
This will give me an indication of their efficacy, and whether to proceed with trying them myself, since Pat's not into listening to things and so can't describe himself the sonic effects of his cable ;)
Pat,
Some interesting replies there. I'll come back to you later on them. I'm glad that you seem to be settling in ok! :smoking:
Marco.
Spectral Morn
02-12-2009, 20:58
So,what length of cable would you recommend using Pat?
On his web site...his digital cable is 16ft long.
Regards D S D L
Primalsea
02-12-2009, 21:35
I have a question, no really I do.
Is comparing the difference of the source signal to what is recovered by the receivers circuitry problematic because the reflections distort what you measure at the sources output as well as what you measure at the receiver?
Hopefully that makes sense.
The G1000HD interconnects you bought recently are coaxial in construction, and therefore double-up as digital cables...
Huh?
Where is it written that digital cables must be coaxial?... or vice versa, for that matter? :scratch:
I didn't mean that, matey - just that Dimitri can use his G1000HDs as either analogue or digital cables :)
Marco.
You could do that with pretty much any cable. Might not be any good though... ;)
Is the MG cable 75ohm?
Hi Dimitri,
I'd like that, thanks! - Just to give me an idea of how well constructed it is, and how 'bulky' the whole lot is when coiled up :)
Let's see some nice close-ups, please!
There's something else I'd like you to do though (unless you've done it already) is to compare the A.R.T digital cable to the Mark Grant one. The G1000HD interconnects you bought recently are coaxial in construction, and therefore double-up as digital cables...
All you have to do is use a single interconnect as a digital cable. You'll obviously have to use some other interconnects as analogue pairs whilst doing the test, but at least you'll be able to compare the A.R.T digital cable to Mark Grant's, sonically - and then of course let us know the outcome! :cool:
This will give me an indication of their efficacy, and whether to proceed with trying them myself, since Pat's not into listening to things and so can't describe himself the sonic effects of his cable ;)
Pat,
Some interesting replies there. I'll come back to you later on them. I'm glad that you seem to be settling in ok! :smoking:
Marco.
Consider that done, Marco, but I need to setup correctly my amplifier chain : I'll do it as soon as the Croft amps are settled.
In the meanwhile, I will make some close-up photos on Saturday, along with the review that I had prepared on Pat's cable (and I never posted anywhere... :(). I hope Pat is not too angry on me, but I had some trouble back then.
Nice one, Dimitri - I'll leave it with you :)
Marco.
You could do that with pretty much any cable. Might not be any good though... ;)
Is the MG cable 75ohm?
I think so. Check with Mark. His cables work superbly well in either digital or analogue applications :)
Marco.
Marco:
If you are making your own, most anything over 2 metres or so would be fine.
"Then why is yours so sodding long?"
("You can't say 'sodding' on the telly.")
OK, bub........want some more honesty?
Years ago, we made a digital cable, designed to sell in retail shoppes, for USD100. Guess what?
The dealers slagged us off. "Get that cheap thing out of here. Don't even bring anything back that isn't at least $200. How are we supposed to make money on anything that cheap? Besides, if it is cheap, it has to sound cheap."
Not sure about the logic of that last part, but................
Anyway. We were faced with making something that was really good, and cost $200. But, no matter how good it is, how expensive the parts are (trust me, the parts in the U-Byte are way out of line for its retail price), it is just wire. Wire, maybe expensive wire, that needs custom-made BNCs, but it is still just wire. We could not live with charging $200 for something that was only 2 m or so in length. The price point would have been more in line with industry norms, but it is just wire. Nothing magical, special, or fancy about it. You just have to be willing to buy a big roll, shell out $$$ for custom BNCs (and wait a month or so, if lucky) for them to show up.
What ended up was a cable that 6 m or so long.
1.) Making it that long made it look like you were getting more value for money, compared to one that was 1 or 2 m long.
2.) We had to buy a really big roll of it, so let's use it up as fast as we can.
3.) It was different. Dealers who wanted to sell it really had to want to sell it. Just sticking a 6 m coil of coax out for display wasn't going to cut it. They had to be willing to invest some time in selling them.
Sadly, most dealers don't want to spend more than 5-10 minutes on a sale. If they sense that it is going to take longer, they lose interest rather quickly. Things might be different in your shoppes, as you guys tend to be a more polite culture, that thinks nothing of "queuing up" for stuff. But not over here.
So, there you have it..........2 m........3 m.........probably works just as well. You get more for your money at 5 or 6 m. And the dealers can't sit on their arse and sell them.
We have some very prominent friends in the industry that say "What are you doing, fooling around with a $200 cable? As much as that must cost to build, it should be priced at $600, if not higher. You will probably sell just as many at that price point, and we won't have to listen to all of you gripe on how poor you all are."
Yeah, we could but.............it is still just wire. Not a damn thing that we could do to it that would justify the price. So, $200 it is. We shortened it a tad, because the bare cable costs about 50% more than it did way back when.
Sad thing is, there are some spawny gits who would actually think that at $600-$1000 that it had to be really good. It is, but then we would have to deal with them. No thanks. We'll stay paupers.
BTW, all of our dealers went out of business throughout the 90s. As the Home Theatre market evolved, it squeezed out a lot of high-end gear. Prior to that time, dealers would sell mid-fi to keep the doors open, and high-end was gravy. The smaller dealers were not able to secure the "Big Boys", so they had to settle for more obscure brands. Like ours. But the smaller dealers could not make the transition. There really weren't small fish players in the HT market. The elite salons still had enough floor space to carry the Big Boys, and had no need for fly-by-night outfits.
More so if they are in some dust bin like Texas!
Pat
Dave Cawley
03-12-2009, 08:35
Hi Pat
Well, return loss doesn't tell you much, and I'm still curious why you set a 30dB spec on it, you must have measurements to back this up? "measure phase noise at 1 Hz offset" this is more like it, can we see some results please?
Thanks
Dave
I thought that I had explained that part. OK, I'll try again.
We found, through careful listening tests, that when got to the point that all cables started to sound alike, and you could not tell a difference, was when we got to the -30 dB range. Initially, we weren't that close. RX chips are notorious for disrupting the input rho. (Almost all of them use Schmitt triggers, which are regenerative circuits. They spit all sorts of stuff back. Some guys stick clamp diodes on the input. Also problematic.)
Anyway, we had to do a lot of tweaking (6 months, in the case of our DAC, before we released it) in order to get the rho that low. We also did some tests on how fast the rise time needed to be, but since we did not make transports, we just made all the ones in the shoppe as fast as possible.
If you want detailed lab tests to document this, there are none. If there were, we are talking over 16 years ago that this work was done. We did not see a need to carefully document this. In the wide world of microwaves, where I came from, we had rules of thumb for what numbers to shoot for, in various applications. A lot of them were having a rho in the range of -26 or -28 dB. A few dB lower for SPDIF did not sound out-of-line.
Or you can just use simple math and calculate that -30 dB rho is about 3% reflection. If -26 db (5%) is good enough for sending 1800 channels of voice traffic up a piece of waveguide, to the antenna, can SPDIF be significantly more demanding? If the far end has a -20 dB rho, then the amount of the first reflection is only 0.3%. Which is only 1.6 mV, on an SPDIF signal.
As for phase noise results, that is not something that we would care to make public. Partly because we don't want to start a row over how bad certain commercial products are. Yes, we have data on that, but a great deal of that was done as part of our consulting work. I don't think the folks who paid us would appreciate us disclosing what we found out about their competitors.
As you might imagine, consulting work goes as the industry goes. It is down quite a bit. One of the main reasons that we get any work is that we are good at not disclosing what we have done, or who it was done for. I hope that you can appreciate our position.
Obviously, that does not apply to anything that we build, but I don't see an urgent need to share that data. Especially since almost no one spec's their products lower than 1 kHz. The numbers at 1 Hz would sound awful, taken out of context. You would have to know what awful truly is to appreciate that ours are good.
Sorry, bub.............about all that I can provide.
Pat
His cables work superbly well in either digital or analogue applications :)
Oh, I don't doubt it... I've yet to see a bad word said about them. ;)
Dave Cawley
03-12-2009, 16:48
Hi Pat
I know it's terribly easy to measure return loss. But I'm having difficulty reconciling 30dB as a bench mark. In another thread you said that you measure everything, and I was hoping that you could define the reasons for 30dB, using measurements, not 16 year old listening tests. But I understand this works for you.
I fully understand your reasons for not sharing your test results, respect that, and will not ask again.
Many thanks
Dave
Oh, I don't doubt it... I've yet to see a bad word said about them.
Hi Mikey,
Well I guess that if a product (cables or anything) is genuinely very good, then there's unlikely to be a bad word said about it...
Credit where credit it due :)
Marco.
Dave Cawley
03-12-2009, 16:55
Hi Pat
I know it's terribly easy to measure return loss. But I'm having difficulty reconciling 30dB as a bench mark. In another thread you said that you measure everything, and I was hoping that you could define the reasons for 30dB, using measurements, not 16 year old listening tests. But I understand this works for you.
I fully understand your reasons for not sharing your test results, respect that, and will not ask again.
Many thanks
Dave
Why is -30 db such a hard concept to accept? As I said, in a lot of commercial applications (from a previous lifetime), 28 db RL was considered more than acceptable.
Just because we love to measure does not mean that we love to document. Quite the opposite. There are almost no written records of anything that we do, or have done.
OK, let me turn the question around: what would you think that the goal should be, and why? Do you have any data to suggest that is should be something else.
Let me try to inject some levity into this. Maybe you will understand my point better by this.
Many years ago, probably on my least fave forum (Audiogon), some guy who knew everything that there was on tubes was telling me how to design a balanced input stage. We were getting nowhere fast, except to grate on my nerves. So, I asked him: "Well, what would you consider that the input CMRR should be?"
"Oh, I would think that it would be at least 80 dB, out to at least 1 MHz."
I did not respond.
I guess to expound on my question to you would be to ask if you have measured any of your gear? If you look at the graphs on our site where we show what happens when you stick an RCA jack into the chain, it goes above 30 dB at 20 MHz. (Assumes that you have something better than that to start with, and few do.)
OK, here is something that you might be able to grasp.
I have demonstrated that you can hear the insertion of a 50R barrel in between a pair of 75R cables. The rho of that is not much. Yet, a careful listener, on a system that they are very familiar with, can hear a difference. Granted, not much. It is subtle But they can do it. (Please don't start in with a demand for DBTs, or I will go bonkers.) You can show what the rho is of that on a TDR. It is more than the rho of any piece of equipment that we have made. So, why is -30 dB hard to accept as a goal?
"Why don't you like Audiogon, since you brought it up?"
The folks who moderate it let any Average Joe question anything that I say, but take everything out of context, and twist and distort it to the point where even I would think that I am an idiot, if I said those things. Simple replies, such as "That is not what I wrote. Please carefully re-read what I wrote. I would appreciate it if you would not put words into my mouth." are always removed.
So, if someone just stumbles upon a thread that I have participated in, the last word is (more times than not) someone pointing out to the entire world that I don't even know the basics. So, why bother? Sad thing is, in this present way of doing things, a presence on the forums is essential. And Audiogon is one of the more important ones in the US.
But, as bad I as I may feel that it is, it not a "DIY"/hobbyist forum. They are the worst. A very famous designer, who some here may know by reputation, participates in one of the larger ones. He gets his teeth kicked in on a daily basis. Seems that the DIY crowd gets their jollies by taking pot shots at someone who has accomplished more in life than they have. Bringing him down a peg (or five!) must make them feel better about themselves.
Art of Sound, from my limited exposure, is great place. Taken either all alone or in comparison to other forums. I will try to answer all questions as honestly and as forthright as I possibly can. However, some things I am unable to share. I am glad that you understand that.
Cheers,
Pat
I didn't realize that DYI forums were cruel to that point. This is really a pity. :(
It seems you're in a good mood for answering questions, so, ! have two :
- Why are there so few designs of class D amplifiers out there ? Is it so hard to design one ? (ok, I'm a bit OOT here, never mind)
- I've come to the conclusion that when people say they hear differences between digital transports, what they really listen to is rather the inability of their dac to cope with transport jitter. Am I wrong ?
Primalsea
05-12-2009, 10:09
Was my earlier question stupid or something, no body replied, not even to tell me it made no sense, jibberish or anything??:(
Dave Cawley
05-12-2009, 16:49
Maybe it fell into : " However, some things I am unable to share. I am glad that you understand that. "
Dave
Was my earlier question stupid or something, no body replied, not even to tell me it made no sense, jibberish or anything??:(
Are you referring to post #109? I missed it, first time around. Now that I have read it, I am not sure what the question is.
If you are asking what reflections look like, or how they affect the waveform, the best way to answer that question is to stick a really long cable in the chain. The "squiggle" that reflections create can then be clearly seen, riding on the top of the waveform. Does that help any?
I have been out of the office, on non-business related matters a lot the last week or two. An unfortunate trend, that may continue for some time. If no one hears from me, I am still around, just too far behind and trying to catch up on essential tasks.
Cheers,
Pat
Primalsea
13-12-2009, 12:26
Hi Pat, The question was partly made on an assumption that the test is based on comparing the test signal from the signal generator to the signal received by the DUT in real time.
I was wondering if the reflections initiated at the DUT end actually then have a detrimental effect on the original test signal. The picture in my mind is of reflections bouncing back down the cable and becoming part of the original test signal.
Been busy, sorry for the delay in response. Trying to get parts is getting tough. Long lead times for many parts is common the last few weeks. Semiconductors are the most likely, but some SMD caps we use are almost impossible to get.
Unless you want to wait until April.
Anyway, your question.........
The test we use is what is known as time delay reflectometer. Not so sure that I can easily explain it in layman's terms, but it operates on the principle much like radar.
A step pulse is sent down the transmission line. As it hits an impedance aberration, you will see its effects, separated by the amount of time it takes to reach it. You have to know the velocity of propagation for the cable, usually around 70% or so. Knowing the amount of time it takes to reach said point, one can then determine where on the cable that point is. We can probably resolve to less than 1 cm.
The trick to making a reliable test is to use a cable that is long enough to swamp out any aberrations that occur at the point where the cable connects to the machine. (The far end is terminated.)
Likewise, if you replace the termination that is normally at the end, you can place an active component (like SPDIF TX or RX), and determine its impedance. With our ability to resolve impedance changes so close together allows us to see the effects of the connector, the cable that goes to the PCB, the circuitry between that point and the active circuit, as well as the active circuit itself.
After all of that circumlocution and blathering...........
Now, if I read your question right, yes, a gross impedance mismatch will show up as a series of glitches. But, it will be reduced by a predictable amount at the sending end. It takes experience working with this stuff to be able to pick out what reflection is caused by what mismatch.
If you had a very short pulse, then in a sense I can see where you would be concerned that the reflection from the far end could distort the waveform enough to throw the measurements off. Our test equipment uses a very long step, so that only the rising edge is used. In addition, we always use a 1 or 2 m cable to make sure that we are not looking at any aberrations that occur at the TX end.
So, let's see how this can work in the real world, instead of our lab.
Let us say that you have a 1 m cable. It will take around 4 nSec for the first reflection to bounce back to the transport. Which is not far from the rise time of the signal itself. (Our TDR has a rise time of around 150 pSec. No problem to see a 1m cable.) So, if you think about that for a second, maybe you can see where the germ of inspiration for the U-Byte came from.
Hopefully, that not only answers your question, but gives you some insight into how it all fits together.
Pat
Primalsea
15-01-2010, 19:01
Thanks Pat, I see that I had a bit of a misconception regarding how impedance mis matches are measured.
While reading the above a question did flash up in my mind. In your opinion is there any benefit to having analogue cable with a propagation of 90% compared to 70%?
No, because it would have to be longer to alleviate jitter problems. Besides, it is not likely that you have something like 600 ohm "ladder" air line for your link.
A bit o/t but...........some of you may have heard this story before.........
When we figured out that length was a problem (about 1/2 hour after our first DAC proto was turned on), we tried every cable in the shop. One was a very expensive 1 m Gore-tex cable, that was part of a microwave network analyser set-up. It cost USD2000. In 1982. (Don't ask why we had it!)
It sounded fantastic! Cleanest top-end we had heard.
Just had zero bass.
Of course, it was 50 ohm. Like most RF cables.
(If it makes you feel better, I eventually returned the cable to the lab that owned it! Not any good for audio, so..............)
Pat
Powered by vBulletin® Version 4.2.3 Copyright © 2024 vBulletin Solutions, Inc. All rights reserved.