Direct Attach Cable Thread

@Superdad:  "Do not expect to be get great performance from a Direct Attach Cable between for example a $20 FMC and an EtherREGEN--even if you use linear power supplies on both." 

I was about to experiment with that, just to see if the SFP+ DAC remained backwards compatible in that application. Not worth the effort? 

53 minutes ago, Superdad said:

 

... even if you use linear power supplies on both.

 

To secure galvanic isolation from the mains (electrical isolation - decoupling the power supply from the mains) a battery solution is preferable. 

Torben

1 hour ago, Superdad said:

But the topic of this thread, Direct Attach Cables, is really about the potential performance difference between usage of a DAC versus SFP fiber-optic transceivers.

Maybe it was not @audiobombers intention, or yours for that matter, but to expand our reach, why not delve into both the theoretical aspects and practical applications of passive DAC cables (and SFP fiber-optic transceivers for comparison)? Here we will also explore the real-world performance of DAC cables, offering practical insights and discussing key characteristics for a comprehensive understanding before choosing a DAC cable.

16 minutes ago, di-fi said:

Maybe it was not @audiobombers intention, or yours for that matter, but to expand our reach, why not delve into both the theoretical aspects and practical applications of passive DAC cables (and SFP fiber-optic transceivers for comparison)? Here we will also explore the real-world performance of DAC cables, offering practical insights and discussing key characteristics for a comprehensive understanding before choosing a DAC cable.

Sounds good to me.  

As I mentioned previously (in this thread I think), we believe that the quality of the dielectric insulation used on the twin-axial wire pairs is the most important performance factor (assuming proper 100-Ohm pairs). Very much doubt that wire gauge or silver-plating will make much if any difference.  

Hopefully we have to put to rest any notion that there could be performance difference between--same brand/wire type--of SFP+ and SFP DAC cables, as they differ only by the former having a pin tied to ground to permit 10G SFP+ cages to send at that rate. [But there certainly can be performance differences between 10G SFP+ fiber-optic transceivers and 1G SFP modules--owing to tighter jitter requirements for the former.]

Melco C1 (2m.) DAC compared to Taiko / FS DAC (2 reviews)
 

• Naturalness (less digititis)
• Better bass
• Better image

Quote

Furthermore you can try to use longer DAC cables, longer = fuller/warmer. This is more likely due to the wire gauge used, it ranges between 30 and 24awg depending on the length (longer = larger diameter). And yeah there’s always soms burn in related effect of course.

https://www.whatsbestforum.com/threads/taiko-audio-sgm-extreme-the-crème-de-la-crème.27433/post-978633

Be careful with cable length and passive DAC cables. I have sfp+ devices that work perfectly with 10g fiber at 10 meters but stutter with 24awg DAC cables of the same length.

1 hour ago, di-fi said:

Melco C1

Too bad they don't specify the most important part of the construction--the dielectric used on the conductors themselves!

37 minutes ago, lmitche said:

Be careful with cable length and passive DAC cables. I have sfp+ devices that work perfectly with 10g fiber at 10 meters but stutter with 24awg DAC cables of the same length.

10 meters? Yikes!  I would not go longer than 1-2 meters with Direct Attach Cable.  Remember, these are LVDS pairs--with the sorts of signal voltages usually used internally on. a board or to feed an active transceiver module to drive a fiber with a laser.

The whole point of playing with DAC connections is to improve things. A long cable, as Larry here has found, will make things worse--even to the point of having dropouts!

Looks like the opportunistic folks over at AfterDark have also jumped into the audiophile silver-plated DAC cable fray:

https://www.adark.co/collections/project-constellation-fiber/products/afterdark-project-clayx-constellation-reference-ix2000-sfp-sfp-lan-cable-for-audiophiles

About half the price of the Melco.

But honestly, except for some fancy coloring and outer braids, I am pretty sure that both firms are just gussying up and rebranding some silver-plated commercial cables they found.  I really think such pricing is rather silly. But to each their own... 

4 hours ago, di-fi said:

Melco C1 (2m.) DAC compared to Taiko / FS DAC (2 reviews)

No reviews there. just a couple of opinions. A real review will need a lot more substance.

7 hours ago, Superdad said:

Do not expect to be get great performance from a Direct Attach Cable between for example a $20 FMC and an EtherREGEN--even if you use linear power supplies on both

I was not intending to try one between the oMD v2 and ER in my main system, because it is a 15 meter run. I was contemplating a DAC for my desktop system (10GTek SFPs, 5 meter Corning SMF, 10GTek 1G and TP-Link MC220L FMCs). The FMCs are not plugged into the same power conditioner, or the same outlet, so I believe your advice would be not to bother?

How ridiculous is it that they chose to call it a Direct Attach Cable (DAC)? They could have called it a Direct Connect Cable (DCC) to avoid confusion, and eliminate the need to use the redundant "DAC cable" term. 

17 minutes ago, audiobomber said:

How ridiculous is it that they chose to call it a Direct Attach Cable (DAC)? They could have called it a Direct Connect Cable (DCC) to avoid confusion, and eliminate the need to use the redundant "DAC cable" term. 

Well this cable type has been in existence for a long time in data centers. And I am sure we can all think of a bunch of inconveniently identical acronyms. A friend of mine (who happens to be black) works for the BLM—Bureau of Land Management, not Black Lives Matter… 

5 hours ago, audiobomber said:

No reviews there. just a couple of opinions. A real review will need a lot more substance.

and also comparison with similar DAC cables ideally, (not) like this? 

Melco C1-D20 SFP+

…..now, who will be the fist to put a comment in that Ear article? (About impedance). 

Maybe suggest a 4 meter has 200 ohm, but if you run them in parallel, it will be 100 ohm and maybe they will make a custom version for those in need as well. 

Replaced the Cisco SFP AOC cable between the opticalModule and opticalRendu with the Cisco Twinax SFP+ DAC cable and holy cow! Who delivered the subwoofer?!! Or in other words, it really lets my 160 do its Naim thing with the bass and mid-bass. Maybe almost so much so that I'm going to have work on some better isolation of my Audio Physic Compacts so as to not overpower the high end so much. I actually have the speakers in bookshelves in my office, which I know isn't the best, but it is what it is...

The DAC cable doesn't have quite the same airiness of the AOC, but it has an organic meatiness about it and is less fatiguing in the treble, and is a much funner listen imo, esp with older or poorly recorded rock records which could be a bit shrill and anemic with the optical cable. It's really something else, and the difference was not in the least bit subtle. Just for the fun of it, I may try another inexpensive DAC cable (should I go shorter? or longer - which would take me from 30awg to 24awg?) which may be much more difficult to parse any difference. 
 

On 6/20/2024 at 2:59 PM, Superdad said:

we believe that the quality of the dielectric insulation used on the twin-axial wire pairs is the most important performance factor

could not find any DAC with dielectric info other than Amphenol with Skewclear and polyolefin dielectric...

52 minutes ago, Mops911 said:

could not find any DAC with dielectric info other than Amphenol with Skewclear and polyolefin dielectric...

While polyolefin (XLPO, what most heat shrink tubing is made from) does not have as low a dielectric constant as the various Teflons (PTFE, PTE, or the best of all foamed PTFE), Amphenol's patented low-skew, bonded-pair manufacturing method is very interesting and advantageous for extremely high speed LVDS applications.

Their brochure about the Skewclear wire is worth reading:

https://www.mouser.com/datasheet/2/18/1/New SKEWCLEAR Brochure 2-25-02-8977.pdf

For the very simple direct attach cables they use that wire in this one series: https://cdn.amphenol-cs.com/media/wysiwyg/files/documentation/datasheet/cableassemblies/hsio_ca_sfp.pdf

Although the part numbers differ from the datasheet, it does seem that the big distributors do have stock of the SFP+ passive DAC cables that use the Skewclear wire:

https://www.mouser.com/c/?m=Amphenol&q=SF-SFPP2EPASS

Perhaps someone here will give it a try...

or this?

https://cablesondemand.com/sfp-cables-sfp-modules-1

in 0.5m https://cablesondemand.com/amphenol-sf-sfpp2epass-000-5-0-5m-sfp-cable-amphenol-10gbe-sfp-direct-attach-copper-cable-1-6-ft-sf-sfpp2epass-000-5

23 minutes ago, Mops911 said:

or this?
https://cablesondemand.com/sfp-cables-sfp-modules-1
in 0.5m https://cablesondemand.com/amphenol-sf-sfpp2epass-000-5-0-5m-sfp-cable-amphenol-10gbe-sfp-direct-attach-copper-cable-1-6-ft-sf-sfpp2epass-000-5

Yes, CablesOnDemand is Amphenol's direct resale site.  Just be sure to stay with the passive cables, as filtered with this link: https://cablesondemand.com/sfp-cables-direct-attach-1

Alex, the link is helpful. 

Quick question:  these cables are about three times as much as the generics available on Amazon, one of which I bought and am appreciating. Is it the dielectric that accounts for the price difference? 

Thanks @Superdad
I now only miss a couple of ER2’s 😀