SPDIF Digital Coax to AES/EBU XLR Inputs?

BrassEar · « **on:** 12 Jun 2006, 11:44 pm »

Thinking of trying the Behringer DEQ2496 EQ. However, the digital input options are optical (OK) or AES/EBU XLR (better?)

Can you use a standard RCA to XLR adapter to input a digital coax to this AES/EBU XLR connector or are special termination resistors required? The Behringer manual states that the AES/EBU XLR inputs also accept SPDIF signals.

Or is the safer approach just to use the optical in? I would be feeding this via a Squeezebox 3 which apparently has a lot more jitter on the optical vs. the coax.

TIA.

Russell Dawkins · « **Reply #1 on:** 13 Jun 2006, 12:38 am »

you cannot use a simple XLR to RCA adapter - the impedance of the input/outputs needs to be matched. The cheapest and simplest way is these transformer adapters by Neutrik, one of the best pro connector manufacturers:

http://mercenary.com/neditr.html

note XLRM = male XLR, XLRF = female XLR

Skynyrd · « **Reply #2 on:** 13 Jun 2006, 03:36 am »

The Behringer DEQ2496 is presently claimed in it's owners manual (available online) to work with either AES/BSU or SPDIF. You'll have to use an RCA to XLR connector if you use RCA SPDIF, but supposedly no transformer is necessary!

I do not own one of these units and haven't proven the manual's claim.

Anyone tried RCA (also called coaxial) SPDIF through an RCA/XLR connector on the DEQ2496?

Coaxial, RCA, SPDIF...what was the quesstion?

Skynyrd

Russell Dawkins · « **Reply #3 on:** 13 Jun 2006, 04:39 am »

SPIF data and AES/EBU data are largely, but not completely, identical. What is different is the voltage and impedance, the AES being, I think, 2V and SPDIF being around .5 - .7 V. AES has a characteristic impedance of 110 ohms, while SPIF is 75 ohms. If you plug one into the other through an adapter it will work but won't sound as good due to reflections back and forth within the cable.

Every time an ultra high frequency electrical signal encounters an impedance mis-match in its path part of it is reflected back and part goes through. If you have a mismatch at each end, which I gather is more common than we would like to know about, the signal is reflected back and forth a number of times before it dies out. Each time it reaches the connector some goes through and some gets reflected back.

Apparently this can cause confusion in the receiving D/A converter as it is receiving multiple word clock signals and has to "sort out" the valid from the invalid. These multiple word clock signals are called "timing jitter" and tend to degrade the sound. The better the D/A, the more immune to jitter. This is actually one of Benchmark's (a pro D/A converter) claims to fame - the ability to deal with very jittery signals. Benchmark used to say save your money - buy the cheapest digital player with a digital out and thier converter.

I know this explanation probably has technical inaccuracies in it, but I think it is roughly correct and is basically as much as I want to know about it. My brain's already nearly full.

By the way, if you have to send digital signals over a long distance, the best system is AES converted to 75 ohms with a BNC connector. Read about it here:

http://www.pcmag.com/encyclopedia_term/0,2542,t=AESEBU&i=37583,00.asp

ryno · « **Reply #4 on:** 13 Jun 2006, 05:01 am »

In one of the input menus you can choose between S/PDIF and AES/EBU.
Why would you need a transformer if the eq thinks you are using S/PDIF? Wouldn't the eq be expecting the 75 ohm 0.5-0.7 V signal?
I've only used glass optical with mine, and been very happy.
Ryan

Russell Dawkins · « **Reply #5 on:** 13 Jun 2006, 05:21 am »

For a definitive answer you would have to approach Behringer. They may change impedances, but I'd be surprised. If the input does not match impedances then one type would be favored over the other, but good luck getting that info from Behringer. On the other hand, they may have great and responsive technical support.

I imagine they probably take the expedient approach and let the chips fall where they may, knowing but not telling which format the device is, in fact, optimized for, and also knowing that it doesn't sound quite as good with the other.

Or they may have a very jitter-resistant input.

I forgot to mention in my last post the superiority of glass, properly terminated with gel (again, to eliminate reflections). That's why phone companies use it for miles without a repeater. Better than any metal. Wadia used (and still may) specify glass for their digital I/Cs.

Dan Banquer · « **Reply #6 on:** 13 Jun 2006, 11:53 am »

SPDIF is based on a 75 ohm impedance and AES/EBU is based on a 110 ohm impedance. If you mix the two without using a transformer/balun to convert impedances than I can guarantee you will VSWR/Return Loss issues.
d.b.

BrassEar · « **Reply #7 on:** 13 Jun 2006, 04:53 pm »

Thanks for the detailed responses. I guess I will play this safe and go optical. Any good sources of decent quality but reasonable optical cables?

TIA

gerchin · « **Reply #8 on:** 13 Jun 2006, 05:43 pm »

This should work just fine on the Behringer:

http://archives.telex.com/archives/Merlin/ISP-100%20App%20Notes/Merlin%20-%20MDM-1%20SPDIF%20Adapter.pdf

dwc · « **Reply #9 on:** 30 Jun 2006, 02:01 am »

I use a pair of simple HOSA RCA->XLR adapter cables to run SPDIF into the DEQ and it sounds just fine. (i.e. absolutely no loss of quality vs glass toslink input.)

I do run glass toslink on the outbound side into my Panny to the Kestrels.

-Dan

Bob Reynolds · « **Reply #10 on:** 30 Jun 2006, 10:47 pm »

Quote from: Dan Banquer on 13 Jun 2006, 11:53 am

SPDIF is based on a 75 ohm impedance and AES/EBU is based on a 110 ohm impedance. If you mix the two without using a transformer/balun to convert impedances than I can guarantee you will VSWR/Return Loss issues.
d.b.

My understanding is that SPDIF and AES/EBU are data formats and should have no specification of cable characteristic impedance.

Common coax cable has a 75 ohm characteristic impedance. It is common to see RCA connectors used on the SPDIF digital outs of CD players though the RCA connector is not 75 ohms. A BNC connector is 75 ohms and would match the coax cable.

Shielded twisted pair (STP) typically has an impedance of 110 ohms and the XLR connector is 110 ohms so they match.

I believe that at higher frequencies the cable begins to act more like a true transmission line and impedance matching becomes important. At digital audio frequencies, it would probably be safer to use an impedance matching transformer as stated above.

Russell Dawkins · « **Reply #11 on:** 1 Jul 2006, 12:17 am »

Canare makes one of the few 75 ohm RCAs out there. Blue Jeans sells them.
http://www.bluejeanscable.com/pages/technicaldocs/canarercap.pdf
It is my understanding, as mentioned above, that impedance mis-matches are a major contributor to a type of timing error known as jitter.

Bob Reynolds · « **Reply #12 on:** 1 Jul 2006, 04:12 am »

Quote from: Russell Dawkins on 1 Jul 2006, 12:17 am

Canare makes one of the few 75 ohm RCAs out there. Blue Jeans sells them.
http://www.bluejeanscable.com/pages/technicaldocs/canarercap.pdf
It is my understanding, as mentioned above, that impedance mis-matches are a major contributor to a type of timing error known as jitter.

I'm familiar with the Canare RCA connector and Blue Jeans cable. Every interconnect I own is from BJC and all the unbalanced ICs use the Canare connector. But, the Canare RCA is not a 75 ohm connector. As far as I know a 75 ohm RCA connector does not exist. I have a lot of respect for Canare and they make excellent products, but they are still a company with a marketing department.

Here's the relevant paragraph from the link you cited:

Quote

Canare’s all new 75W type RCAP Video crimp plugs are Impedance Matched
and achieve outstanding analog and digital electrical performance with a usable
bandwith to 200MHz.

Note the word "type" following 75ohm and the fact that impedance matched is first-letter capitalized. It's a subtle way of letting you think their connector is truly 75ohm and matches the impedance of 75ohm coax.

Now here's the relevant paragraph about their BNC connector:

Quote

Canare offers a full line of high performance MIL-C-39012
true 75Ω BNC connectors with impedance matched performance
characteristics and specifications that properly address
the latest generations of high bandwidth digital video
equipment.

Here they actually use the word "true" and impedance matched is spelled normally.

I believe that if you drop an email to BJC they will confirm that the Canare RCA connector is not a true 75ohm connector.

Russell Dawkins · « **Reply #13 on:** 1 Jul 2006, 05:05 am »

Good catch, Bob.
It looks to me like you may be quite right.
All these years I have taken it as gospel that whenever the Canare 75 ohm RCA was mentioned, it was in fact 75 ohms.
I wouldn't have thought a company with the credibility (and widespread usage within the industry) of Canare would be playing loose with such terms.
I have an Apogee WydEye digital interconnect I used for 7 years, until the center pin on the Canare "75 ohm" RCA fell off.
All that time I believed they were true 75 ohm connectors. They were sold as such.

They still are, really. Is not the page linked to titled "75 ohm RCA crimp plug"?
If it does not have a true 75 ohm characteristic, then this is intentional obfuscation, and I think it is intentionally misleading and should'nt be allowed, marketing notwithstanding.

I'm going to look into this next Wednesday.

Bob Reynolds · « **Reply #14 on:** 1 Jul 2006, 02:13 pm »

Hi Russell,

From the Canare catalog (pdf) that I downloaded from their web site, the table of contents uses the same nomenclature: 75ohm Type RCA Crimp Plugs, 75ohm Type RCA Bulkhead Receptacles.

I'm sure their legal department has signed off on the wording. From the specs comparing the true 75ohm BNC connector to the 75ohm type RCA connector you get an indication that something is amiss:

BNC: DC to 2GHz; >=26dB Return Loss (<=1.1 VSWR)

RCA: DC to 200MHz ; >26dB Return Loss (<=1.1 VSWR)

I think the factor of 10 in the bandwidth tells the story.

I also see a Note in the BNC section that says that an impedance mismatch between older 50ohm BNC and newer 75ohm BNC should not be a problem with NTSC video frequencies, but will be an issue at higher analog and digital video frequencies.

So I guess I can accept the RCA connectors on my analog component video cable and on digital audio cables, though I would prefer BNC. Obviously it makes no difference at all at analog audio frequencies.

Does this make any difference with the new class D switching amps? Might be better to use STP and XLR. I'm a huge fan of balanced lines anyway.

-- Bob

Dan Banquer · « **Reply #15 on:** 1 Jul 2006, 04:26 pm »

"RCA: DC to 200MHz ; >26dB Return Loss (<=1.1 VSWR)"

With spec's like this for SPDIF there should be no issue. As Jocko Homo pointed out in the Lab forum, the real problem is not connectors or 75 ohm coax, the real issue typically is the driver from the source unit.
d.b.