"Proper" XLR to RCA cable configuration?
Sep 30, 2007 at 2:51 AM Thread Starter Post #1 of 13

sejarzo

Headphoneus Supremus
Joined
Jan 1, 2006
Posts
1,964
Likes
21
Location
Indiana
I recently acquired a mint Marantz Pro PMD321 CD player with strictly balanced XLR outputs for donation to a high school band program, thinking it would be connected to a balanced input on a mixer. That is not the case now though, as it must be connected to an unbalanced input on a preamp.

There are trim pots on the outputs that allow adjustment of the output level down to something equivalent to consumer -10dBu level. I played a CD with a 0 dBFS 1 kHz tone cut, and found that I could easily adjust the hot-to-ground down to 2.0 Vrms.

I went ahead and built a pair of cables that left pin 3 (cold) on the XLR open, tied the shield and the blue conductor to pin 1 (ground) and the white conductor to pin 2 (hot) on the female XLR. I then tied the blue to the ground of the RCA and the white to signal on the RCA; the shield floats on the "RCA end".

It worked just fine driving my CK2III on a short test, but later I found a recommendation on the Rane site that the preferable method is to tie the shield to the chassis ground on the XLR, the "hot" to the signal on the RCA, the "cold" to the normal ground on the RCA, and float the shield on the RCA end.

I have verified continuity between the pin 1 on the XLR output and a chassis screw on the player. Also, I verified that I can turn down those pots even more to obtain 2.0 Vrms between the hot and cold pins on the player output with the 0 dBFS test tone.

Will I damage the cold side output by leaving it run open with the cable as-is, or should I rebuild it as Rane suggests with no connection to pin 1 on the XLR end, or should I resolder the XLR end with shield to pin 1, white to pin 2, and blue to pin 3 and adjust the output down again?

Thanks for your advice on this one--there are apparently two schools of thought out there on the best way to handle this situation!
 
Sep 30, 2007 at 3:24 AM Post #2 of 13
You will not damage anything by leaving the OUT- disconnected. The advantage of using OUT+ and OUT-, is you get double the voltage swing.

Where you could run into trouble using OUT- is if you use a piece of equipment downstream that shares ground between left and right channels (not uncommon). In this scenario, you will be shorting the OUT- of each channel together (generally bad).

I would stick with it the way you have it, and adjust your gain accordingly. You lose 6dB without OUT-, but that's easy to compensate for in a pre, or it sounds like, by adjusting the output pots on the player.
 
Sep 30, 2007 at 3:40 AM Post #3 of 13
Here you cna get some ideas...
 
Sep 30, 2007 at 3:55 AM Post #4 of 13
Thanks Brian, for a great explanation! I hadn't thought about the fact that a downstream component could tie the colds together, which is obviously not a good thing even to this non-EE.

Losing 6 dB really doesn't hurt, as the hot-to-ground or cold-to-ground levels were over 4 Vrms when I received the unit. All the other gear that will be connected is consumer-level output, so my existing solution seems to be just fine.

I was pleasantly surprised that it was easy to adjust the trim pots down to get 2.0 Vrms--I was afraid that they would truly be "trim only" within a narrower range, and wouldn't allow the output voltage to be dropped so low.

One of the pages by Rane almost disagrees within itself on the "right" way to do this. One figure suggests leaving pin 3 open when using a single-ended coax cable to the RCA, but another figure for use of 2-conductors-plus-shield cable clearly shows tying XLR cold to RCA ground.
 
Sep 30, 2007 at 6:22 AM Post #5 of 13
I am highly impressed with your skills...cool! I have yet to dig too far into balanced lines, even though my big rig downstairs would allow me to.

You might consider looking into the Cardas adapters which would allow you to use your present single-ended cables. They are all slightly different than each other, but it might be an elegant way to attack.
 
Sep 30, 2007 at 12:59 PM Post #6 of 13
Quote:

Originally Posted by bhjazz /img/forum/go_quote.gif
You might consider looking into the Cardas adapters which would allow you to use your present single-ended cables. They are all slightly different than each other, but it might be an elegant way to attack.


Or since its for a high school band, he could keep it under $4 for the pair with these: http://www.pulsartech.net/product/PA...)_adapter.html
 
Sep 30, 2007 at 1:32 PM Post #7 of 13
I previously tried the pin 2-to-RCA signal & pin 3-to-RCA ground (no bridging of pin 3 to pin 1) method, and I suffered from intermittent line noise as well as too high an input volume (just a little volume pot movement sent volume sky-high already). I second the recommendation to stay with what you have done originally (pin 2 to RCA signal, pin 1 to RCA ground, shield to pin 1 and floating on RCA end, and pin 3 unpopulated). This is the exact same configuration I am using for my own XLR-to-RCA cables.

Cheers!
 
Oct 1, 2007 at 2:13 AM Post #9 of 13
Quote:

Originally Posted by eddiewalker /img/forum/go_quote.gif
Or since its for a high school band, he could keep it under $4 for the pair with these: http://www.pulsartech.net/product/PA...)_adapter.html


Quote:

Originally Posted by bhjazz /img/forum/go_quote.gif
....You might consider looking into the Cardas adapters which would allow you to use your present single-ended cables. They are all slightly different than each other, but it might be an elegant way to attack.


But none of those show how the pins on the XLR are connected to the contacts on the RCA.......and that is really the gist of the problem. You can probably get sound out of the unbalanced component with them, but the big question is whether that adapter is actually wired correctly for the situation. There is definitely no universal agreement, even in pro audio, as to how signal and chassis grounds are connected.....which is why there are all sorts of problems with hum at times when new gear is hooked up.

Quote:

Originally Posted by Zorander /img/forum/go_quote.gif
I previously tried the pin 2-to-RCA signal & pin 3-to-RCA ground (no bridging of pin 3 to pin 1) method, and I suffered from intermittent line noise as well as too high an input volume (just a little volume pot movement sent volume sky-high already). I second the recommendation to stay with what you have done originally (pin 2 to RCA signal, pin 1 to RCA ground, shield to pin 1 and floating on RCA end, and pin 3 unpopulated). This is the exact same configuration I am using for my own XLR-to-RCA cables.

Cheers!



The nice thing about my situation is that the output level from the PMD321 is variable all the way down to nearly 0 Vrms, so it was just a question of playing a 0 dBFS test tone and dialing in the right level on my DMM.

Quote:

Originally Posted by BrianDonegan /img/forum/go_quote.gif
It's likely your Rane unit does not share grounds between channels, which would explain their recommendation.


Oh, I really don't own any Rane gear at all. It's just that their technical reference info is normally so well executed that I, as an engineer (but non-EE) can easily understand the issues discussed and why they recommend various solutions. The pages to which I referred are:

"Grounding and Shielding Audio Devices":

http://www.rane.com/note151.html

and "Sound System Interconnection":

http://www.rane.com/note110.html

which is the one that shows how to construct cables for various combinations of "unlike" gear!

Scroll down about halfway, look at that table of various connectors and a bit further down, the diagrams of how to build the cables. Excellent explanations throughout both those pages about how various grounds differ and why they have suggested a somewhat different way to connect fully balanced gear via XLR's.
 
Oct 1, 2007 at 2:28 AM Post #10 of 13
Ah, that explains it. If you look closely at the picture, the XLR to RCA connection (XLR source) has OUT+ connected to IN+ and OUTgnd to INgnd. For the opposite direction (RCA source) they show OUT+ connected to IN+ and OUTgnd to IN-. This is correct. Typically when feeding balanced gear from a single-ended source, you would tie IN- to GND at the input.
 
Oct 1, 2007 at 2:52 PM Post #11 of 13
Yes, I agree.

What they don't show, though, is how to use standard shielded balanced cable (twisted pair plus a shield) to route the signal from a balanced XLR output to an unbal RCA--they only show the single internal conductor plus shield version. I have all available documents for the PMD321, but they don't identify the type of balanced output circuit used in the unit, so I presumed it best to tie the shield to ground at the XLR and use the other conductor in the twisted pair for signal ground.
 
Oct 1, 2007 at 3:06 PM Post #12 of 13
I think you could probably safely leave the shield out of the equation all together unless you end up with a bunch of noise on the line. I have yet to experience a difference between shielded and unshielded connectors, but maybe I'm just lucky.
 
Oct 1, 2007 at 3:18 PM Post #13 of 13
I agree, Nate, for home systems that the shield seemingly offers little benefit.

My concern is that there will be all sorts of cabling in a rather small lockable 10U or 14U rack enclosure for security reasons.....power strip, CD player, cassette player (band contest judges still use cassette for comments), patch panel, small balanced preamp, an eq, and power amp (if we don't end up with powered speakers.)

Lots of gear, PSU's, and interconnects and power cords very close together.....and it seems that the differences in opinion between balanced gear manufacturers about how to connect signal and chassis grounds often causes issues.

The gist of one reference is that, while unbal connections are technically not so great in a noisy environment, those inconsistencies can cause noise nightmares even in totally balanced systems. This one will have to be assembled from a mix of gear due to cost considerations.
 

Users who are viewing this thread

Back
Top