Balanced input

xcortes · 4084

0 Members and 1 Guest are viewing this topic.

Offline xcortes

  • Sr. Member
  • ****
    • Posts: 997
on: March 17, 2011, 02:11:46 PM
I'm in the process of adding digital to my system. Nope, not BH's DAC (yet) but a beautiful Pacific Microsonics Model Two. In the past digital has not cut it for me so this time I decided to do an all-assault effort. The digital part is yet undecided in terms of platform, card, playing software, etc although it looks like I'll be using a very high end card with pro software running on XP (who would have said that after all this years I would buy a Windows machine for my house!).

The PM2 outputs are balanced with XLR connectors, pin 2 or 3 high (software selectable) -12dBu to +24dBu full scale, 20 ohms impedance.

Eventually I'll be building a new preamp but right now I need to put this baby to work with my EFP3.

The manual says the following: If the audio input from the Model Two must drive an unbalanced load, use pin 2 as high, pin 1 as ground, and lift pin3 at the load or use pin 3 as high, pin 1 as ground, and lift pin 2 at the load. CAUTION: Never connect the unused low pin to ground, (pin1) as that will short circuit one side of the Model Two's actively balanced, DC coupled, output amplifier, and seriously degrade performance.

So the manual gives me one option to connect it. But I'm wondering if using transformers wouldn't be better.

Jensen Transformers sells the PC-2XR which is designed just for this application. It has a 4:1 transformer with the primaries connected to pins 2 and 3 while pin 1 goes to ground. On the secondaries side one side is hot and the other ground:

http://www.jensen-transformers.com/datashts/pc2xr.pdf


That's one option. Another option would be to build my own devices. One alternative is using a pair of Harman Kardon XT-3 transformers that I have on hand. These are 1:10  input transformers and have a 500/600 ohm CT primary with a 50,000 ohm single ended secondary. I would connect it as the Jensen with the 50K side as a primary and the 500 one as a secondary, maybe loaded with a 600 ohm resistor?

What about a pair of B7 15K:600 in the same configuration? These are earmarked for my big preamp project but I could use them in the meantime.

BTW, I'm using 33K resistors on the FP3 shunting scheme.

What would be my best option?

Thanks

Xavier Cortes


Offline JC

  • Sr. Member
  • ****
    • Posts: 485
Reply #1 on: March 17, 2011, 04:43:05 PM
On the output, which pin is the third-wire ground?  IOW, which pin would be used for shield for the balanced pair?  I would be tempted to use that shield pin for "ground" at the unbalanced end, then use whichever side of the balanced line you wish for "hot" at the unbalanced end.  As the manual says, leave the remaining half of the pair floating.

My reasoning is that you may want to use either side of the balanced pair as the unbalanced "hot" depending on whether or not a signal inversion would be useful or not.  Leaving the other line of the balanced pair unconnected, like the manual says, keeps you from running the risk of burning up an amp on the balanced output.
Using the "shield" connection as the signal ground at the unbalanced end mitigates against there being an issue between safety grounds between the two pieces.  Having said that, I think I would follow the manual for starters, since presumably they know their gear.

Personally, I always find that a piece of wire sounds better than a transformer wherever possible.  The reason to use a transformer is if you needed to keep the balanced line in tact, such as in the case of a long interconnect between the two pieces or a particularly noisy environment for the interconnect.

Just my 2 cents, though.

Jim C.


Offline Grainger49

  • Hero Member
  • *****
    • Posts: 7175
Reply #2 on: March 18, 2011, 02:00:14 AM
If I understand balanced pin 1 is the ground/shield for the standard balanced output.  That can be used for the "common" of an unbalanced.  The balanced signal gives you the possibility of an absolute phase correct (pin 2) or absolute phase inverted (pin 3) output from your balanced output.  I would put in a switch that gave me the ability to change according to the musical source as I mark music either positive or negative.



Offline Paul Joppa

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 5847
Reply #3 on: March 18, 2011, 05:57:56 AM
The main reason for using balanced interconnects is to reduce common-mode or ground-current noise. It is possible that the proposed DAC will have some ground noise, especially if it shares a ground with a computer - but if it does not, then the balanced line isn't doing anything for you.

A step-down transformer also allows you to run the DAC outputs at a higher signal level, reducing the impact of noise which is usually independent of signal level. Again, if noise is not a problem then that advantage is moot.

The thing about transformers is that they are designed for specific situations. For example, I recently ran some simple tests on an Edcor line stage output transformer, 15K:600 ohms. Measured performance was quite good with a source impedance of 8K to 10K, but it had a lot of ringing if the impedance was lower than that. Another thing is that a transformer designed for step-down will not in general work as well when reversed as a step-up. Finally, transformers are usually designed for either single-ended or balanced operation on each winding; this is especially important for the higher impedance winding.

We have listened to a few well-designed transformers with excellent measurements, from the likes of Jensen and CineMag for example, and more often than not found comparable classic designs that sounded better to our ears.

The bottom line for me is that I would want to listen to a variety of units before committing to a particular transformer for a particular application, if I wanted the highest possible performance. The Jensen PC-2XR has the virtue that the transformer inside is designed for this application, and it's a ready-built unit with many other potential uses. The other transformers you mentioned are not optimized for this application, and their measured performance would almost certainly fall short of the Jensen. Whether other factors might make them sound better or worse overall is hard to say. It would be an interesting set of experiments to compare all four options (including the balanced-to-single-ended cable as specified by the manufacturer) - but I would not attempt to predict which would come out on top.

Paul Joppa


Offline xcortes

  • Sr. Member
  • ****
    • Posts: 997
Reply #4 on: March 18, 2011, 08:56:59 AM
Thanks all,

I guess I'll just make me a pair of XLR to RCA cables with pin 2 as "hot" and pin 1 as "common". And maybe later start playing with the transformers.

Grainger, the DAC allows to select which pin is "high" so no need to add an additional switch!


Xavier Cortes


Offline Grainger49

  • Hero Member
  • *****
    • Posts: 7175
Reply #5 on: March 18, 2011, 09:04:56 AM
Cool!  I have had two DACs that allowed output to be inverted, my current one, Ack! dAck!, doesn't.



Offline ironbut

  • Sr. Member
  • ****
    • Posts: 470
Reply #6 on: March 20, 2011, 07:43:35 PM
Hey, hey Xavier!

Moving forward with your own digital project I see.
IIRC, Larry Toy's Model 2 needed to have a breakout cable built for it. I think that Tim Marutani built it or had it made to get the best interface with Larry's system.
If you haven't contacted Tim about your PM2, you should.
« Last Edit: March 21, 2011, 05:12:17 AM by Doc B. »

steve koto