[David Gravereaux]

The AES3 standard (AES-2id-2006) for AES/EBU digital transmission of 2-ch audio
states in section 7.1.1.3 that the cable is supposed to be grounded at both ends
(aka a ground loop).

Yet the standard also says that the transmitter and receiver are based on the
RS-422 hardware UART.

According to documents about RS-422 hardware
http://www.bb-elec.com/tech_articles/rs422_485_app_note/transient_protection.asp#ground,
ground is only a reference for the single-ended signals and not to be made common
between end-points. Also, shield of the cable is lifted at the modem side
http://www.interfacebus.com/Design_Connector_EIA530.html.

The electronics themselves can only handle a common-mode voltage between -5V and
+12V. And best isolation is achieved with opto-couplers (LED-opto transistor).
The AES standard does mention the use of transformers, but digital (ie. 1 and 0
levels of a carrier) doesn't need an "analogous" coupling.

Bill Whitlock of Jensen Transformers @ the SF AES show last October did mention in
his talk the importance of attention to ground loops in data hook-ups. So how is
AES/EBU different than RS-422, aside from half-duplex?

Hmmmm... Is it possible the AES standard is not exclaiming the best manner of
realization? It certainly is worth asking about.

--
"That's the whole problem with science. You've got a bunch of
empiricists trying to describe things of unimaginable wonder."
--- Calvin


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (MingW32)

iD8DBQFGKHsUlZadkQh/RmERAnTKAKC8o7eyXYK+0EUG/cbAZl3vEnBItgCgoACv
BGzMxYzta9ILj7DCu6hAAak=
=JRMJ
-----END PGP SIGNATURE-----

[Aaron J. Grier]

David Gravereaux wrote:
The AES3 standard (AES-2id-2006) for AES/EBU digital transmission of
2-ch audio states in section 7.1.1.3 that the cable is supposed to be
grounded at both ends (aka a ground loop).

that's funny, I don't find the word "ground" once in the standard.

maybe you are referring to a shield?

I think you might want to re-read section 8 "electrical requirements".

--
Aaron J. Grier | "Not your ordinary poofy goof." |
The United States is the one true country. The US is just. The US
is fair. The US respects its citizens. The US loves you. We have
always been at war against terrorism.

[Mike Rivers]

On Apr 20 (Geez, was it THAT long ago?), 4:34 am, David Gravereaux
wrote:
The AES3 standard (AES-2id-2006) for AES/EBU digital transmission of
2-ch audio
states in section 7.1.1.3 that the cable is supposed to be grounded
at both ends
(aka a ground loop).

It's only a ground loop if there's another path to complete the loop.
Connecting the cable shield at both ends is good practice because it
doesn't leave a hole in the shielding for EMI to get in or out. The AES
standard also calls for transformer isolation on both the receiver and
transmitter (though not every manufacturer complies with this) so the
data signal is independent of chassis ground (where the shield normally
goes).

Yet the standard also says that the transmitter and receiver are
based on the
RS-422 hardware UART.

According to documents about RS-422 hardware

http://www.bb-elec.com/tech_articles/rs422_485_app_note/transient_pro...,
ground is only a reference for the single-ended signals and not to be
made common
between end-points. Also, shield of the cable is lifted at the modem
side
http://www.interfacebus.com/Design_Connector_EIA530.html.

There's no modem in a digital audio interface. That's to keep telephone
company **** away from the electronics. If you're talking about an A/D
or D/A converter with analog connections as well as digital, you may
want to lift the shield on one end of the analog side, but still, it's
better if you don't have to do that.

Hmmmm... Is it possible the AES standard is not exclaiming the best
manner of
realization? It certainly is worth asking about.

There are a lot of compromises that a manufacturer must consider when
trying to make a piece of equipment that will interface with a different
manufacturer's equipment, particularly when one (or both) manufacturers
is trying to cut every corner possible to keep the price down. If price
was not an object, then everyone could follow the same interface
standard. When the first guy says "I don't need a transformer on the
input because the the other guy is putting one on his output" that's
OK. But if the second guy, at the same time, says "I don't need a
transformer on the output because there will be one on the other guy's
input" then you got trouble when you try to hook those two devices
together.

Tell your vendors that you'll happily pay an additional $100 per input
and output if they'll only build the gear right. Or only buy gear that
does it right. But this means you have to do more homework. Sometimes
it's worth it, sometimes you can get away with cheating and buy a better
mic or use a better grade coffee in the studio with the money you save.