[LAU] Behringer and Linux

[Chris Caudle]

On Wed, April 1, 2015 2:03 am, Patrick Shirkey wrote:
> So you are saying that Behringer manufactured and released
> an entire range without testing *any* of them before they
> went out the door for ground loop issues at the board/design level?

Something much more limited and specific:  I am saying that the
description provided by Fons of the behavior of that specific model of low
cost mixer is consistent with a "pin 1 problem" design flaw, although that
is a bit of a misnomer with a USB connector.  "Pin 1 problem" is much
easier to say than "reference conductor common mode impedance noise
coupling" though.  Possibly only affecting the headphone output, Fons
never mentioned whether he also checked the main or monitor outputs.

> Seem pretty unlikely even for cheap low end manufacturing from China.

Where the unit was manufactured has no bearing on where it was designed,
and there are examples of pin 1 problems even in expensive equipment. 
With proper PCB layout noise from the computer chassis should not be a
problem.  The problem and how to avoid it were popularized in the June
1995 Journal of the Audio Engineering Society, they devoted an entire
issue to grounding and shielding problems, and Jensen and Rane have also
spread the word through several white papers and tutorials.  Many seminars
have been taught by Bill Whitlock of Jensen and Jim Brown of Audio Systems
Group consulting among others.  This is not something esoteric.

> More likely this one just slipped through the (random) QC process

As Fons pointed out, the typical design style for that class of equipment
has all connectors soldered to a single PCB, and most of the assembly of
the PCB is performed by automated equipment.  The balance of likelihood
between an assembly flaw that allows the equipment to work, but not work
fully properly, and a design flaw comes down on the side of design flaw
for this particular behavior.  The solution is relatively straight
forward, but you have to fight against the easy way of doing things using
most PCB layout software, so it is commonly not done correctly.

> From his description it sounded like the device was working
> pretty good with Linux in every other way. Just a bit of
> hum at the hardware level.

Indeed, the improved standardization of USB interfaces has been very good
for linux audio.  It would be interesting to see if the problem affected
every unit or was a manufacturing flaw as you speculated, and whether the
problem affects all the outputs or only the headphone amp.

Fons, did you happen to get the specific model number of that mixer?

-- 
Chris Caudle