[mythtv-users] Slightly OT: receiver input buzzing/humming

f-myth-users@media.mit.edu f-myth-users at media.mit.edu
Thu Jan 12 00:08:33 UTC 2006


    Date: Wed, 11 Jan 2006 17:11:10 -0600
    From: matthew.garman at gmail.com

    On Wed, Jan 11, 2006 at 04:58:36PM -0500, f-myth-users at media.mit.edu
    wrote:
    > Eh?  If he's sending digital information down the coax, then no
    > amount of ground hum will be audible at the receiver, until the
    > hum is so bad it starts flipping bits in the data stream (at which
    > point, he'll hear either silence or godawful artifacts, depending
    > on the protocol---but it sure won't be 60 Hz hum).

    How bad is "bad"?  :)  I mean, there's definately noise between the
    computer and the receiver.  Will I hear random bits being flipped (I
    assume that noise will cause SOME flippage)?  Although I suppose
    there's some kind of parity or error correction built in to the
    protocol...

If the hum is enough to flip bits (and I doubt it would be), you'd
either hear nothing at all (if the protocol suppresses packets with
errors) or perhaps fuzz (like clipped square waves or a fuzzbox on a
guitar) or perhaps white noise (hiss).  Or all kinds of other things.
But it's very unlikely to sound like hum.  If you take a random MP3
(say) or WAV file and start flipping 1% of the bits, what do you hear?

    > supposedly the -only- source of ground interconnection, then going
    > to an optical interconnect might help, but that's a fragile
    > solution.

    What makes it fragile?  Is it because of the point above, that I
    might be close to enough bit flipping that I hear garbage?

It's fragile because if the only thing keeping the system working is
keeping two dissimilar grounds separate, then all it takes is one wire
between two components that didn't used to be connected to possibly
connect those grounds and give you your hum back.  That might amount
to adding hardware, or changing what outlet something's plugged into,
or even someone touching the cases of both components simultaneously.
There are loads of sneak paths, and it only takes one.  Better to
eliminate the problem at the source than depend on isolating grounds.

    > It's ambiguous from the original poster's comment whether the
    > problem is new, or he's just started noticing it.  My guess, if
    > the latter, is that the shield on some cable got damaged, or
    > something was recently changed in the hardware configuration
    > elsewhere (new component? new cable?), and certainly replacing

    I *think* it's new.  If it's not, then it's definately worse than it
    used to be.  And yes, there were *major* changes---I got a new TV,
    and got rid of the VCR and DVD (mythtv makes them obsolete).

Then lots of things might have changed in your cabling, and maybe a
bad shield that was papered-over by some -other-, good shield (hence
good ground) being connected between two things isn't connected any
more---so maybe the problem is that you used to have a better ground
connection between one place and another, and now you don't.

Or maybe the physical strain of unplugging a cable or plugging it back
in broke a marginal solder joint in a shield somewhere.  You see the
point---you're going to have to start swapping cables & so forth
around and see if you can get the hum to move around or vanish.

(You might also try turning off the TV and seeing if you still hear
hum from the stereo.  Instead of being induced ambient powerline hum,
you -could- be getting hum from the flyback transformer or something
else coupling into a signal line somewhere that runs too close to the
TV (and has a bad shield).  This used to be an issue with power supply
transformers, but modern electronics doesn't use transformers in its
power supplies any more.  And if this TV is in fact not a CRT, there's
no flyback, so that's not it either...)

    > cables is the easier & quickest way to debug that.  If he has
    > another way of producing audio, I'd also try substituting that at
    > the end of the cable, once the cable is known good.  (After all,
    > it -could- be that some filter cap in the sound board on the
    > computer blew or is getting leaky; such things do happen.)

    Hmmm... it seems like my sound card is unusually quiet.  Part of the
    reason I even noticed this problem is that I have to turn up the
    receiver louder than I do for anything else when using my HTPC
    (mythtv, xmms, whatever).  Even with PCM cranked, I still have to
    turn the volume more than I usually do.

It's possible you're just hearing hum from the soundcard that no
amount of cabling can get rid of---the card itself just might be
hummy, and you never noticed it at lower volumes.  I'd try swapping
cables first, then any other sound output your machine can generate,
including perhaps a different soundcard if you don't have any other
audio outputs available.  (You might also make sure that aslamixer
claims that all your masters are up, and so forth.)

    > (I've seen strange currents induced on so-called power grounds
    > when different phases meet through equipment, especially in older
    > structures with poor wiring [which is one reason why certain lab
    > and audio gear keep chassis ("power") and signal grounds
    > rigorously separate]

    In my embedded systems class in college, my professor had this great
    analogy of why you ground signal and power separately: think of
    indoor plumbing, and how the toilet output is generally routed
    separately from every other drain.  In the case of a backup, you
    DON'T want stuff from the toilet coming up!

Yeah, that's a pretty good way to think about it... :)  But most
consumer-grade stuff isn't as rigorous about that as you might like,
and in most cases it doesn't matter, anyway.  Where it matters is when
you're trying to measure microvolt quantities at high impedances,
probably in the presence of either motors or high-speed electronics.


More information about the mythtv-users mailing list