Spontaneous phase inversion?!

[Binba]

My good ole' mini stereo amplifier started to behave extremely weird
recently. Beyond not being able to use it, I'm really puzzled with the
symptoms:

There's a strong reverb effect and the vocals are canceled and slightly
to the right - seems like a phase shift. I put on a Beatles song to
better test it:
- The vocals are canceled even when I hold my ear against one speaker,
or when I flip the polarity for one of the speakers.
- I hear both channels in each speaker, and each channel on both
speakers.
- With balance full left, the left speaker sounds slightly louder, with
full right it sounds centered.
- I tried the two available inputs ("DAT" and "AUX"). Same thing.

I opened up the unit, and as far as I can tell, everything looks fine.
Nothing burnt or melted, no excessive dust (and they were smart enough
that no board is facing up). I vacuumed it just to be sure, didn't
change anything. You can see snapshots at
http://www.flickr.com/photos/31278470@N00/sets/72157594269637399/

The unit is a Sony MHC-1600. Its amp, if that means anything, is
STK4132II. The only thing I did to it was to move it to another room
for one night.

So my only conclusion is that the unit spontaneously crosses (mixes)
both channels, with inversed phase, before splitting it back to the
stereo amp. How the hell can that happen, not to mention how to fix
it?!

 

[N Cook]

My good ole' mini stereo amplifier started to behave extremely weird
recently. Beyond not being able to use it, I'm really puzzled with the
symptoms:

There's a strong reverb effect and the vocals are canceled and slightly
to the right - seems like a phase shift. I put on a Beatles song to
better test it:
- The vocals are canceled even when I hold my ear against one speaker,
or when I flip the polarity for one of the speakers.
- I hear both channels in each speaker, and each channel on both
speakers.
- With balance full left, the left speaker sounds slightly louder, with
full right it sounds centered.
- I tried the two available inputs ("DAT" and "AUX"). Same thing.

I opened up the unit, and as far as I can tell, everything looks fine.
Nothing burnt or melted, no excessive dust (and they were smart enough
that no board is facing up). I vacuumed it just to be sure, didn't
change anything. You can see snapshots at
http://www.flickr.com/photos/31278470@N00/sets/72157594269637399/

The unit is a Sony MHC-1600. Its amp, if that means anything, is
STK4132II. The only thing I did to it was to move it to another room
for one night.

So my only conclusion is that the unit spontaneously crosses (mixes)
both channels, with inversed phase, before splitting it back to the
stereo amp. How the hell can that happen, not to mention how to fix
it?!

I would suggest your speakers are now wired
L-C-R
where C is a common connection but not connected to ground G (black)
L-GG-R
properly

 

[[email protected]]

I would suggest your speakers are now wired
L-C-R
where C is a common connection but not connected to ground G (black)
L-GG-R
properly

the other possibility is an o/c connection on the balance control

NT

 

[Binba]

o/c as in open-circuit? can you elaborate?
The source is a laptop, I tested it with other amp systems and the
balance control works perfect, so this part of the chain is functioning
properly. The Sony amp doesn't even have balance controls...
If you know these units, they're very simple - no settings to spoil.

 

[Binba]

Interesting... I'm still wondering how something like that can just
happen.
And if I'm trying to think of a workaround...should I try to ground the
speakers? Perhaps run a jumper wire from one (--) (black) terminal to
the unused AM antenna's GND? Or is that just gonna cause one big
short-circuit...?
The amp's mains connection is 2-prong.

 

[Jim Land]

Interesting... I'm still wondering how something like that can just
happen.
And if I'm trying to think of a workaround...should I try to ground the
speakers? Perhaps run a jumper wire from one (--) (black) terminal to
the unused AM antenna's GND? Or is that just gonna cause one big
short-circuit...?
The amp's mains connection is 2-prong.

First thing to do is *test* to see if this is the case.

Disconnect the G leads (both speakers) from the unit, tie them together,
and then temporarily connect them to any known ground on the unit (such
as the one you mention).

Listen to tell whether the problem goes away. If it does, then you can
open the unit and figure out why the G speaker connectors aren't
connected to ground, and fix it.

If the problem is still there, this is not the solution. Reconnect both
G speaker leads to the speaker connectors on the unit. Continue looking
for cause elsewhere.

 

[John]

You could try turning the karaoke switch off!
John.

 

[Mr.T]

Disconnect the G leads (both speakers) from the unit, tie them together,
and then temporarily connect them to any known ground on the unit (such
as the one you mention).

Listen to tell whether the problem goes away. If it does, then you can
open the unit and figure out why the G speaker connectors aren't
connected to ground, and fix it.

If the problem is still there, this is not the solution. Reconnect both
G speaker leads to the speaker connectors on the unit. Continue looking
for cause elsewhere.

I'm not familiar with the particular unit, but I wouldn't attempt this if it
uses bridged amplifier outputs like some do.
There aren't any "G" connectors in that case of course, but then most amps
don't have terminals actually marked "G" in any case.

MrT.

 

[Binba]

Actually, I found the whole datasheet for the amplifier (unless there's
more than one STK4132II amp in the world), the system is a Sony but the
amp is made by Sanyo...whaddyaknow.
http://service.semic.sanyo.co.jp/semi/ds_e/ENN2668A_d.pdf#search="STK4132"

From my basic electronics knowledge, it doesn't look like it's a

bridged amp.
But if someone is actually interested in examining the circuit diagrams
and verifying, that'll be cool. Maybe you could spot some design flaw
that'd explain this behavior

Thanks for the replies, I'll try the grounding workaround.

-D

 

[Binba]

And yes, there's a chasis ground terminal, actually 2, for the AM+FM
antennas.

Oh and it totally sounds like the kareoke switch is on... except that
there's no such feature in this system

 

[[email protected]]

o/c as in open-circuit? can you elaborate?

if you disconnect the balance pot's wiper you get mono output, in
phase.

- With balance full left, the left speaker sounds slightly louder, with
full right it sounds centered.

The source is a laptop, I tested it with other amp systems and the
balance control works perfect, so this part of the chain is functioning
properly.

Someones sure confused.


NT

 

[Arfa Daily]

And yes, there's a chasis ground terminal, actually 2, for the AM+FM
antennas.

Oh and it totally sounds like the kareoke switch is on... except that
there's no such feature in this system

The problem will almost certainly be that the grounds to the speaker " black
" connectors have gone missing - it's a common problem. The really big clue
to this is that the vocals disappear on a Beatles recording. Most of their
stuff was produced in the early days of stereo recording, and was
studio-made rather than genuinely recorded with multiple mics. The vocals
were almost always just a mono-split across the two channels, to help with
keeping them 'bright' and defined with mono pickup cartridges that were
billed as " stereo compatible ". The effect of this when you lose the ground
to the speakers is just as Cook-y from Diverse Electronics said at the top
of the thread - you get almost total phase cancellation of the vocals due to
the L-C-R connection that's left when the ground goes ( it's almost like
having the two speakers connected in series across a bridge connected amp,
but with each half of the amp being driven by different signals ). Most
likely bet is a bad joint or cracked print right behind the connector block.

Arfa

 

[N Cook]

The problem will almost certainly be that the grounds to the speaker " black
" connectors have gone missing - it's a common problem. The really big clue
to this is that the vocals disappear on a Beatles recording. Most of their
stuff was produced in the early days of stereo recording, and was
studio-made rather than genuinely recorded with multiple mics. The vocals
were almost always just a mono-split across the two channels, to help with
keeping them 'bright' and defined with mono pickup cartridges that were
billed as " stereo compatible ". The effect of this when you lose the ground
to the speakers is just as Cook-y from Diverse Electronics said at the top
of the thread - you get almost total phase cancellation of the vocals due to
the L-C-R connection that's left when the ground goes ( it's almost like
having the two speakers connected in series across a bridge connected amp,
but with each half of the amp being driven by different signals ). Most
likely bet is a bad joint or cracked print right behind the connector block.

Arfa

I think this was the weirdest sympom I ever saw.
I thought the customer was winding me up initially.
Kit was a play only car cassette player and whatever tape you put in it ,
self recorded or shop bought, it recorded the sound of a cracked vinyl
record over the music, permanently.

If no one gets the answer I will post back in a few days time.

 

[Arfa Daily]

I think this was the weirdest sympom I ever saw.
I thought the customer was winding me up initially.
Kit was a play only car cassette player and whatever tape you put in it ,
self recorded or shop bought, it recorded the sound of a cracked vinyl
record over the music, permanently.

If no one gets the answer I will post back in a few days time.

There used to be a Panasonic answer machine that used to do the same thing
in that it recorded a rythmic "crack - crack - crack " when certain tape
types ( mostly of their own manufacture, oddly ) were used. Problem was
caused by a static build up on the head's surface from the passage of the
tape over it. The head as I recall, was mounted on plastic, so 'floating'.
This may not be exact as it's been some years ago, but something along those
lines anyway. Cure was to add a ground wire to the head. Something similar
to this perhaps ??

Arfa

 

[N Cook]

There used to be a Panasonic answer machine that used to do the same thing
in that it recorded a rythmic "crack - crack - crack " when certain tape
types ( mostly of their own manufacture, oddly ) were used. Problem was
caused by a static build up on the head's surface from the passage of the
tape over it. The head as I recall, was mounted on plastic, so 'floating'.
This may not be exact as it's been some years ago, but something along those
lines anyway. Cure was to add a ground wire to the head. Something similar
to this perhaps ??

Arfa

A tiny piece of magnetic material squashed into the rubber of the pinchwheel

 

[Mike]

The problem will almost certainly be that the grounds to the speaker " black
" connectors have gone missing - it's a common problem.

Also note that the *inputs* to the amp can cause a similar problem, I repaired
an amp that was doing this intermittently. It turned out the phono-RCA-jacks
*in* to the amplifier had no ground connection. This was due to them being
dual phono sockets, stacked vertically, with no physical support other than
the solder joints (not good), coupled with super tight fit, gold plated,
heavy cable interconnects -- the insertion force required to get the line-in
to go in was such that it broke the amp input.

likely bet is a bad joint or cracked print right behind the connector block.

Either at the input, or output

 

[Arfa Daily]

A tiny piece of magnetic material squashed into the rubber of the
pinchwheel

Yep ! That's a good one ...

Arfa

 

[Arfa Daily]

Also note that the *inputs* to the amp can cause a similar problem, I
repaired
an amp that was doing this intermittently. It turned out the
phono-RCA-jacks
*in* to the amplifier had no ground connection. This was due to them being
dual phono sockets, stacked vertically, with no physical support other
than
the solder joints (not good), coupled with super tight fit, gold plated,
heavy cable interconnects -- the insertion force required to get the
line-in
to go in was such that it broke the amp input.


Either at the input, or output

Hi Mike

This is absolutely true, as a general rule. I too have seen this many times
on even high end Yammies and Teacs. I assumed it not to be the reason in
this case, because the unit has a built in CD, which I took to be the OP's
source when playing his test Beatles track. [ " I put on a Beatles track
to better test it " ] It sort of reads like the available input sockets
on the back secondarily produce the same result, [ " I tried the two
available inputs, ( "DAT" and "AUX" ) - same thing " ] so this would
indicate an output rather than input issue. It is a shame that he did not
post just one more picture looking down on the print side of the output
connector PCB, which he has otherwise completely exposed. It may have been
possible to see whether or not the joints were bad there.

I wonder if the OP has got anywhere with it yet, as a result of all the
advice offered so far ?

Arfa