[EADGBE]

Good News!

This headphone amplifier has been REPAIRED and is now working
perfectly.

Here's what happened:

I have been taking parallel measurements between the "good" and "bad"
channels since the beginning.

I took voltage measurements, resistance measurements, tested ALL
capacitors, and tested ALL transistors.

But someone in my original thread (I can't remember who offhand --
sorry) did mention checking the circuit board traces.

"No need," I thought. "The circuit board and all solder joints look
fine to me."

But in a fit of desperation, I nonetheless decided to whip out a
tester I hadn't used yet: my magnifying glass.

That's when I found it -- a *microscopic* crack in the trace between
the emitter of Q208 and its adjoining 3.3 ohm resistor. The crack was
so incredibly tiny that I easily overlooked it--repeatedly.

I took one of my jumper leads and clipped one end of it to Q208's
emitter and the other end of it to the 3.3 ohm resistor.

I then took voltage measurements. EVERYTHING WAS NOW ON SPEC:

Q205's collector now measures +7.09V, and Q206 now has the following
measurements:
B: +6.51V
C: +7.08V
E: +5.84V

I was able to bridge the crack with a well-placed solder trail, and
the headphone amp now sounds great.

A BIG "THANK YOU" TO EVERYONE WHO HELPED ME OUT! I OWE YOU BIG TIME.

[Richard Crowley]

"EADGBE" wrote ...
That's when I found it -- a *microscopic* crack in the trace between
the emitter of Q208 and its adjoining 3.3 ohm resistor. The crack was
so incredibly tiny that I easily overlooked it--repeatedly.

I took one of my jumper leads and clipped one end of it to Q208's
emitter and the other end of it to the 3.3 ohm resistor.

I then took voltage measurements. EVERYTHING WAS NOW ON SPEC:

Its hard to imagine how the output node (where the two
3.3 ohm resistors join) could be correct (6.4V) without
the path through the lower 3.3 ohm resistor and Q208?

But at least you got it working. Good for you. Trouble-
shooting ("fault-finding" as they call it in the UK) discrete
circuits is rapidly becoming a lost art.

[Don Pearce]

On Tue, 22 Apr 2008 12:38:19 -0700, "Richard Crowley"
wrote:

"EADGBE" wrote ...
That's when I found it -- a *microscopic* crack in the trace between
the emitter of Q208 and its adjoining 3.3 ohm resistor. The crack was
so incredibly tiny that I easily overlooked it--repeatedly.

I took one of my jumper leads and clipped one end of it to Q208's
emitter and the other end of it to the 3.3 ohm resistor.

I then took voltage measurements. EVERYTHING WAS NOW ON SPEC:

Its hard to imagine how the output node (where the two
3.3 ohm resistors join) could be correct (6.4V) without
the path through the lower 3.3 ohm resistor and Q208?


Pure coincidence - it was the breakdown voltage of the reverse biased
base-emitter of Q207.

But at least you got it working. Good for you. Trouble-
shooting ("fault-finding" as they call it in the UK) discrete
circuits is rapidly becoming a lost art.

Quite so - when all the components are right, but there is still a
problem, go for the board.

d

--
Pearce Consulting
http://www.pearce.uk.com

[John Phillips[_2_]]

On 2008-04-22, Don Pearce wrote:
On Tue, 22 Apr 2008 12:38:19 -0700, "Richard Crowley"
wrote:

"EADGBE" wrote ...
That's when I found it -- a *microscopic* crack in the trace between
the emitter of Q208 and its adjoining 3.3 ohm resistor. The crack was
so incredibly tiny that I easily overlooked it--repeatedly.

I took one of my jumper leads and clipped one end of it to Q208's
emitter and the other end of it to the 3.3 ohm resistor.

I then took voltage measurements. EVERYTHING WAS NOW ON SPEC:

Its hard to imagine how the output node (where the two
3.3 ohm resistors join) could be correct (6.4V) without
the path through the lower 3.3 ohm resistor and Q208?

Pure coincidence - it was the breakdown voltage of the reverse biased
base-emitter of Q207.

Rubbish. ( Well, possibly :-) )

Q207 was indeed hard off with its base at 0V. Q208 was not connected
to the output because of the fault. The output junction was therefore
biassed only by the emitter voltage or Q204 (8V nominal) via the
15 k resistor.

If the measurement was made with a 20 kohm/V multimeter on the 10 V range
(200k) then I would expect the potential divider effect to pad that 8
V down to about 7.4 V. EADGBE actually posted that the node was 7.3V
(not the nominal 6.4 V of the schematic).

OK, I admit it could have been reverse bias breakdown but I prefer my
version as it doesn't require that coincidence. (Well OK I do make an
assumption about the measurement kit instead ...)

--
John Phillips

[Don Pearce]

On 22 Apr 2008 20:09:23 GMT, John Phillips
wrote:

On 2008-04-22, Don Pearce wrote:
On Tue, 22 Apr 2008 12:38:19 -0700, "Richard Crowley"
wrote:

"EADGBE" wrote ...
That's when I found it -- a *microscopic* crack in the trace between
the emitter of Q208 and its adjoining 3.3 ohm resistor. The crack was
so incredibly tiny that I easily overlooked it--repeatedly.

I took one of my jumper leads and clipped one end of it to Q208's
emitter and the other end of it to the 3.3 ohm resistor.

I then took voltage measurements. EVERYTHING WAS NOW ON SPEC:

Its hard to imagine how the output node (where the two
3.3 ohm resistors join) could be correct (6.4V) without
the path through the lower 3.3 ohm resistor and Q208?

Pure coincidence - it was the breakdown voltage of the reverse biased
base-emitter of Q207.

Rubbish. ( Well, possibly :-) )

Q207 was indeed hard off with its base at 0V. Q208 was not connected
to the output because of the fault. The output junction was therefore
biassed only by the emitter voltage or Q204 (8V nominal) via the
15 k resistor.

If the measurement was made with a 20 kohm/V multimeter on the 10 V range
(200k) then I would expect the potential divider effect to pad that 8
V down to about 7.4 V. EADGBE actually posted that the node was 7.3V
(not the nominal 6.4 V of the schematic).

OK, I admit it could have been reverse bias breakdown but I prefer my
version as it doesn't require that coincidence. (Well OK I do make an
assumption about the measurement kit instead ...)

Mmmmm...... 20k/V multimeters are pretty rare beasts these days. I
have an old AVO 8 but I don't think I'd have the confidence to
estimate two decimal places in its voltage readings ;-)

My money is still on reverse breakdown. Trannies are almost as good as
zeners in this configuration.

d

--
Pearce Consulting
http://www.pearce.uk.com

[Geoff]

EADGBE wrote:

That's when I found it -- a *microscopic* crack in the trace between
the emitter of Q208 and its adjoining 3.3 ohm resistor. The crack was
so incredibly tiny that I easily overlooked it--repeatedly.


Jeepers. You would normally find that around any PCB-mount jack or DC
socket. Unless the whole thing has had a severe jarring. May have been a
heat/cool fatigue thang.


geoff

[isw]

In article
,
EADGBE wrote:

Good News!

This headphone amplifier has been REPAIRED and is now working
perfectly.

Here's what happened:

I have been taking parallel measurements between the "good" and "bad"
channels since the beginning.

I took voltage measurements, resistance measurements, tested ALL
capacitors, and tested ALL transistors.

But someone in my original thread (I can't remember who offhand --
sorry) did mention checking the circuit board traces.

"No need," I thought. "The circuit board and all solder joints look
fine to me."

But in a fit of desperation, I nonetheless decided to whip out a
tester I hadn't used yet: my magnifying glass.

That's when I found it -- a *microscopic* crack in the trace between
the emitter of Q208 and its adjoining 3.3 ohm resistor. The crack was
so incredibly tiny that I easily overlooked it--repeatedly.

I took one of my jumper leads and clipped one end of it to Q208's
emitter and the other end of it to the 3.3 ohm resistor.

I then took voltage measurements. EVERYTHING WAS NOW ON SPEC:

Q205's collector now measures +7.09V, and Q206 now has the following
measurements:
B: +6.51V
C: +7.08V
E: +5.84V

I was able to bridge the crack with a well-placed solder trail, and
the headphone amp now sounds great.

You should *always* bridge the crack with a strand or two of fine wire;
that'll keep it from cracking again, Real Soon Now.

Isaac

[Mr.T]

"isw" wrote in message
...
I was able to bridge the crack with a well-placed solder trail, and
the headphone amp now sounds great.

You should *always* bridge the crack with a strand or two of fine wire;
that'll keep it from cracking again, Real Soon Now.

Why use fine wire? I just use an off-cut from a resistor leg for normal size
traces. Much easier to place and they're pre-tinned as well.

MrT.

[Richard Crowley]

"Mr.T" wrote ...
"isw"wrote ...
I was able to bridge the crack with a well-placed solder trail, and
the headphone amp now sounds great.

You should *always* bridge the crack with a strand or two of fine wire;
that'll keep it from cracking again, Real Soon Now.

Why use fine wire? I just use an off-cut from a resistor leg for normal
size
traces. Much easier to place and they're pre-tinned as well.

"Fine" is a relative term. And some PC boards are too dense
and use traces that are too narrow to use resistor lead wire.
I typically use 30Ga wire-wrap wire for that purpose. But I
agree that just bridging the gap with a blob of solder is not
an optimal solution.

[Mr.T]

"Richard Crowley" wrote in message
...
You should *always* bridge the crack with a strand or two of fine wire;
that'll keep it from cracking again, Real Soon Now.

Why use fine wire? I just use an off-cut from a resistor leg for normal
size traces. Much easier to place and they're pre-tinned as well.

"Fine" is a relative term.

But not so much when you say "a strand or two".

And some PC boards are too dense
and use traces that are too narrow to use resistor lead wire.

Which is why I already said "for normal size traces". That covers a lot of
audio gear.

MrT.

[isw]

In article ,
"Mr.T" MrT@home wrote:

"isw" wrote in message
...
I was able to bridge the crack with a well-placed solder trail, and
the headphone amp now sounds great.

You should *always* bridge the crack with a strand or two of fine wire;
that'll keep it from cracking again, Real Soon Now.

Why use fine wire? I just use an off-cut from a resistor leg for normal size
traces. Much easier to place and they're pre-tinned as well.

Depends on the width of the trace...

Isaac

[Mr.T]

"isw" wrote in message
...
You should *always* bridge the crack with a strand or two of fine
wire;
that'll keep it from cracking again, Real Soon Now.

Why use fine wire? I just use an off-cut from a resistor leg for normal
size
traces. Much easier to place and they're pre-tinned as well.

Depends on the width of the trace...


Which is why I said "for normal size traces".
The post I replied to OTOH said "*always*". That is what I objected to.
(Doesn't anybody read before replying?)

MrT.

[Richard Crowley]

"Mr.T" wrote ...
"isw" wrote ...
You should *always* bridge the crack with a strand or two of fine
wire;
that'll keep it from cracking again, Real Soon Now.

Why use fine wire? I just use an off-cut from a resistor leg for normal
size
traces. Much easier to place and they're pre-tinned as well.

Depends on the width of the trace...


Which is why I said "for normal size traces".
The post I replied to OTOH said "*always*". That is what I objected to.
(Doesn't anybody read before replying?)

I suspect that the standard "normal" trace width may have
shrunk while you weren't looking. Or maybe you have the
luxury of working on older, more human-scale equipment.

[jakdedert]

Richard Crowley wrote:
"Mr.T" wrote ...
"isw" wrote ...
You should *always* bridge the crack with a strand or two of fine
wire;
that'll keep it from cracking again, Real Soon Now.
Why use fine wire? I just use an off-cut from a resistor leg for normal
size
traces. Much easier to place and they're pre-tinned as well.
Depends on the width of the trace...

Which is why I said "for normal size traces".
The post I replied to OTOH said "*always*". That is what I objected to.
(Doesn't anybody read before replying?)

I suspect that the standard "normal" trace width may have
shrunk while you weren't looking. Or maybe you have the
luxury of working on older, more human-scale equipment.


In any case, a small wire (like one strand of some zip cord--30 gauge?)
is an improvement over nothing at all.

jak

[Mr.T]

"Richard Crowley" wrote in message
...
I suspect that the standard "normal" trace width may have
shrunk while you weren't looking.

Nope, but what's normal for an audio amp is not necessarily "normal" for a
computer motherboard. Since we weren't discussing computers, they are not
included in my comment.

Or maybe you have the
luxury of working on older, more human-scale equipment.

Yes, since most micro-miniature equipment is not really designed to be
worked on, and we were discussing an audio amp, not digital electronics.

As usual, people can argue endlessly about linguistics, frankly I'm not
interested.

MrT.

[EADGBE]

On Apr 24, 9:47 pm, "Mr.T" MrT@home wrote:

Yes, since most micro-miniature equipment is not really designed to be
worked on, and we were discussing an audio amp, not digital electronics.

MrT:

I must add that most micro-miniature equipment is also NOT WORTH being
worked on!

Analog rules!

:-)

P.S.: I am taking everyone's advice and adding a bit of fine wire to
the solder trail I used to fix the circuit board. Good advice, it is!

[GregS[_3_]]

In article , EADGBE wrote:
On Apr 24, 9:47 pm, "Mr.T" MrT@home wrote:

Yes, since most micro-miniature equipment is not really designed to be
worked on, and we were discussing an audio amp, not digital electronics.

MrT:

I must add that most micro-miniature equipment is also NOT WORTH being
worked on!

Analog rules!

:-)

P.S.: I am taking everyone's advice and adding a bit of fine wire to
the solder trail I used to fix the circuit board. Good advice, it is!

Sometimes its better to add a complete wire from end to end and forget the trace.

greg

[Arny Krueger]

"GregS" wrote in message

In article
,
EADGBE wrote:
On Apr 24, 9:47 pm, "Mr.T" MrT@home wrote:

Yes, since most micro-miniature equipment is not really
designed to be worked on, and we were discussing an
audio amp, not digital electronics.

MrT:

I must add that most micro-miniature equipment is also
NOT WORTH being worked on!

I dunno. If you have a few $100s invested in something, or say the better
part of a $grand, even a few $100 spent on getting it back into service can
be justified.

Analog rules!

In the larger view, there is no dichotomy between analog and digital.

P.S.: I am taking everyone's advice and adding a bit of
fine wire to the solder trail I used to fix the circuit
board. Good advice, it is!

Sometimes its better to add a complete wire from end to
end and forget the trace.

Good advice.

The worst case of circuit board repair I ever encountered were complete
longitudinal fractures of 2 major boards in a Heath AR-15 receiver. I seem
to recall that both the multiplex and input boards were severed. When I
picked up the package at the APO, parts literally streamed out of a ripped
corner of the box.

I used the wire-replacement/reinforcement technique, and the receiver ran
flawlessly for years afterward.