"ixuv" wrote in message
oups.com...
On 28 elo, 20:34, "Arfa Daily" wrote:


That's an odd one. What might be a clue is that the next point that
crackles
is always 7.5dB up on the previous point, which would suggest that it is
something 'digital', such as the volume control IC.

I agree.


I seem to have the
service manual for just about every other STR series amp from around that
time, but not the DB795, so unfortunately, I can't look any closer.
However,
I would say that often, the volume control IC also does input selecting,
so
it would be worth having a look to see if that is the case with this amp.
I
have had these ICs go faulty, and this is the sorts of problems that they
can cause. I had one recently that kicked up a lot of noise on one
channel,
above a certain volume level.

I put couple block diagrams in he

http://www.aijaa.com/v.php?i=896688.jpg

If I understand right, there is separate ICs for volume control(each
speaker pair) and input selector.

I would think that you would be able to see
this crackle on a 'scope, and then know for sure where it was getting
into
the chain, but failing that, dabbing a 0.47uF cap to ground from various
key
points, such as the output of the volume control IC, might prove
revealing.

I don't have oscilloscope, only basic multimeter. And I don't fully
understand what you mean this dabbing.


Looking at the block diagram, without doing any 'scoping, I think that my
first favourite would definitely be the volume control IC on the offending
channel. If you have the necessary desoldering equipment and skills to get
the IC out without damaging either it, or the print that it's soldered to,
perhaps you could swap it for the one on the other front channel ? That
would at least prove the point one way or the other.

As far as using a cap to help with diagnosing where noise is getting into a
channel, it works best wher you have a low-level crackle or hiss. Crackles
and hisses contain significant high frequency components. At these
frequencies, a cap of value 0.47uF, has a low reactance - around 300 ohms or
so at 10kHz. So, if you connect one side of it to ground, and touch the
other end to various places in the amplifier chain, such as inputs and
outputs of ICs, transistor collectors and so on, then wherever the noise is
present, it will be significantly attenuated when you touch or 'dab' the
capacitor onto such a point. Working backwards, when you find the first
place that dabbing the cap on has no effect on the noise, then you know that
the point that the noise is originating is the next stage forward. This can
be a very effective way of finding a noisy opamp or transistor or resistor,
when the noise signal is just too small to see effectively on a 'scope. I
would normally use only a low-leakage polyester or similar cap for this, and
the value is not critical. 1uF or higher is fine, but not in an electrolytic
format, as the leakage from one of these might in some cases, be enough to
upset operating conditions in a DC coupled amp.

Arfa