On Mon, 18 Aug 2008 09:59:21 +0100, "Arfa Daily"
put finger to keyboard and composed:


"Franc Zabkar" wrote in message
.. .
On Mon, 18 Aug 2008 15:44:29 +1000, Franc Zabkar
put finger to keyboard and composed:

Here it is:
http://www.eserviceinfo.com/download.php?fileid=37140

Page 25 appears to be the relevant section.

It appears that the protection circuitry is in two parts. Q551 and
Q651 sense the current in the emitter resistors, whereas Q552 and Q553
sense the DC offset at the amplifier outputs. Both protection circuits
trigger the Protect signal at pin 24 of the LM6405G microprocessor via
Q554, and the uP then trips the relay via pin 30.

Disconnecting R556 will disable the DC offset protection, and
disconnecting R553 and R653 will disable the current overload
protection on the left and right channels, respectively. Disconnecting
each of the three resistors one at a time may help narrow down the
source of the problem.

By my reckoning, you would need a DC offset of 2.7V before Q552 would
turn on, and you would need 1.2A in the emitter circuit of Q552 and
Q553 in order for Q551 to turn on. The latter is much higher than your
out-of-spec 60mA idling current.

- Franc Zabkar
--

OK then, how about this for a scenario. As both pots are now set in about
the same place to get the same idle current, and one of them wasn't before,
it follows that the one which was wrong, had almost certainly been set
wrongly for a long time. This means that the transistors and heatsink on
that channel have been running much hotter than they should have. If it uses
mica washers with heatsink paste, rather than silicone rubber insulators,
the excess heat may have dried the paste out over the years, such that it is
now doing a poor job, and the transistors can get hot enough to start
running away, which creates a large emitter current, which is then sensed,
and causes the shutdown. Even if the insulators are rubber, a close
inspection might reveal that the excess heat has done them no good, and they
have gone hard. Just a thought, as I agree with your thoughts regarding
offdet and current trip points.

Arfa

The OP has stated that the relay trips and closes again, presumably
straight away. Is this consistent with a thermal runaway scenario?

It seems to me that the uP is sensing a fault condition and opening
the relay to disconnect the speakers. But opening the relay achieves
nothing in the case where the amp is unloaded. So if there is a
thermal runaway condition, then surely it will persist even after the
relay is deactivated, in which case the uP should not reactivate it.

BTW, there is an additional 1 ohm resistor in parallel with the two
0.47 ohm emitter resistors, so the misadjusted idling current appears
to have been 120mA, not 60mA as I previously calculated. This current
would result in 10W of dissipation in the output transistor pair
instead of only 2.5W when correctly adjusted. But surely even this
high figure is much less than that generated by the amp when running
under load?

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.