No known cause for one channel to catastrophically burn out -
literally a burnt hole in the board and loss of many components.
At this stage (justified in the event ) replacing the expensive o/p 15 amp
2SA1216
and 2SC2922 with 25 Amp TIP36C and TIP35C; mid power trannies 2SA1668 and
2SC4382 with TIP41C and TIP42C. Replaced low power trannies and passives
and tracks.
Had to change biasing as driven too strong with original Rs but settled
amplifying
undistorted 1KHz, phones only, with variac at 60 per cent of mains.

BUT Every 20 seconds or so a strong short duration power drain - monitoring
on variac ammeter recovering each time without damage. Assuming
high-frequency instability somewhere any
ideas before I blow up another pair of TIP35/36s where to place some hf
suppression ?.
I was just about to isolate pre-amp stage but putting a DVM probe to i/p of
this channel caused signal injection and permanent allway B-C-E s/c of both
the o/p trannies.
The emitters of the TIP35/36s are at rail volts in this amp.
Seems to be mainly DC coupling in this amp so theoretically could be
DC surges in the preamp as I didn't get around to a scope to monitor the
excursions.

electronic hints and repair briefs
http://homepages.tcp.co.uk/~diverse

Nigel,Diverse Devices,Southampton,England

I replaced the TIP35/36 and replaced both collector 10W 0.2 ohm droppers
with 1/3W 1 ohm resistors temporarily.
Putting scope on o/p there is continuous low level 10mV or so 20KHz or so
oscillation
on this channel.
Trying my old freezer spray trick to find any pitch variation a squirt
in the biasing area caused full level oscillation. I had changed biasing ,R
plus
pot range, from 1.1K to 1.3K to 1.7 to 2.1K but not changed the trannies
in this area. Could well be the original problem.
Bringing my fingers close to components in this area presumably
changes parasitic capacitance and noticeably changes the oscillation
frequency.
Before dropping the gain or changing all the trannies and caps in this area
any ideas to
localise problem better ? I tried card dividers around components and
then squirting but no help.
Putting a scope on this area causes full level, otherwise fatal,
oscillation.

"Nigel Cook" bravely wrote to "All" (18 Mar 04 16:51:23)
--- on the heady topic of "Audiolab 8000A repair well half there"

I'm inclined to think it is not a hf oscillation but rather a lf
instability. It could be high esr supply caps problem in the power
supply (regulator) letting the signal through or a dry electro in a
preamp feedback stage. For example some input bootstrap circuits are
prone to oscillate when there is a critical value of capacitance, not
too much nor not too little and source resistance could affect it too.
One thing for sure if it oscillates then there is a loop to be opened
somewhere!

NC From: "Nigel Cook"

NC No known cause for one channel to catastrophically burn out -
NC literally a burnt hole in the board and loss of many components.
NC At this stage (justified in the event ) replacing the expensive o/p 15
NC amp 2SA1216
NC and 2SC2922 with 25 Amp TIP36C and TIP35C; mid power trannies 2SA1668
NC and 2SC4382 with TIP41C and TIP42C. Replaced low power trannies and
NC passives and tracks.
NC Had to change biasing as driven too strong with original Rs but
NC settled amplifying
NC undistorted 1KHz, phones only, with variac at 60 per cent of mains.

NC BUT Every 20 seconds or so a strong short duration power drain -
NC monitoring on variac ammeter recovering each time without damage.
NC Assuming high-frequency instability somewhere any
NC ideas before I blow up another pair of TIP35/36s where to place some
NC hf suppression ?.
NC I was just about to isolate pre-amp stage but putting a DVM probe to
NC i/p of this channel caused signal injection and permanent allway B-C-E
NC s/c of both the o/p trannies.
NC The emitters of the TIP35/36s are at rail volts in this amp.
NC Seems to be mainly DC coupling in this amp so theoretically could be
NC DC surges in the preamp as I didn't get around to a scope to monitor
NC the excursions.

NC electronic hints and repair briefs
NC http://homepages.tcp.co.uk/~diverse

NC Nigel,Diverse Devices,Southampton,England

"Nigel Cook" wrote:

I replaced the TIP35/36 and replaced both collector 10W 0.2 ohm droppers
with 1/3W 1 ohm resistors temporarily.
Putting scope on o/p there is continuous low level 10mV or so 20KHz or so
oscillation
on this channel.
Trying my old freezer spray trick to find any pitch variation a squirt
in the biasing area caused full level oscillation. I had changed biasing ,R
plus
pot range, from 1.1K to 1.3K to 1.7 to 2.1K but not changed the trannies
in this area. Could well be the original problem.
Bringing my fingers close to components in this area presumably
changes parasitic capacitance and noticeably changes the oscillation
frequency.
Before dropping the gain or changing all the trannies and caps in this area
any ideas to
localise problem better ? I tried card dividers around components and
then squirting but no help.
Putting a scope on this area causes full level, otherwise fatal,
oscillation.


First of all, you need to stop the oscillation. Try connecting a 470pF
capacitor between collector and base of one or the two driver
transistors. If this doesn't work, put another in parallel with the
feedback resistor (the one that goes from the channel output to the
input circuit) If the imput circuit is an Operational Amplifier, then
try connecting another capacitor between its output pin and its
negative input pin. Once the oscillation has stopped, you can find
what else is wrong.
---
Ing. Remberto Gomez-Meda
http://ingemeda.tripod.com/
INGE - Ingenieria Electronica.
Puerto Vallarta, Jalisco, Mexico.