Cambridge A5 amp crowbarring?

[N_Cook]

Headscratching time. On first receipt tried the amp out as owner had
been using 1 channel only , after having blown the other by shorting it.
Both LV ac fuses ok but one SAP15 p-type blown.
Had to cross-over power Darlington terminals as mirror footprint from
the usual BCE orientation.
Cleared out and replaced with that fudge for the SAP and other blown
1/8W Rs and tried out at 50% mains.
Variac meter swung right over to the right and both LV fuses blew.
Replaced fuses and tried at 20% mains, half amp or so of mains draw.
Diconnect mains and main +ve and main -ve rail measure 0 ohm to ground
for 10 or 20 seconds and then goes high impedance.
Tried at 10% mains and sometimes +ve rail would be at zero ohms for 20
seconds or so and sometimes -ve rail like that, on repeating this game.
Cannot see anything in the way of a crowbar SCR anywhere or relay and
circuit looks much as the A500 that is out there , including overlay
numberings.
There is conductive brown glue to the main caps but that measured 2 meg
or so minimum, with close DVM probes.
All output devices and my fudge cross-compare DVM-D cold fashion, fine
still.
When I say 0 ohms I do mean that , well 0.1 ohm or less resistance.
Unless anyone has an idea what is going, the next will be to remove my
fudge and power up again.

 

[N_Cook]

Thats enough of this malarkey for today.
About 10% mains for 20 seconds or so , the only thing warming up on IR
thermometer was the 2 fuses, so something big presumably taking the
power. That with about .2 amp mains current draw, 5V ac over secondaries
and +/-1.8V on the DC rails

 

[Phil Allison]

Nucase Kook wrote:

Thats enough of this malarkey for today.
About 10% mains for 20 seconds or so , the only thing warming up on IR
thermometer was the 2 fuses, so something big presumably taking the
power. That with about .2 amp mains current draw, 5V ac over secondaries
and +/-1.8V on the DC rails

** Bet you have revered the positions of the PNP and NPN Darlingtons - causing the internal C-E diodes to conduct soon a power is applied.



..... Phil

 

[N_Cook]

I'd wired in the train of 5 Shottky SMD diodes , the wrong way round so
no biasing, doh!

 

[N_Cook]

On 02/11/2016 13:36, N_Cook wrote:

I'd wired in the train of 5 Shottky SMD diodes , the wrong way round so
no biasing, doh!

So far good output level match and sinuosity with no load ,with 1KHz
input at 60% mains ,and reasonable biasing DC levels match between L and
R channels with original preset setting. Will add 8R loads and repeat
these tests and adjust bias while still at 60% mains ,interim, before
going up to 100% mains

 

[N_Cook]

On 03/11/2016 09:46, N_Cook wrote:

On 02/11/2016 13:36, N_Cook wrote:
I'd wired in the train of 5 Shottky SMD diodes , the wrong way round so
no biasing, doh!


So far good output level match and sinuosity with no load ,with 1KHz
input at 60% mains ,and reasonable biasing DC levels match between L and
R channels with original preset setting. Will add 8R loads and repeat
these tests and adjust bias while still at 60% mains ,interim, before
going up to 100% mains

Required 6 diodes of the SM" 2 in 1 pack BAT54S type I used, not 5, to
get the right bias zone for the preset, 100% mains and stable wiht load,
just quality of sound check to go