Marantz 2270 speaker protect problem; req. for help

[Jelcke]

Hello,
I have a Marantz 2270 that will not let speaker relay engage.

Things I checked:
No DC offset on both channels (measured at J757)
Bias 5mv / no shorts or heavy current pulling from power boards / output is nice waveform.
Checked components and voltages on P800. (replaced H813 and caps)
Relay clicks when I apply 24V with power to amp off.
Only problem is voltage on collector of H809 = +40V at startup drops off to +27.5v and should be +0.8.
Can't figure out what the source is of this voltage. Your suggestions are very much appreciated.

Greetings from France,
Jelcke

Schematic:
http://akdatabase.org/AKview/albums/userpics/10007/Marantz 2270 Schematic HiRes ALL.pdf

 

[mbz]

Power for the protection circuit comes from a separate transformer winding
and rectified at H811, 812. Thinking that the H809c voltage is high since
there is no current flow (and subsequent volt drop).

Some things to try would be,
measure H809, 810 b/c/e VDC to GND/chassis, outside chance one of the transistors
has failed open. Check or replace C807 220uf/10V, I think this one does the power
on mute timing with R816. Certainly would check for dry solder joints in the area.

The voltage measurements will help the marantz gurus when they come online.

 

[Jelcke]

Thanks, I have made a picture of the measurements on p800 and components I checked.
Caps have been replaced. Checked transistors H808 is off and H809 + H810 are on.
R803 was burned so I replaced it. Perhaps a clue to the location of the problem. H803 is OK.

 

[mbz]

Good work on the measurements. It almost looks like H809, 810 is not conducting, ie, it's gone (c-e?) open.
I'd be tempted to replace both (KSC2383). Marantz guys should be along shortly.

 

[Jelcke]

H809 E -> C 1.3v
B -> E 0.57v

H810 E -> C 17.8 V
B -> E 0.6v

When I measure H809 E->C the relay clicks...

 

[Jelcke]

Replaced H809 with a BC337-40 and problem is fixed....
The old transistor H809 test fine on a Atlas DCA75 pro. So that sucks a bit...

Tested H809 resistor with old Heathkit IT-3120 and see no problems with it. Beta is 35. No leakage in any direction.

Any ideas on how this is possible are welcome....

 

[mbz]

Any ideas on how this is possible are welcome....

Sometimes the silicon needs a few "volts/amp" before they breakdown. Also the testers are only so-so, if it passes
diode test then always consider that the part may be faulty, however, you have nice testers... The other possibility
is a bad solder connection/joint at H809, but you would have inspected these joints.