Sansui BA-3000 blowing mains fuse following restoration

[Hyperion]

Here are the results of the variac testing of the secondary windings @Hyperion .

Looks fine to me - I am very pleased for you!

 

[BradWil]

Sanken Rectifiers

D01 SG-5TS common cathode
D1 Vf=0.572V If=4.401mA
D2 Vf=0.573V If=4.401mA

D02 SG-5TR common anode
D1 Vf=0.578V If=4.393mA
D2 Vf=0.555V If=4.416mA

Tested with Atlas DCA55
Mica seems intact



D03 C08PN (1850) common cathode
D1 Vf=0.615V If=4.358mA
D2 Vf=0.619V If=4.3545mA

D04 C08PR (1850R) common anode
D1 Vf=0.613V If=4.359mA
D2 Vf=0.607V If=4.366mA

F-2511 top and bottom photos

Transistors next

 

[Hyperion]

They look OK - perhaps surprisingly - leaving only the remote possibility of 'breakdown under load'.

Next step I recommend: - connecting the secondary to the low current supply, (Yellow-Green-Yellow) and measuring voltage +ve & -ve to ground, applying power with DBT & Variac if wanted. For this test it shouldn't matter that the output of the rectifiers can 'see' the regulator circuitry. Should get a small flash from the DBT and dimming to almost nothing.

You could continue selectively reconnecting stuff one at a time, in the above fashion - 'baby steps'

 

[BradWil]

I just realised that even though the driver amps are isolated, pins 18, 24, 25 on F-2511 and the blue/red leads to the filter caps are connected to the output array. The outputs are all original matched Sanken and tested fine during restoration. Any outputs shorting to ground would also cause a DBT fail, even with the driver amps isolated, right?

I will remove the heatsinks and test each one next ...

 

[Hyperion]

Any outputs shorting to ground would also cause a DBT fail, even with the driver amps isolated, right?

Yes, I think so, if the short were C-E, but probably not if B-E for example.

 

[BradWil]

Output modules removed, driver amps connected, DBT pass, therefore likely a shorted output ... will replace all with On Semi MJ21194/93 and try to match them as best I can.

 

[richardpp]

@Hyperion is correct, a C-E short will fail the DBT even with the driver amps isolated.

You can quick-check the output transistors for C-E shorts without removing the heatsink assemblies by probing the transistor case (C) together with the orange wire on the output board (E) shown in your last photo above. Because the transistors are in parallel you only need to test one to test all three, so test one of the NPN's on the front side of the heatsink and one PNP on the back, for each heatsink.

To properly isolate the load from the unregulated supply when testing with a DBT, disconnect terminals 7 and 8 (output stage) and 29 and 30 (driver stage) on F-2511.

Just saw your last post re: shorted outputs. Old uniwatt pre-driver transistors a common cause for shorted outputs.

 

[BradWil]

@Hyperion is correct, a C-E short will fail the DBT even with the driver amps isolated.

You can quick-check the output transistors for C-E shorts without removing the heatsink assemblies by probing the transistor case (C) together with the orange wire on the output board (E) shown in your last photo above. Because the transistors are in parallel you only need to test one to test all three, so test one of the NPN's on the front side of the heatsink and one PNP on the back, for each heatsink.

To properly isolate the load from the unregulated supply when testing with a DBT, disconnect terminals 7 and 8 (output stage) and 29 and 30 (driver stage) on F-2511.

Just saw your last post re: shorted outputs. Old uniwatt pre-driver transistors a common cause for shorted outputs.

Many thanks, very helpful.

Going through the outputs now. Second one tested, TR708 2SA908 shorted on CBE.

Will replace all outputs then check and replace the pre-drivers on the driver amps. Thank goodness it's a completely straightforward fix. The lack of rail fuses really threw me and I really should have checked the outputs first! Live and learn.

 

[BradWil]

so far, 3 outputs on the right channel dead

 

[BradWil]

Only right channel failure but will do both channels anyway. Next driver amps - drivers and pre-drivers. Hopefully the drivers are ok, as replacements are slower.

 

[richardpp]

Just wondering what your failed criteria is in the above chart? If it helps I have 1 spare 2SA908 and 1 spare 2SC1585.

R34, R35 and /or R36, R37 on driver board usually show signs of overheating with shorted output transistors.

 

[BradWil]

Just wondering what your failed criteria is in the above chart? If it helps I have 1 spare 2SA908 and 1 spare 2SC1585.

R34, R35 and /or R36, R37 on driver board usually show signs of overheating with shorted output transistors.

The shorted ones, literally shorted all 3 (CBE) leads. The failed one registered as a diode so one side was open.
Thanks, will check those resistors, very helpful.

Appreciate the offer but I'm going to run with OnSemis as per below.



With the power amps, I plan to replace TR08 (2SC1628), TR09 (2SA818) and TR04 (2SA818), TR05 (2SC1628) - may use Toshibas TTC004B/TTA004B but need to research this.
Failed Right power amp pictured.

 

[BradWil]

DBT passed after replacing all outputs and with driver amps removed.

 

[BradWil]

Interestingly, all the right driver amp's transistors tested fine. I plan to repalce TR01, TR02, TR03 as they have the black oxidised legs.
I will also replace TR04, TR05, TR08, TR09 - researching replacements.

Looks like one or more outputs failed on their own?

 

[BradWil]

Replaced the right driver transistors as per below. Passed DBT but now I'm unable to dial any BIAS, even on full mains - any ideas?

Perhaps I should reinstall the originals, which did test OK? I only changed them as it was mentioned the original pre-drivers are a potential point of failure.

 

[BradWil]

Something's up. Installed the left driver board, unaltered with all original semis - no BIAS either?
DBT passes fine and unit is comes out of protection.
Rails +/- 68V correctly feeding F-2512 connector board.

Any experience using the MJ21195/96 On Semis on the BA3K? Do some resistor values need to be changed on the drivers in order to BIAS? In the case of the left driver, it is only the outputs that have changed, all drivers, pre-drivers and other semis original.