RE: Sansui G-4700 Power supply

Hello to all Sansui & solid state enthusiasts.

this receiver needed many transistors replaced, they have been carefully replaced with modern parts, list available upon request. Upon slow power-on via Variac (approximately 70 Volts) I am noticing immediately the power transformer making a buzzing noise (excessive noise based on previous repairs) it is also becoming warm to touch and you can begin smelling the resins or paint softening, not a good sign. The voltage rails on the power transistors are ok on one collector side and not good on another. If I disconnect the tree secondary wires from the transformer to the amplifier board, the buzzing is gone. My gut instinct tells me that there is something wrong with the amplifier board or the power supply itself for the high voltage rails. I also cannot get the -23 VDC rail to show (the positive is fine). DC Offset & BIAS seem to be very responsive. Any pointers may be much appreciated, thanks
All semiconductors have been checked, resistors & fusistors as well, caps are within specs.


  • G-4700.pdf
    897.7 KB · Views: 111
Some thoughts:

Having a Variac is a great tool. Does the negative power supply generate DC voltage even though current draw is excessive?

Sometimes there are wire jumpers on the PCB that can be disabled to isolate board sections.

If there aren’t removable jumpers, here’s a technique that I use to help trace shorts. Turn Variac drive up to a safe level that won’t provoke damage but forces some current into problematic circuits. Put one meter lead on the big supply cap on the problematic negative rail and probe along the length of suspect circuit traces. Look for increasing voltage drops of only a few mV. If you venture down an incorrect branch, the voltage drop won’t change and you’ll know to find your way back to the trace that shows increasing voltage drop. Often you’ll be led to a defect. Having a meter with 0.1mV resolution helps.

Good luck!
See the primary voltages to and from transformer with it out of circuit.

Check if the Zener diodes are leaky or short, along with the caps across them.
Then follow the circuit.

Hum or smell is a sign of overload in transformer.
Lucky it has not blown...put new fuses, sometimes the old ones lose rating.
I have done extensive component testing on and off the board, with equipment & testing circuit jig. Nothing seems to make any substantial difference. I can bring the Variac up to 50 Volts and for short periods of time I can do measurements without douing any damage. On the collector of TR01 & TR02 I am getting at the most -5.4 VDC.
If this can be of any useful information for the output devices I used Sanken complementary pair 2SA1386 & 2SC3519, which I have used before many times on this topography.
Pre-drivers are KSA modern transistors. I have also tried to isolate the power supply so that it will have "no load", again no difference..........attaching three pictures for the sake of it.


  • 20230616_193911.jpg
    460.8 KB · Views: 54
  • 20230616_193920.jpg
    339.2 KB · Views: 67
  • 20230616_193928.jpg
    350.2 KB · Views: 58
You mean the transformer gets hot when the secondary side is out of circuit?

Check the winding impedance with a good digital meter, with a fresh battery, and deduct the zero reading when the leads are shorted in x1 ohms range. Most meters do not read zero when shorted.
Positive rail is better than negative? For perspective, when you observe about -5.4V on collector of TR01, what voltage do you find on collector of TR03?

You said earlier transformer hum stopped when blue 32V transformer leads were disconnected? Are you certain D602 and D604 are OK, especially not shorted?

With -5.4V on negative rail, would you put black lead of DVM on negative terminal of C602 and probe voltage on collector TR1 and TR3? It may be only a few mV. If you find any drop, would you then measure the voltage re ground of the R01 and R02 resistors, i.e. amp output and the emitter voltages at TR01 through TR04?

Joined 2007
Paid Member
Quick thoughts.

1/ Fit a link across C41 and C42 to force zero bias.

If that makes no difference then:

2/ Remove D605 and D606 which isolates the power amps (excluding the output pair) from the supply.

If the rails still do not come up then:

3/ Remove R01 and R02 which prevents current flow through the output transistors... has to be done at this point.

If the rails still do not appear then either the output transistor collectors are not insulated from the heatsink or there is a proper PSU issue which really only leaves bridge and caps.
Hi if I disconnect the secondary rails 19, 20, 21 from the board the transformer is perfectly fine, not hot, no buzzing.
You mean the transformer gets hot when the secondary side is out of circuit?

Check the winding impedance with a good digital meter, with a fresh battery, and deduct the zero reading when the leads are shorted in x1 ohms range. Most meters do not read zero when shorted.
Hello dear members, big thanks to all of you for contributing to another painful frustrating & successful repair.
I gather all the information that you provided then started removing components from the negative rail starting at the diode rectifiers, ceramic caps, main filters etc.....The strange thing was that I was measuring a very low resistance or impedance between the collectors of TR02 & TR01 (1.2 ohm almost a short). Once I loosened the screws that were probably overtightened to make the back-plate in almost contact through the mica foil with the heat-sink which is grounded. This solved the problem & the negative rail is back in order. Many thanks see you at the next adventure.
  • Like
Reactions: 1 users