Sansui Alpha AU-907i MOS Limited

Member
Joined 2021
Paid Member
Hello, I have Sansui Alpha AU-907i MOS Limited that sometimes lose the right channel at low volume, when I increase the volume the right channel becomes alive also after 1 or 2 hours the problem disappeared. I changed all the electronics capacitors to good ones Nichicon MUSE KZ Series (the original was Nippon) I adjust bias to 15mV and set the DC 0V to 0V on both channels and the problem presist as before. The same problem as before. The only thing that I did on changing the capacitors is putting the 47uf 100V instead of 50V also the radiator of rights channel is much hotter than the left. Please help. Thank you.
 

Attachments

  • 20250106_203739.jpg
    20250106_203739.jpg
    600.5 KB · Views: 95
  • 20250106_204229.jpg
    20250106_204229.jpg
    776.5 KB · Views: 88
  • 20250106_204515.jpg
    20250106_204515.jpg
    744.3 KB · Views: 80
  • 20250106_204539.jpg
    20250106_204539.jpg
    689.7 KB · Views: 75
  • 20250106_204941.jpg
    20250106_204941.jpg
    826.3 KB · Views: 78
  • 20250106_205000.jpg
    20250106_205000.jpg
    802.5 KB · Views: 76
  • 20250106_205120.jpg
    20250106_205120.jpg
    795.4 KB · Views: 90
Yes of course all the transistors are mounted on the heatsink with mica between them and thermal compound. How much voltage should be there? Because I've already adjusted the bias voltage to 15mV (should be 8.8mV I increased it to that level) hot and cold on both channels.
 
Yes. Those Alpha series amps have hot and cold. I don't have the service manual to know the values. All the small transistor doesn't came with heatsink except the two ones in the middle. Any idea how can I test it to see if it's the amp section or the preamp? There's direct input that the tone and balance control doesn't effect. I think it's direct input without the preamp.
 

Attachments

  • 20241018_151504.jpg
    20241018_151504.jpg
    713 KB · Views: 75
Last edited:
Just wanted to note that both comments in #10 can be correct/appropriate--- an intermittent channel could still have an independent bias defect.

Wild guesses: are there Zobel networks? Are their components functional? Are the Zobel resistors hot? I suggest a 'scope to check for possible oscillation.

Good luck!
 
Hi, thank you very much. I will check the relay and the voltage across the bias resistor. What should be the voltage across the resistors? and should I put the two leads of the meter across the two resistor legs or one lead on one resistor leg and the other refer to the ground? What is Zobel network or resistors? For the bias settings on the service manual that mbz (#11) shared (Thank you for that) the bias should be 26mV. The manual is in Japanese (thanks God that the numbers are international 🙂) Maybe I should start from there and see what happens. Thanks for all the comments.
 
Across the resistor legs. First after turning on the amp, a second time after 1 hour or when it is warmed up.

when I increase the volume the right channel becomes alive also after 1 or 2 hours the problem disappeared.

Maybe a bad solder joint or a defective part that "works" at higher temperature when the amp has warmed up.
When the amp is cold and you turn up volume, the current "jumps over" in the broken part or solder joint.

Different heatsink temperatures and lost channel may or may not be two different issues.
 
The bias test points are only 1 sample out of 4 resistors.
So, if you have adjusted all 4 pots to 26mV on the test points, all 16 resistors should show the same 26mV.
You should adjust bias after the amp is warm.

After that, turn off, let it cool down. Turn on again and check again all 16 resistors.
 
Last edited: