My Sansui AU-D9 doesn't want to work!

[Aldo M.]

Hello everyone, in the mean time I introduce myself, I'm Aldo and I've been following this forum for some time, just since I regained the habit of listening to music properly and buying some amp;

I hope this is the right section to post my question.

I have a problem with an amplifier that is driving me crazy; I've been in possession of a Sansui AU-D9 for 5 years now; when I left home I immediately turned it on and with great happiness I see that it worked, so I remove its wooden lid without even scratching, to give it a clean of the dust (not even much) and make the adjustments of the case; I immediately realize that After a few minutes it gave off a slight smell of burnt components but I don't notice anything burnt, so I start touching the transistors, to see the temperature and I notice that the low power transistors (Q19-Q20) are very hot, (around 80°); I check all the voltages BE and these are within limits; unfortunately, trying to make other measurements, I carelessly cause a short circuit and burn an A1169; after some time I replace all four transistors with new ones, but some prove to be fake and in addition to some resistance outside I also make C2238 and A968; since then I have put back The cover at the Sansui and it's been 5 years and three moves.... about a month ago I took the Sansui back in hand and tried to count the damage caused by the last ignition...... in short, resistors, transistors and final transistors; for the rest of the material I had no problems but for the 2SA196/2SC2773 pairing it seems that there is no hope of finding good ones, so I start looking for current transistors, in production and above all that are compatible.... found, even under the advice of a friend, the 2SC5200/2SA1943 coupling, I take them and put them on;

In my research I find several discussions about the damage that the famous "black flag" capacitors could cause, and my Sansui has many..... well, let's change those too, maybe they cause an auto oscillation to the transistors. Well, I have all the material and very calmly I carry out my work, once everything is mounted and checked well I proceed with turning on the amplifier using a Variac and the classic bulb in series and I start to pray; as I go up with the voltage I check that the voltages are within the range of the Service Manual, until I get to the working voltage, so far everything is ok. I remove the lamp and resume with the Variac until I get to the voltage of 225V, I leave the device with the volume at minimum and without rising and I start to make the DC and Bias adjustments, everything perfect, I'm about to do a power test with connected dummy load and 1kHz signal at 100mV; everything is fine until half of the volume burns a resistor (R39) and two final transistors; luckily I was doing the test only on one channel and therefore the damage is limited to only one channel, the left one. I have another pair of 5200/1943, I mount them and redo all the checks of the case, everything normal; this time I want to connect a pair of speakers instead of the fictitious load; music finally comes out of the speakers, and I leave everything working for about 30 minutes at volume, about 10 o'clock; then I connect the Sansui to a turntable and to my speakers at home, after about 15 minutes and a bang coming from inside the amplifier, everything stops, then no music comes out of my speakers, **** it will have happened again, I checked and rechecked everything several times before making it play. ok, the power transistors and the usual resistor burned out again, this time on the right channel;
Today I finally get back on the Sansui and do some investigations, one thing I never noticed is that on the SM a voltage is reported, at the exit of the rectifier bridges, of 62.5V, but I measure about 55.7V;

The voltages here refers to the SM in the POWER AMP circuit are fine for one channel but in the other I have about 1V more;

Ultimately the machine turns on and works but as soon as the volume is turned up the ends burn;

Sorry for the very long place but I think it is necessary to know every aspect of the attempted repair;

I thank those who have had the patience to read everything to the end and those who will want to give me a hand;

Thank you

PS

My English is Goggle Translator made

 

[Aldo M.]

This is how the amp showed up when I bought it.

 

[Aldo M.]

these resistances that burn...

 

[PF Sound]

Hi......I haven't worked on a D9 but have done a little on a D11 which is appears to be very similar. First of all, when you checked the voltage at the rectifiers you say they are low. Are they balanced? Are you getting +55v & -55v? If so, don't worry about it.
It's hard to discern exactly what work you've done because you've swapped parts with some that you say were fake, & the other part numbers you mention may be different than what was used in the D11, which is the schematic I'm looking at. The fakes could've caused other damage. But it appears that the problem is on the negative rail of both channels (R39 blowing). I'd start by checking all components, upstream & downstream, of that resistor & compare them to the same components on the positive rail. When failures occur they can damage other components. Maybe one of the transistors you used has the wrong pin-out? I'm curious as to why it doesn't go into protect as these units seem to have a reputation of going into protect if you just look at them wrong.

 

[Aldo M.]

Hi, thanks for reading the post and answering, even if you don't have the same device, I'll answer your questions one at a time:

Are they balanced? Are you getting +55v & -55v?

Yes. the voltages are balanced, what I read in the left channel I read in the right channel, to the tenth of a volt.
when I had the first short circuit I had to replace several resistors, some driver transistors (Q15/Q16), and the distortion correctors (Q19/Q20), but the protection does not intervene, so I think that at rest everything is fine, as soon as power is requested the damage occurs.

Maybe one of the transistors you used has the wrong pin-out? I'm curious as to why it doesn't go into protect as these units seem to have a reputation of going into protect if you just look at them wrong.

the only transistors replaced with non-original parts are the two power transistors 2SC2773/2SA1169 with the respective 2SC5200/2SA1943, the pinout is the same, for the rest I used all original parts.

 

[Magnus Back]

For R39 to burn there must be too large current going through it and it needs to come from somewhere.
Q15 and Q16 are more or less the only options here, followed by Q17 and Q18. Also check C14 or just replace it.
The lower Power supply voltage can be because of a short or too high current draw or the fact you're on DBT (which is consistent).

 

[Aldo M.]

Hi, unfortunately I don't always have time to work on the amplifier, so today is that day;

I checked all the bias resistors, I also checked the others around, then I replaced those resistors that are close to the +/- 5% tolerance;

 

[Aldo M.]

In the meantime I have checked the following transistors but I notice that there are differences in gain between them, can this affect the correct balance of the system?
one channel is still working while the other has burned the final transistors (Q17-Q18)

 

[Aldo M.]

Hello everyone, here I am again..... I finally had the time to put my hands on the amplifier, well I solved two important problems:
1° in looking for every little detail that could give me attention I realize that between the radiator fins, looking at the speaker connector board, I see two pads with two holes, as if some components had been unsoldered then looking better from the component side I find the surprise, two resistors soldered only on one side; these are the resistors of the Zobel network at the speaker output; since they were unsoldered these caused some self-oscillation sending my power amps up in smoke as soon as I raised the volume a little without having time to realize what was happening;

2° always looking for the detail, I stupidly try to position the fork diode, which is mounted on the fin of one of the drivers, in the back part of the transistor, or the part that heats up the most; by doing so the temperature of the drivers does not deviate from 63 ° C;


but I still have a problem, as written in the first post this amplifier, unfortunately, does not have its original final transistors (Sanken 2SA196 / 2SC2773), I replaced them with the TTA1943 / TTC5200 pair from Toshiba;
I set DC to 0V on the speaker terminals, BIAS to 10mV on the emitter resistors, I check all the voltages starting from the differentials up to the drivers and finally to the finals; everything corresponds perfectly except for the voltage of 0.4V that I should have on the emitter of each final transistor, there is none;
I mount a dummy load and inject signal, 100mV@1kHz;
I check that there is something at the output and indeed the signal comes out both from the distortion correctors and from the finals; I bring the volume knob to 12 o'clock and leave it there for 10 minutes, noting with satisfaction that the transistors remain just the right amount of heat, they do not burn; the oscilloscope tells me 1.45Vac rms, which at 8 Ohm is about 0.2W!
how is that possible? this is a 90W beast, how is it possible that at half volume it only gives me 0.2W? I don't know, I disconnect the dummy load and put in some speakers, I send some music and everything works, the power amps also start to heat up, I take the voltages at their bases and I find 550mV, they seem a little low, so I raise the bias a bit and bring it up to 600mV (as per the service manual);
in the meantime the bias current has reached 80mV, I don't like it, the power amps start to heat up a bit more, I go back down with the voltages and I'm left with 15mV of bias and 560mV at the bases of the power amps;
at this point I let everything go at a high volume for about 40 minutes, and finally my AU-D9 works!
I wanted to ask you, with more experience and knowledge than me, considering that the finals are not the original ones, therefore with parameters outside the project, would there be a way to understand if this bias current that I set is good or not?

Thanks

sorry for my english but google is helping me