One more thread from the AU 4900 and they go....Over the years, this amplifier becomes noisy, generating random crackles. It can work well for a few hours or many days, and at some point begin to generate violent bangs, crackles. It can happen on only one channel, and then appear (or not) on the other.
We usually find the advice on the web to replace the TR 2SA798, pointed out as the only and unequivocal culprit. We will see that it is not so.
It is a differential pair, two TRs in a single package, 5 pins, which is why in the Sansuí manual it is called IC (integrated circuit). These are on the power board, it is the first input TR.
These TRs are impossible to get new, so there are two solutions, build it using two KSA992FB transistors, (the most recommended by the repair community) joined with heat shrink and thermal paste, or use the dual BCM856DS TR, for the latter a adapter for easy assembly. Given the availability here, I chose the BC557B, purchasing 40 units and selecting two pairs with the most similar HFE. I have seen tutorials from repairers who have bought these very damaged amplifiers, only to remove the 2SA798, but it is no guarantee for what has already been said. Of course, I don't believe that the entire Sansui AU4900 production with these TRs failed, but it seems that the failure rate was quite high, they were even used as well.
(maybe the same batch) on Pioneer amplifiers, with the same problems. We are talking about audio equipment from the 70s. There are people who never repaired the AU4900 (or also AU3900) and still enjoy it. Many people say that this amp. It has a "tube sound", something very debatable. But it is very complete, it has tones, filters, loudness, attenuator, two speaker systems, headphone output and also a practical microphone mixer. With good quality speakers, it really sounds great.
We usually find the advice on the web to replace the TR 2SA798, pointed out as the only and unequivocal culprit. We will see that it is not so.
It is a differential pair, two TRs in a single package, 5 pins, which is why in the Sansuí manual it is called IC (integrated circuit). These are on the power board, it is the first input TR.
These TRs are impossible to get new, so there are two solutions, build it using two KSA992FB transistors, (the most recommended by the repair community) joined with heat shrink and thermal paste, or use the dual BCM856DS TR, for the latter a adapter for easy assembly. Given the availability here, I chose the BC557B, purchasing 40 units and selecting two pairs with the most similar HFE. I have seen tutorials from repairers who have bought these very damaged amplifiers, only to remove the 2SA798, but it is no guarantee for what has already been said. Of course, I don't believe that the entire Sansui AU4900 production with these TRs failed, but it seems that the failure rate was quite high, they were even used as well.
(maybe the same batch) on Pioneer amplifiers, with the same problems. We are talking about audio equipment from the 70s. There are people who never repaired the AU4900 (or also AU3900) and still enjoy it. Many people say that this amp. It has a "tube sound", something very debatable. But it is very complete, it has tones, filters, loudness, attenuator, two speaker systems, headphone output and also a practical microphone mixer. With good quality speakers, it really sounds great.
Last edited:
In this link you can see it in its entirety.
https://rochester.craigslist.org/ele/d/rochester-sansui-au-4900-stereo/7699770713.html
https://rochester.craigslist.org/ele/d/rochester-sansui-au-4900-stereo/7699770713.html
Thank you for passing by ! This will take a few days, too many photos and things to tell.
You can see how to join the two TRs to form one, the emitter is common, so we have the five contacts of the 2SA728. I recommend using the minimum heat necessary for soldering and desoldering, the tracks are very prone to lifting, probably the adhesive used no longer has the necessary adhesion given the years that have passed, or simply because the one used on these plates was not of the best quality. (I'm sorry for fans of the brand, but it's reality, plus the tracks - rails - are made of copper that are too thin, very brittle.)
So I replaced the two TRs, the amplifier worked without crackles.....but not for long time ! The same ones came back, sometimes stronger on one channel, sometimes on the other. It is important to emphasize that they have no relationship with the volume of the amplifier. They can make you jump out of your chair even with the volume at zero. So I had to isolate the problem. Was it the power board or was it the preamplifier? Is it possible that some TRs of the same one also failed? Searching the internet, a forum highlighted that the 2SC1313 and 2SA726 TRs were also prone to generating noise, and the recommendation was to also change ALL the preamplifier TRs! ......lucky me, noises everywhere!
So now it was necessary to confirm this, was the preamplifier the new culprit? I don't have a signal generator or an oscilloscope, I am an amateur who thought I would make a living from electronics, I really like it a lot, but the twists and turns of life meant that my income came from other areas. Then the noble Nad 3020 came to my aid. The 3020 was a classic and revered amplifier of those years, with the possibility of using the pre or power separately. So I disconnected the signal r input to board power, and used the NAD pre. Successful test, no noise, and it was tested exhaustively. (that means many hours of use) So the opinions about the remaining TRs were true! Again, I looked for replacements available here in my country. Again the BC557 (I remember when I was studying electronics that a professor told us, "Phillips' BC line is "all-terrain", they serve many applications and replacements" with its complementary pair, the BC 547 (this only has 5 Volts less than VC , 45 instead of 50 volts, but the Sansuí's power supply has 35 volts +/-, there were no problems with that) came to my aid. At least, that's what I thought.....
So I replaced the two TRs, the amplifier worked without crackles.....but not for long time ! The same ones came back, sometimes stronger on one channel, sometimes on the other. It is important to emphasize that they have no relationship with the volume of the amplifier. They can make you jump out of your chair even with the volume at zero. So I had to isolate the problem. Was it the power board or was it the preamplifier? Is it possible that some TRs of the same one also failed? Searching the internet, a forum highlighted that the 2SC1313 and 2SA726 TRs were also prone to generating noise, and the recommendation was to also change ALL the preamplifier TRs! ......lucky me, noises everywhere!
So now it was necessary to confirm this, was the preamplifier the new culprit? I don't have a signal generator or an oscilloscope, I am an amateur who thought I would make a living from electronics, I really like it a lot, but the twists and turns of life meant that my income came from other areas. Then the noble Nad 3020 came to my aid. The 3020 was a classic and revered amplifier of those years, with the possibility of using the pre or power separately. So I disconnected the signal r input to board power, and used the NAD pre. Successful test, no noise, and it was tested exhaustively. (that means many hours of use) So the opinions about the remaining TRs were true! Again, I looked for replacements available here in my country. Again the BC557 (I remember when I was studying electronics that a professor told us, "Phillips' BC line is "all-terrain", they serve many applications and replacements" with its complementary pair, the BC 547 (this only has 5 Volts less than VC , 45 instead of 50 volts, but the Sansuí's power supply has 35 volts +/-, there were no problems with that) came to my aid. At least, that's what I thought.....
Attachments
Replacing TR's is not easy at all, we have to choose the parameters from the manufacturer's data sheets that are closest to the original. So my choice, given the availability here, was to go again to the European line. 2SA726 = BC557 and 2SC1313 = BC547. Since I did not find precise information about the HFE of the Japanese originals, (in a data sheet that I found after much searching on the web it said 50, much less than the ones I bought, because none were lower than 300, and others that I had saved for years , they did not go below 200)...! So, checking with a well-known technician, he told me: "Choose the TRs with the highest HFE, because the bias circuit associated with each TR is what will determine the gain, therefore the HFE of the replacement TR may be higher than the TR." original, but never lesser." Regarding this, I would like to know the opinion of those in the know here, that information would be very useful to clarify any reader who reads this and is going through the same dilemma. So I replaced ALL TRs in the preamp. I started with the first one (the signal input) and so on until the last one (output to the power board). With each pair I replaced (taking great care in the EBC arrangement, because they do not match) I put the amplifier back into operation and the crackling always continued, until the end! (my hope was that at some point the crackles would go away, then I would know exactly which TRs were bad, but that didn't happen) Instead, with each TR torque change the volume went up, slightly, but it went up! What do you think about it? Was my technical friend wrong? Everything would seem to indicate yes....😳
Attachments
" So, checking with a well-known technician, he told me: "Choose the TRs with the highest HFE, because the bias circuit associated with each TR is what will determine the gain, therefore the HFE of the replacement TR may be higher than the TR." original, but never lesser." Regarding this, I would like to know the opinion of those in the know here, that information would be very useful to clarify any reader who reads this and is going through the same dilemma."
Some thoughts ?
Some thoughts ?
There were no answers, but I found this.
" Therefore, for robust and reliable circuit designs, engineers try not to rely on a specific hFE value for the correct operation of a circuit. Instead, they design circuits in a way that they operate properly over the entire specified range of hFE values. "
https://www.ovaga.com/blog/transistor/what-is-transistor-hfe
So from this paragraph a question arises, in the seventies did designers/manufacturers take this into account? Due to the behavior of the BC547 and BC557 that I used in this circuit, I think not, at least in this Sansui AU4900. And I have a new doubt, will the highly recommended KSA992 replacing the 2SA726 and the KSC1845 replacing the 2SC1313 have an excessive current gain like the ones I used? If anyone had that experience, can they tell me?
Thanks guys for stopping by.
" Therefore, for robust and reliable circuit designs, engineers try not to rely on a specific hFE value for the correct operation of a circuit. Instead, they design circuits in a way that they operate properly over the entire specified range of hFE values. "
https://www.ovaga.com/blog/transistor/what-is-transistor-hfe
So from this paragraph a question arises, in the seventies did designers/manufacturers take this into account? Due to the behavior of the BC547 and BC557 that I used in this circuit, I think not, at least in this Sansui AU4900. And I have a new doubt, will the highly recommended KSA992 replacing the 2SA726 and the KSC1845 replacing the 2SC1313 have an excessive current gain like the ones I used? If anyone had that experience, can they tell me?
Thanks guys for stopping by.
