I embarked in the restoration of a Sony TA-5650. It's been my first ever restoration project, and I am not an expert at all, so it's been quite a huge learning curve and looking at it now, I probably should have started with something simpler.
The amp was without V-Fets when I received it, the previous owner tried to replace them with some other FETS modifying parts of the bias circuit, but he ultimately gave up. The power supply board was also damaged, with a couple of capacitors and one transistor burnt.
I rebuilt the power supply board and applied the voltage doubler mod, I also restored the power amp board replacing the killer diodes, the resistors and cap values according to the service bulletins, and finally recapping it entirely. I tested all the transistors for hfe and made sure they were all functioning.
I surprisingly found all the new v-Fets, thanks to Johsvi from the forum who sent me 6 of them from India, and thanks to Ebay.
I mounted all the boards and checked the voltages before installing the V-Fets, to make sure I wouldn't burn them.
The power supply board runs well giving me perfect 100V and 20V;
the 44V rail runs at +/- 48V, and the 77V rail runs at about +/- 76V with the preamp connected, and about +/- 80V with the preamp disconnected and the inputs shorted. These values are pretty much the same as other people restoring these amps found, and pretty much within specs. As others here, I also have oscillation in the output.
With the preamp disconnected and the bias pot adjusted for maximum readings, I get a gate voltage of about |54|V, with the preamp connected I get about |39|V. These values don't seem too far off the |60|V and |40|V that other people read on their amps here and on Audiokarma, although I am a bit concerned about the |54|V, slightly low compared to the |60|V in the specs.
That's where my problems start: when I check the Source to Gate voltages for the VFets I can get between 3V to 10V depending on the bias adjustment pot, and that is much lower than the 0-28V that I should get. These values are the same in both channels, and generally, the two channels are absolutely symmetrical in all their values.
I kept the original potentiometers in the restoration.
Does anyone have any clue where that limited bias adjustment range could come from?
Thanks
The amp was without V-Fets when I received it, the previous owner tried to replace them with some other FETS modifying parts of the bias circuit, but he ultimately gave up. The power supply board was also damaged, with a couple of capacitors and one transistor burnt.
I rebuilt the power supply board and applied the voltage doubler mod, I also restored the power amp board replacing the killer diodes, the resistors and cap values according to the service bulletins, and finally recapping it entirely. I tested all the transistors for hfe and made sure they were all functioning.
I surprisingly found all the new v-Fets, thanks to Johsvi from the forum who sent me 6 of them from India, and thanks to Ebay.
I mounted all the boards and checked the voltages before installing the V-Fets, to make sure I wouldn't burn them.
The power supply board runs well giving me perfect 100V and 20V;
the 44V rail runs at +/- 48V, and the 77V rail runs at about +/- 76V with the preamp connected, and about +/- 80V with the preamp disconnected and the inputs shorted. These values are pretty much the same as other people restoring these amps found, and pretty much within specs. As others here, I also have oscillation in the output.
With the preamp disconnected and the bias pot adjusted for maximum readings, I get a gate voltage of about |54|V, with the preamp connected I get about |39|V. These values don't seem too far off the |60|V and |40|V that other people read on their amps here and on Audiokarma, although I am a bit concerned about the |54|V, slightly low compared to the |60|V in the specs.
That's where my problems start: when I check the Source to Gate voltages for the VFets I can get between 3V to 10V depending on the bias adjustment pot, and that is much lower than the 0-28V that I should get. These values are the same in both channels, and generally, the two channels are absolutely symmetrical in all their values.
I kept the original potentiometers in the restoration.
Does anyone have any clue where that limited bias adjustment range could come from?
Thanks
You can power up the amplifier with no load and without VFETs and check the bias adjustment range (pot fully one, then the other way). This is really the only good way to know. When the VFETs are in and if there are oscillation problems, the power supplies will change (oscillation creates a fake load) and also an AC component on the measurement points will usually confuse a normal multimeter.
Secondly, if you get VFETs from ebay (especially from China), do not trust the rank numbers. They simply print a number on there (and unlike normal markings it's very easy to rub off). Not only that, they will also re-label 2SK82/2SJ28 to 2SK60/2SJ18 and the other way. At the moment the only trustworthy VFETs are those pulled from a known working unit, or ones tested with a curve tracer.
How have you distributed rank 55 and 56 VFETs amongst channels? Both of these should have a Vgs of >10V when the idle current is set to 100mA. Rank 56 should have a higher Vgs than 55 when you measure it in the amp.
Secondly, if you get VFETs from ebay (especially from China), do not trust the rank numbers. They simply print a number on there (and unlike normal markings it's very easy to rub off). Not only that, they will also re-label 2SK82/2SJ28 to 2SK60/2SJ18 and the other way. At the moment the only trustworthy VFETs are those pulled from a known working unit, or ones tested with a curve tracer.
How have you distributed rank 55 and 56 VFETs amongst channels? Both of these should have a Vgs of >10V when the idle current is set to 100mA. Rank 56 should have a higher Vgs than 55 when you measure it in the amp.
Tonight I will check the Vgs voltages with an oscilloscope. I just found that there was oscillation with the simple frequency monitor of my multimeter.
The DC multimeter, with no V-Fets installed, only gives me those 3 to 10V ranges for both channels when I adjust the bias pots from min to max. For the rankings, I will make sure the same fets have the same ranking in each channel, but unfortunately I could not find all 55 or all 56 ranks. I might have another amp for parts arriving soon, which I hope will help me with the matching.
I got the V-Fets on Ebay from Estonia (I did not trust the Chinese stuff fond on Alibaba or Ebay), and when I tested them with my Chinese transistor tester they all checked as double diodes, like the original ones. Also, the resistances and voltages are within the specs in the service bulletins, so I am fairly positive they are real new old stock as advertised and that nobody messed up with the ranking numbers. Fingers crossed there!
The DC multimeter, with no V-Fets installed, only gives me those 3 to 10V ranges for both channels when I adjust the bias pots from min to max. For the rankings, I will make sure the same fets have the same ranking in each channel, but unfortunately I could not find all 55 or all 56 ranks. I might have another amp for parts arriving soon, which I hope will help me with the matching.
I got the V-Fets on Ebay from Estonia (I did not trust the Chinese stuff fond on Alibaba or Ebay), and when I tested them with my Chinese transistor tester they all checked as double diodes, like the original ones. Also, the resistances and voltages are within the specs in the service bulletins, so I am fairly positive they are real new old stock as advertised and that nobody messed up with the ranking numbers. Fingers crossed there!
I rechecked all the voltages with an oscilloscope, and it seems that, weirdly enough, the problem was that I was using as testing ground with my multimeter the speaker ground instead of the check points on the power amp board. When I plugged the oscilloscope, using the ground on the power amp board, all voltages checked nicely (still with that oscillation behavior and the drop in voltages when shorting the power amp inputs or connecting the preamp). The bias range would go from about 0 to 30V.
I went and put back the vfets, adjusted the bias, checked briefly that it worked properly with the oscilloscope and a signal generator, and the rest is just incredible sound.
I am glad I went for it, and it's just become my main amplifier.
I went and put back the vfets, adjusted the bias, checked briefly that it worked properly with the oscilloscope and a signal generator, and the rest is just incredible sound.
I am glad I went for it, and it's just become my main amplifier.
CAREFUL!!!!!! wih the oscilloscope ground, as it is connected to earth, and if the amp has an IEC mains socket, the amp ground is ALSO connected to earth. This means that the oscilloscope ground mus either be isolated from earth if you want to connect it anywhere (NOT RECOMMENDED!!!!) or you must connect the oscilloscope ground to the amplifier ground and not any ground either, as you are automatically introducing a ground loop. And if you connect the oscilloscope ground anywhere in the amp, that point will be shorted to the amp ground through the oscilloscope-amp earth connection.
Also, on the 5650 power amp board the power and signal ground are completely separate and come from different ground points in the amp, so when measuring it's not at all the same which one is used (especially for a scope ground). In fact, if you accidentally disconnect the input connector, the amp loses it's signal ground and goes completely crazy.
Also, on the 5650 power amp board the power and signal ground are completely separate and come from different ground points in the amp, so when measuring it's not at all the same which one is used (especially for a scope ground). In fact, if you accidentally disconnect the input connector, the amp loses it's signal ground and goes completely crazy.
See service bulletin 62 for biasing new FETs.
Sony TA-5650 - Manual - V-FET Integrated Stereo Amplifier - HiFi Engine
Sony TA-5650 - Manual - V-FET Integrated Stereo Amplifier - HiFi Engine
Dear Ilimzn, you are a treasure for us, eternal beginners!
Just remain patient and tolerant while reading the stupid things we sometimes do as careless children...
Thanks for your permanent help and support.
Best wishes,
M.
Note: my amp lacks ground pin on power cable. I imagine this is standard.
I second that!
Dear Ilimzn, you are a treasure for us, eternal beginners!
Just remain patient and tolerant while reading the stupid things we sometimes do as careless children...
Thanks for your permanent help and support.
Best wishes,
M.
Note: my amp lacks ground pin on power cable. I imagine this is standard.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.