EXPERIENCED Guidance Sought for Sony TAN 8550 Rebuild (Pt.2)

Vinylcafe

Linvin' the Dream
This is Part 2 of a thread started in April this year.

I felt it it better to start fresh and not have to wade through a long/ old dead thread - sorry for the length of this one.

This project may also end up being the poster child for the old axiom, "if it aint broke, don't fix it".

The monsoon season has set in here, so time to get back to work.

I am rebuilding the power supply board of my 8550, with new electolytics, transistors,resistors, and diodes.
Happy to supply complete list of components replaced when needed.

My challenge at the moment is that voltages are both low and high (depending on the winding in question), and I cannot determine why.

I've floated the entire board for easier access and testing. Also allows for reading directly off the transformer.

I fear the issue is as follows ... during the rebuild one of the shielded leads from the filter caps somehow nicked a post on the power supply board - odd given the metal sleeves of the connector are slightly recessed inside an insulated shield.
The contact was with the post with the green wire, supplying a nominal ~85 volts direct from the transformer.

The unit was unplugged, but the caps were still charged.
Big bang/spark ensued and a slightly melted lower post and metal sleeve inside the shielded connector resulted.

Also a resistor and diode were knocked out on the amp board (these two components are just voltage regulation for the lamp).
Who knows what else - there must have been a bit of wire or solder protruding from the bottom of that post.

My fear is I have damaged the transformer, which is right up there with frying a V-FET.

Is it possible that there is a component on the board that is somehow dropping the voltage right off the full wave rectifiers?

Am really hoping someone is willing and able to do a step-by-step walk-through of trouble shooting.

I have already taken many measurements and tested virtually all of the components. Happy to share all the data if there are any takers.

I feel like I've driven a vintage Porsche into a pole.

Hoping someone can please help out?



Here are the voltages off the transformer's various windings:

Ground>red 45.1 Vac Board marked ~60
Ground>orange 45.2 Vac Board marked ~60
Ground>green 66.7Vac Board marked ~85
Ground>yellow 66.6 Vac Board marked ~85

Measuring straight across red/orange and green/yellow without reference to ground/center tap - doubles those reading.

This was with all of these leads floated off the power supply board.

While both halves of the two secondary windings match, they appear low.
There is continuity between them, and no continuity between them and the 11 volt secondary winding for the lamp.

As for resistance in the secondary windings, they too are symmetrical:


Here are the resistance measurements for each winding:

yellow/green (80v) 4.5 ohms
yellow>center tap 2.5
green>center tap 2.5

red/orange (60v) 0.8 ohms
red>center tap 0.5
orange>center tap 0.5




Here are the voltages at C 335/336 off the 2 full wave recitfiers:


C-335 positive lead (+88v nominal) actual +62.5v
negative lead (+60v nominal) actual +72.5v

C-336 positive lead (-60v nominal) actual -70v
negative lead(-88v nominal) actual -58v

This is with no connection of the multi pin connector to the power amp board, except for a ground connection to the center tap of the transformer.

Any and all insight appreciated.


Service manual here for anyone interested:

https://drive.google.com/file/d/0B2_JfFthlwCqd0FCUFF3VUdESzg/view?usp=sharing

You can view the entire service manual here.
 

Attachments

  • _1080347.jpg
    _1080347.jpg
    83.4 KB · Views: 78
Last edited:
I doubt you damaged the transformer - if you had, you'd most likely have 0 volts on one or more points, rather than the wrong voltage. Rather, you've likely blown one or more of the transistors or zeners in the regulator circuit.

Note that the 1T243 zeners Sony used are not all the same part. This is actually a series of zeners with different reference voltages. There are two in the TAN-8550 power supply. Unfortunately, Sony didn't specify which was which in the service manual (at least in the copy I have - I was unable to open the link you posted). From memory, I think D336 is either 9.1V or 10V, while D337 is 7.5V.

Good luck - the TAN-8550 is an outstanding amp!
 
I doubt you damaged the transformer - if you had, you'd most likely have 0 volts on one or more points, rather than the wrong voltage. Rather, you've likely blown one or more of the transistors or zeners in the regulator circuit.

Note that the 1T243 zeners Sony used are not all the same part. This is actually a series of zeners with different reference voltages. There are two in the TAN-8550 power supply. Unfortunately, Sony didn't specify which was which in the service manual (at least in the copy I have - I was unable to open the link you posted). From memory, I think D336 is either 9.1V or 10V, while D337 is 7.5V.

Good luck - the TAN-8550 is an outstanding amp!

Thanks Bob, and thanks for weighing in!

I really want to get this up and running!!!


You raise a really good point on the zener diodes D336/337; it was one I was wondering about.

The voltage out of D336 is clearly marked on the schematic as 9 volts. However there is no nominal voltage indicated at D337.

I was getting 13 volts at D336 so pulled it and replaced BOTH D336/337 with a 9.1 volt zener diodes.

(Mouser Part # 78-BZX85B9V1)

Looking at the case markings of both components, they were identical with a designation of "2249" after the 1T243.

They are also both listed on the same line on the parts list - both these circumstances lead me to believe they were the same value.

I have no way of testing the forward voltage of these two components, though I know from measurements in circuit that 336 was passing too much.

Is it common knowledge that D337 is a 7.5 voltage diode? I guess I order one?


As for the five transistors replaced ... all were new or new old stock.

Q 333 > ZTX 694B (replacement equivalent as per Echowars)
Q 334 > ZTX795A (replacement equivalent as per Echowars)
Q 332 > Hitachi 2SC1060 ("C" ranking) Original ranking unknown
Q 331/335 > Hitachi 2SCA671 ("A" ranking) Original ranking "C"

Alas I cannot find the original pulls of the three Hitachi transistors (everything else I still have).

However a couple things of note ...
I have a semiconductor analyser kit that will do auto pinout ID, gain measurement, Vbe/Ib, etc.
None of the transistors show to be defective, though I have not compared the readings to a spec sheet as yet (they were all new or NOS).

The other thing that comes to mind is that the rankings of the Hitachi transistors were a different ranking than the pulls.
Believe originals were a "C" ranking and the new matched pair are "A" ranking.
Could this account for the difference in voltage?

Any and all thoughts welcome.

PS: here is the proper link for the service manual.

https://drive.google.com/file/d/0B2_JfFthlwCqd0FCUFF3VUdESzg/view?usp=sharing
 
Last edited:
Here's the PS Sch block for reference.

attachment.php
 

Attachments

  • TAN-8550 PS.jpg
    TAN-8550 PS.jpg
    65.1 KB · Views: 202
Here are the voltages at C 335/336 off the 2 full wave recitfiers:


C-335 positive lead (+88v nominal) actual +62.5v
negative lead (+60v nominal) actual +72.5v

C-336 positive lead (-60v nominal) actual -70v
negative lead(-88v nominal) actual -58v
Those should be two very dead capacitors...reverse-biased by 12V. Installed backwards?

Note that +/-88V rides ON TOP OF the +/-60V (which is why you can use a 35V cap on an 88V supply). I don't know how you tested it all, but I'd suspect the integrity of those two caps, the D331 and D334 diodes, and the R340 and R341 resistors. I'd also check those caps across the diodes (C331/C334).
Is it common knowledge that D337 is a 7.5 voltage diode? I guess I order one?
7.5V? Where does that come from?

D337 is on the emitter of Q333. 10V is given at the base. If the base is 0.7V higher than the emitter (as it must be for a functional NPN transistor), and we round that 0.7V off to 1V for convenience (as Sony is wont to do on their scats), then what's the voltage at the emitter (and thus the zener voltage)?

If you are not getting the output from the 65V regulator that you should be getting, and assuming that all the semiconductors are functional and properly installed, then understand that this circuit is a simple amplifier. Gain is given by ([R336 divided by R337] + 1) * the voltage at the base of Q333.

So, by manipulating the feedback resistors you *could* use any number of zener diodes. But your use of a big-ass 1.3W zener here means you're operating on a soggy part of the zener curve (knee). I do believe a plain-Jane 500mW zener is much more appropriate.
 
Last edited:
If the rest of the markings on the 1T243s are the same, then they should be the same voltage, which should be 9.1V. There is a 7.5v 1T243 elsewhere in the amp, or maybe just in the matching TAN 8450 preamp.

I will defer to EchoWars on the rest of what could be wrong - he's forgotten more about restoring vintage gear than I'll ever know!
 
Thanks for jumping in Glenn!

Those should be two very dead capacitors...reverse-biased by 12V. Installed backwards?

No they are in the right way and when tested on my Fluke 175, both read about 450uf. I do not have a dedicated capacitor tester.

Note that +/-88V rides ON TOP OF the +/-60V (which is why you can use a 35V cap on an 88V supply). I don't know how you tested it all, but I'd suspect the integrity of those two caps, the D331 and D334 diodes, and the R340 and R341 resistors. I'd also check those caps across the diodes (C331/C334).

D331/334 test good out of circuit. About 0.5 - 0.6 voltage drop across each. No continuity when leads are reversed.
That said, there seems to be something funny going on.
The meter sometimes shows about 2.5 volts passing the other way and a second later no continuity at all.
When placed back in circuit, one of the diodes sometimes displays 2.5 volts travelling in reverse.

R340/341 are almost dead on, as are C331/C334.


7.5V? Where does that come from?

Will forget that number.


D337 is on the emitter of Q333. 10V is given at the base. If the base is 0.7V higher than the emitter (as it must be for a functional NPN transistor), and we round that 0.7V off to 1V for convenience (as Sony is wont to do on their scats), then what's the voltage at the emitter (and thus the zener voltage)?

If you are not getting the output from the 65V regulator that you should be getting, and assuming that all the semiconductors are functional and properly installed, then understand that this circuit is a simple amplifier. Gain is given by ([R336 divided by R337] + 1) * the voltage at the base of Q333.

All semiconductors are installed properly, based on pinout from semiconductor analyzer. NOS for the Hitachis and replacemnts for Q 334/334 you suggested in April.

So, by manipulating the feedback resistors you *could* use any number of zener diodes. But your use of a big-ass 1.3W zener here means you're operating on a soggy part of the zener curve (knee). I do believe a plain-Jane 500mW zener is much more appropriate.

I will order lower wattage zeners and if you could suggest a replacement for D331/334, I could order those at the same time.

Is there anything more that can be tested or measured at this time?
 
A plain-Jane 1N1004 or UF1004 or 1N4004 will work fine.

I do NOT understand how you can get a lower voltage on the 88 volt rails than on the unregulated 60V rails.
 
Pics would help.. Not just for us viewers but I just got a new faster camera and it really helps..

Even just take a big step back and snap it and it gives you some perspective.
 
Was there something wrong with the original D336 and 337 zeners? If not, why not just put them back if you still have them. They'll probably out last the ones you're replacing them with ...
 
A plain-Jane 1N1004 or UF1004 or 1N4004 will work fine.

10-4, I will order those up.

I do NOT understand how you can get a lower voltage on the 88 volt rails than on the unregulated 60V rails.

I hear you sir!
Are there voltages we can check elsewhere in the chain to see if it's other components causing an issue?

I have attached some pics to show you what and where I am measuring.

There is a wide overall shot from the top.
The colors of the wire leads match those from the transformer to the proper posts on the board ... with the exception of the white lead, it represents the orange wire from the transformer.

Note all electrolytic caps placement match the markings on the board.
I checked them against the schematic to see if maybe the board was improperly marked. It is not, from what I am able to read.

Then there is the voltage readings I am taking ... the order of pics is as follows;

C336 minus
C336 plus
C335 minus
C335 plus

You can see the actual readings on the meter.

The new zener diodes were put in place based on the schematic, as there are no markings on the circuit board, given the originals were in a different package.
With D336 the cathode end is connected to Pin 8 on the multi pin connector and that pin is marked "+10".
D337 the cathode end is connected to the plus side of C340, emitterof Q333 and R335. So think I've installed that correctly.

I am wondering if maybe the next step it to put all the original components back in? Or start in some sort of methodical way?
 

Attachments

  • C336 minus.jpg
    C336 minus.jpg
    70.7 KB · Views: 57
  • C336 plus.jpg
    C336 plus.jpg
    65.5 KB · Views: 35
  • C335 minus.jpg
    C335 minus.jpg
    73.8 KB · Views: 36
  • C335 plus.jpg
    C335 plus.jpg
    76.1 KB · Views: 35
  • _1080360.jpg
    _1080360.jpg
    79.8 KB · Views: 53
Last edited:
Yes...yes it does.

As I stated way back in post #5, the +/-88V rides on top of the +/-60V. Therefore until these connections are made, NONE of your voltage readings are valid.
 
Yes...yes it does.

As I stated way back in post #5, the +/-88V rides on top of the +/-60V. Therefore until these connections are made, NONE of your voltage readings are valid.

Sorry about that, I didn't really know what "rides on top of" implied.
Glad we got that sorted! And thanks for checking.

Before I charge those caps up however, I did a voltage check on those two posts. It comes in at -70 and +72.

Seems high, but within the caps' spec ... good to go???

Will that drop once connected?
 
Without those caps in place you won't be able to trust what your meter is telling you because there will be lots of 120Hz ripple fooling your meter. Hopefully nothing else has been damaged as a consequence. :tears:
 
Things are looking up ....

With the filter caps now connected here are my readings across C335/336:

C335 plus side 88.9
minus side 62

C336 plus side -62.8
minus side -90

Voltage on the filter caps is +/- 63 volts.

Is this close enough to spec?
Can I leave things connected for a while now (without fear of damaging components in the PS board) and check other voltages on that board?

If so, what are the critical ones that I need to check and confirm in order to start hooking the power amp board and the left and right channel boards up (with no V-Fets yet)?

Clearly need to progress in stages here.
 

Attachments

  • _1080361.jpg
    _1080361.jpg
    91.9 KB · Views: 38
Last edited:
Back
Top Bottom