SONY TA-5550 VFET amp - need help

@MattiaM your analysis makes sense, so today I replaced temporarily the transistors on the PSU board with other similar ones, just to see if anything changed. I used some MJ for the TO220 types, and BC639/640 for the rest. Unfortunately voltages did not change at all.

I noticed that by replacing R407 (4.7ohm) with a slightly higher resistor (~6ohm) and voltages got much closer. I wonder if this modification may have some benefit or it's just pointless.

I'm getting ready for inserting the VFETs back in their sockets. Today I tried to add 68 ohm 5W resistors on +-53V rails in order to limit the current in case anything goes wrong, I noticed that on the negative rail I normally have -55V but now I get 52.5V. Not exactly sure why, but if I remove connector 1 (specifically the black wire, which is a ground connection) from the class B/Protection board voltage goes back to -55V.
 
Could some unbalance in the differential pairs in the class A board load the +60V supply rails differently and, due to regulation of the 60V, end up pulling down the +82V rail? :idea:
 
FWIW, in many Sony sibling amps, the 1T243 zener diode was causing issues, a 7.3V~7.8V zener diode. Can you measure the voltage across D401 please ?

And, can you tell what character is printed on it? That actually defines the real zener voltage.
M ?
 
Could some unbalance in the differential pairs in the class A board load the +60V supply rails differently and, due to regulation of the 60V, end up pulling down the +82V rail? :idea:

I removed all the transistors in the class A amp section and totally disconnected class B amp board. Good thing I have bought a nice desoldering station that makes this kind of experiments much easier and almost effortless. So, there was practically no load on the amp. The Rails voltage imbalance was still there. What changed things was removing R410(2.7K, 2W)... without it the +-82 voltages are perfectly symmetrical. That resistor tests perfectly at 2.7K. There was still a 1V imbalance in the +-60V rails.

Since I could not find any issue with anything else (also swapping the mylar caps did not change a thing) I put all transistors back in place.
 
FWIW, in many Sony sibling amps, the 1T243 zener diode was causing issues, a 7.3V~7.8V zener diode. Can you measure the voltage across D401 please ?

And, can you tell what character is printed on it? That actually defines the real zener voltage.
M ?

The voltage drop is 9.5V. On the diode, besides the model code, there's printed 2549. Unfortunately I don't have any 7.xV zeners around, so I can't make a test right now and see if this changes anything.
 
The voltage drop is 9.5V. On the diode, besides the model code, there's printed 2549. Unfortunately I don't have any 7.xV zeners around, so I can't make a test right now and see if this changes anything.

The 7V zener Oilmaster is the 2547, so no need to look for a 7v zener.
The 2549 is a 9.1v, I believe. I measured the zener voltage for a couple of them and it was at 9.5v, so yours should be good.

There was still a 1V imbalance in the +-60V rails.

I too, am having some issues figuring unexplicable voltage differnces in the PSU of my TA-8550. While doing research on the subject I ran across a writeup from EW that might be applicable here, if the 5550 psu is anything like that of the 8550.

http://audiokarma.org/forums/index....aught-for-sony-tan-8550-rebuild.587938/page-3

read post #45
 
I too, am having some issues figuring unexplicable voltage differnces in the PSU of my TA-8550. While doing research on the subject I ran across a writeup from EW that might be applicable here, if the 5550 psu is anything like that of the 8550.

http://audiokarma.org/forums/index....aught-for-sony-tan-8550-rebuild.587938/page-3

read post #45

@rottalpha thank you for the suggestion. I already had all 1/4w and 1/2w resistors replaced with good quality ones (1% tolerance).
I also had already tried replacing the transistors, but all of this did not fix the issue. The only original components that are left are mylar caps, and I'm thinking of replacing them as well as soon as I arrange a new order on Mouser.

I think that maybe at this point I can put the VFETs back, as the imbalance will not affect them too much (they likely won't be killed) and see what happens. Maybe it's like the TA-4650 and the TA-5650 models that, when the output transistors aren't in place, can show some voltage fluctuations unless you do the diode modification by adding a diode on the PSU board.

Unfortunately, on this specific amp there's not the extensive literature that exists for the other models mentioned above.
 
Did you check whether the differential pairs Q101 (the two sides of the IC), Q102/103 and Q105/106 are closely matched for beta?
 
Vgs values are very close.

If the Vgs at the vfet pins are close, that is because the DC offset pot compensates potential mismatches in the differential pairs. However, potential mismatches, while compensated downstream, might result in uneven loading of the power rails upstream. At least this is my guess. If of course you already matched the transistors for hFE, then this isn't the issue and the problem lies somewhere else.
 
@MattiaM All the transistors are original now. They're not perfectly matched but they're quite close.

Anyway I eventually put the VFETs back and the amp is now working. I was able to set DC offset to 0V and I set the bias to the conservative value of 35mV.

The voltage imbalance remains, as one Rails stay at +80 and the other at -82.
 
@MattiaM All the transistors are original now. They're not perfectly matched but they're quite close.

Anyway I eventually put the VFETs back and the amp is now working. I was able to set DC offset to 0V and I set the bias to the conservative value of 35mV.

The voltage imbalance remains, as one Rails stay at +80 and the other at -82.
Congrats!
 
Hi all - I just bought one of these - along with a TA-4650.

If there are any basic checks that I need to do on this - it would be great to know. Unfortunately it probably wont be in my actual hands until a couple of months or so.

From reading this - it appears that I need to replace the death diodes (with paired 1n4148's) and D401.

I want to check all voltages with the VFETs removed. Is this straightforward to do?

cheers
 
Back
Top Bottom