Sony TA-F6B PSU repair

Jump to Latest
johnm

johnm

Member
Joined 2003
#1
Hi folks!

I have a Sony TA-F6B which I recently was given, virtually for free. It's a lovely looking - and by all accounts sounding - amplifier so I'd like to get it working again.

The last owner said it made a loud popping sound one day, and hasn't worked since. After taking the amp apart, it's obvious the owner before him either did some tinkering, or gave it to a tech who cut a few corners...!

I have traced the problem to the PLPS (Pulse Locked Power Supply - basically a late 70s Sony SMPS I think). A heatsink/clamp to four power transistors was not replaced by previous owner/tech so in the end two of these met their maker! Worse he replaced two of the transistors with a different make, and all four are meant to be hFE/batch matched. There is also evidence of lifted solder pads and a jumper wire too in the following pics:

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


Here's what it should look like with the missing clamp:

http://picasaweb.google.com/895720/SONYTAF6BPLPS?feat=directlink#5325951198508682210

So basically, I need to know what an ideal modern substitute for the 'R' hFE grade, and 'O' hFE grade Sanken 2SC2023 power transistor would be please, as the original devices seem next to impossible to get hold of now.

Any help/advice would be gratefully received 🙂

Thanks,

- John
 
Last edited:
#3
Those transistors with gold leads are bad news... look to replace them.

Fortunately there should be loads of suitable TO-220 transistors now that SMPS are all the rage. Fairchild are probably best to look at for this. The high Vceo and fast(ish, for those days) switching speed will be whats important here.
 
#6
Also just to add that 99% of SMPS failures are down to electroylitics developing high ESR (drying out).

It would be a big help to see the circuit... those 2335's were used from memory in the Sony SLC9 and SLC7 VCR PSU's
 
#8
Thanks for the suggestion guys - much appreciated 😀

I'm really no more than a 'solder by numbers' DIY guy so lack the knowledge about what parts are best to substitute.

The caps on the main board will be replaced with Panasonic and Rubycon 105C long life types (mostly FC/FM and ZLH low esr types).

Here's a scan of the PSU (or PLPS!):

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


- John
 
Last edited:
#11
Hi johnm

I'll try and have a look at the PSU in more detail later...
SMPS can be drastic in their failure modes, and it might help to see if the PSU would run off load and with a "bulb" limiter in the incoming mains supply. It's worth removing (if needed) semiconductors and checking accurately all the resistors as any that see a high voltage can be prone to change value and go high.
 
#12
I think I can stitch the PSU together from the hifi engine link... it's the only way I can visualise what's going to have it complete on paper 🙂

I'll have a gander over the weekend hopefully... fixing stuff like this used to be the day job lol
 
#13
Oh that's very good of you Mooly - thank you Sir! 😀

And I know what you mean about having it out in front of you to make sense of it - I already have pages of photos and scans all covered in notes haha!

Cheers again!

- John
 
#14
Looking from the circuit, Q601 and Q613 are just pass transistors - so as long as you've got a transistor with a high enough Vceo and Ic here, you'll be fine.

The other transistors are the primary side switching transistors in the SMPS. These are probably a bit more critical. I would say the BUT11A transistor that Mooly mentions would be suitable.

One advantage of old circuits like these - parts such as the 2SC2023 would have been top spec at the time, but now SMPS's are so common, so are equivalent (and evenb superior) specced transistors.
 
#15
jaycee said:
Looking from the circuit, Q601 and Q613 are just pass transistors - so as long as you've got a transistor with a high enough Vceo and Ic here, you'll be fine.

The other transistors are the primary side switching transistors in the SMPS. These are probably a bit more critical. I would say the BUT11A transistor that Mooly mentions would be suitable.

One advantage of old circuits like these - parts such as the 2SC2023 would have been top spec at the time, but now SMPS's are so common, so are equivalent (and evenb superior) specced transistors.
Click to expand...

Yeah it's pretty old - the Sony TA-F6B amp came out 1977, and was manufactured until 1981 (Sony TA-F6B on TVK) so I'm hoping it will be fairly straight-forward to sub in new/better parts and get her working again.

Here's a cheeky link to another thread I started as 'ST70fan' over at Audio Karma yesterday (there's a few TA-F6B owners there):

http://mail.audiokarma.org/forums/showthread.php?p=4297238#post4297238

Once I have a consensus on here which parts are best to sub for the transistors I'll get to work on it - afraid I totally lack knowledge about transistor subs alas. Really grateful for all the help here 😀
 
Last edited:
#16
The main reason they specify the same hFE on Q609-Q612 is because they form a self oscillating circuit. If the transistors were wildly different on gain, the circuit would oscillate unevenly - or may not at all. Gain should not be a problem on modern transistors
 
#17
I've printed the PSU out... I'll see what I can come up with as regards transistors and a means of testing.

You need to 100% sure all the resistors are OK, particularly the 0.22 ohm (R622 is it... bit blurry). Check them all.

I think a small 60 watt or filament bulb would be OK as a light (ha 🙂) load across the secondaries with the feed to the amp disconnected (looks like plug/socket... if not then by lifting L607 and 608).
 
#18
jaycee said:
The main reason they specify the same hFE on Q609-Q612 is because they form a self oscillating circuit. If the transistors were wildly different on gain, the circuit would oscillate unevenly - or may not at all. Gain should not be a problem on modern transistors
Click to expand...

I think that explains (along with the missing clamp/heatsink) why this PSU blew as the parts seem to have been placed with no heed given to the service manual at all. Q609-Q612 are currently: 2SC2023-0p, 2SC2023-R, and two NEC C2335s which are from different batches. I guess it probably was oscillating, and over-heating too :s

First thing first - will desolder all those transistors over the weekend, and really give the board a really good clean up, and repair any damaged/missing solder pads. Will then be in a better position to install new parts.

EDIT: Q602 (the one to the right of the two red ones in the pictures) looks a bit dodgy too so will replace that also - it's originally a 2SC1775F for the UK model.
 
Last edited:
#19
That's it for tonight 🙂

Check all those passives john... hopefully this can be a success story.

Do we assume the amp proper is Ok... have you checked the audio outputs just for interest to make sure they are not short.
 
#20
Mooly said:
I've printed the PSU out... I'll see what I can come up with as regards transistors and a means of testing.

You need to 100% sure all the resistors are OK, particularly the 0.22 ohm (R622 is it... bit blurry). Check them all.

I think a small 60 watt or filament bulb would be OK as a light (ha 🙂) load across the secondaries with the feed to the amp disconnected (looks like plug/socket... if not then by lifting L607 and 608).
Click to expand...

Wow speedy gonzales haha! Thanks Mooly - so you mean (once all fixed up again) I'd apply power to the PLPS and then see if the bulb lights up or not collected to the secondaries?

Is it straightforward to test the resistors when in circuit? I thought some of them might have different values as they're all interacting?

p.s. the photo back in post.8 is clearer than the scan in that manual.
 
You must log in or register to reply here.
Top Bottom