Yamaha B-3 V-fet restoration

Sweet, good to know for the future as you say. :)
On a similar note, what are your thoughts on the TO-202 transistors on the amp boards? Are these part of VAS, or for bias generation?
As they are low-CoB parts I've been wondering if that spec is important for the sub or not. TO-126 seems to give the closest matching specs all round, but I'm not sure if heat dissipation is important given the factory package, which is surely better with TO-202 than sealed-back TO-126.
Appreciate your thoughts as always.

I've edited the posted schematics to reflect the mistakes noted so far, and added the suitable modern subs.
I'll post them for everyone as soon as I'm sure they're actually correct.

TR204 on regulator board is unidentified, from the schematic I'm guessing it's the complement to TR209 (2SA872A), so 2SC1775A?
 
TR204 on regulator board is unidentified, from the schematic I'm guessing it's the complement to TR209 (2SA872A), so 2SC1775A?

Exactly; it's a mirrored circuit, so by given parts numbers of only one rail one could figure out the other half ;)
 
It's beyond this place to explain the B-3's VAS stage in detail, but let's summarise it as a enhanced current-mirrored cascoded VAS. The true VAS transistor here is TR505, but controlling the current (swing) only, while its cascode partner TR507 provides the voltage swing. This provides a number of advantages, such as that the TR505's capacitance has no influence, it increases the bandwidth dramatically and linearizes the VAS stage as well (lower inherent distortion). TR507 is operating in saturation mode (and Vce(sat) must therefore be LOW), C511 is the dominant capacitor. D505/D606 ensure symmetrical clipping. The VAS itself is (current) mirrored to improve PSRR to which it otherwise might be sensitive to.

So, those TO-202 (TR506/507/511), with nasty rare EBC pin layout.

TR506/507 = 2SA818 = 150V/120MHz/50mA and Cob of 5.0pF..... note that low Cob at that high rated voltage... (which it ought to be for the cascode function). Dissipated power is around 350mW.

There shouldn't be an urgent need to replace them; they are quite robust transistors. And the swapped pin layout only makes things uglier.
TO-126 subs could be KSA1142 or KSA1381, but again.... not needed and installation will be ugly.
 
Last edited:
So, those TO-202 (TR506/507/511), with nasty rare EBC pin layout.

TR506/507 = 2SA818 = 150V/120MHz/50mA and Cob of 5.0pF..... note that low Cob at that high rated voltage... (which it ought to be for the cascode function). Dissipated power is around 350mW.

There shouldn't be an urgent need to replace them; they are quite robust transistors. And the swapped pin layout only makes things uglier.
TO-126 subs could be KSA1142 or KSA1381, but again.... not needed and installation will be ugly.

2N5401 overkill?
TO-92 EBC
160V/100-400MHz/600mA/625mW/Cob 6pf
Low collector saturation voltage : VCE(sat)=-0.5V(MAX.)
 
Last edited:
2N5401 overkill?
TO-92 EBC
160V/100-400MHz/600mA/625mW/Cob 6pf
Low collector saturation voltage : VCE(sat)=-0.5V(MAX.)

Probably a good sub, but 350mW is a lot of heat for a TO-92 to dissipate and it MUST therefore be provided with a (clamp-on) heat sink.
Which boils down to..... the transistors on the same circuit location in the B-2...... which were too small for operating without a heatsink.
 
Probably a good sub, but 350mW is a lot of heat for a TO-92 to dissipate and it MUST therefore be provided with a (clamp-on) heat sink.
Which boils down to..... the transistors on the same circuit location in the B-2...... which were too small for operating without a heatsink.
crazy that someone would design a TO-92 that can disipate 625mW. You think they would have thought of the heatsink issue

there are many other transistors in a TO-92 case that can, on paper, dissipate even more than 625mW
 
yes, they might dissipate it.... as long as you evacuate it..... (as is the case for all power transistors/FETs BTW)
TO-92 by themselves have too little surface to transfer 350mW to ambient air and will burn out.

I assume that's why Yamaha turned to TO-202 style transistors when designing the B-3 after the B-2, having metal fin trannies in, which also would have been some good cost cutting compared to the former B-2 which had additional heatsinks added to the small TO-39 transistors, also requiring manual labour to solder them in (as where the TO-202 could be automated soldering).

Anyway...labour is for free in our case, so you can put a TO-92 in as long as you find a decent little heatsink clamped on to it....they exist !

F7124320-01.jpg
 
This thread is pure gold for me as I intend to do the same treatment to my B-3 this summer... As far as I know, this is the first documented thread about Yamaha B-3 restoration.
I want to be sure that I get all the correct info, so I have some questions:
- how did the big power supply caps capacitors (27000 microF/63V) perform on measurements?
- transistors TR301,304,306 are 2SC1175 or 2SC1775? It appears to be 2SC1175 from the pictures.
- capacitor C310 is what? I cannot see from the schematics...
- the relay from the Muting board NA 06961 is replaced by?
Thanks, vintagejapan!
P.S.: I was not able to find a second B-3 here in China. They were sold like hot cakes...
My apologies, been away and have not replied to the thread.
I replace the 2 filtering caps. There was one leaking.
The relay on muting board is a MY4 24Vdc, I replaced with HC4 DC24V relay. The transistors were NPN 2SC1775, capacitor C310 electro 10uf/16v
 
Which ones? I can see if mine have more legible writing.
I could make them out on the other board. 2SA673. The relay I got it is a very tight fit. Needed to cut flash all the solder points and even then it slightly touches the volume road.
 

Attachments

  • IMG_2572.jpg
    IMG_2572.jpg
    71 KB · Views: 66
I was am just being lazy, but I guess I will have to open one of my units to get the actual dimensions. It looks like the originals were ~63mm.
 
I have yet to hear how DC sounds :)
:rflmao:

Bass sound tighter and slightly deeper to my years with the new cans! The reason I looked in the first place was because turning up the volume made the highs slightly grainy... Was no visible signs of leakage but when out and measured one cap was way off specs
 
vintagejapan, looks great! I am sure everyone understands your assertion regarding the sound with the brand new filtering capacitors. I hope the B-3 performance is way much better than before the refurbishment.
If you have more pictures with the internal boards of the B-3 to see what electrolytics you installed, I will be delighted to follow your updates.

As a curiosity, how do you open the metallic top lid of the B-3? There are no screws visible. I have opened many times the bottom lid (easy procedure), but never measured the main caps dimensions.

What preamp and speakers you pair with the B-3?

I know, too many questions...Thanks!
 
Back
Top Bottom