Time to unveil a new Yamaha V-Fet transistor, named 2SJ24A

Unda Maris

Mr. V-FET
Just another VFET ? Yes, it was time (only a few weeks until Christmas).

Where did I find it ? ..... In the Yamaha B-1 ! ..... Say what ?

It was clear, that Yamaha used the 2SK77 mammoth V-FET transistors for the output stage and 2SK75 as drivers in various stages.

No source ever mentioned the 2SJ24A to be a V-Fet, but it is definitely one.
A normal J-Fet wouldn´t stand the 150V applied to it. It´s a small signal device in a TO-39 metal can package.

And it´s a P-channel type, which shows to us, that Yamahas first generation also included such a device and not only N-channel devices.

Adding these to the list, the B-1 contains 18 V-Fets in total (of 5 types including the different rankings) ! This is no good news for people, who want to repair such an amplifier.

There is now again a little hope, that the C-1 wasn´t the only V-Fet preamplifier ever made.

Does anybody know any other gear (e.g. preamplifiers), which use this device ? Does perhaps the C-2 use it ? Does anybody have the C-2 Service manual to check it ?
 
Amazing. Two new small VFETs in the last year. Will it never end?

C-2 has only two FETs. Both are in the power supply and are TR401 & TR402 (both halves of the power supply). They are 2SK30AY transistors. I'll google them and see what comes up.

Seems the 2SK30A-Y is a mil-spec n-channel JFET. Nothing particularly special. No other FETs in the C-2. But, has a very interesting phono stage design. Looks as if it would be a great candidate to update with modern caps and quieter/faster transistors. If it wasn't for those peaky NE06001/NE06003 ICs. Probably a pair of the Yamaha specific ICs.
 
Not a single FET in the signal path, everything bipolars ?
I would expect at least a few more similarities.

That´s quite a big step:

C-1 full of FETs (139) -> C-2 two FETs.

So all-FET was obviously:

- way too expensive and/or
- not en vogue anymore for some reason,

when they introduced the B-2/C-2 combo.

The C-1 really seems to be a only-one-of-the-kind type thing and only the few people, who ever heard a B-1 and a C-1 together had the chance to get an idea, how different an all FET system (or Solid State in general) can sound.

Imagine how exclusive these two devices (and especially the C-1) are in reality ! I have to repair my C-1 as soon as possible ...
 
Other than the two ICs I mentioned above, the C-2 has four additonal ICs, but everything else is BJT. But, as I don't have C-2 to look in at the moment I don't know if these are more of the Yamaha specific ICs or something else. Perhaps someone else can pull the top (actually, the bottom) of their C-2 and let us know.

Here's a nudie of the C-2. Can you pick out the ICs and FETs?

Mirko - Cab you post a nudie of the C-1? Don't think we've ever seen on here.

C-2Nudie.jpg
 
I found the 2SK30 datatsheet, it´s a typical low transconductance J-Fet optimized for low noise applications and high impedance current sources. I´d bet it is used to bias the reference diodes of the power supply.
 
Hmm... guess I have to look around for a C-1 of my own, I have a feeling that someone won’t part with his :)

Nice pic B-2.
 
OK. Here is the C-2 PS schematic. There are the two FETs. Bias for the diodes? Seems the zener is doing that.

C-2PSSchem.jpg
 
the-real-mandak said:
Hmm... guess I have to look around for a C-1 of my own, I have a feeling that someone won’t part with his :)

Not necessarily !

At least I had no problem to give it to you as a loan ... I can imagine, that you are also curious about that all-FET sound ...

I tested my B-1 again last weekend. I adjusted the dependency of bias current against power supply voltage fluctuations.

The V-Fet equipped channel really sounds worse compared to the cold triode side. I had a friend listening to it and he made the same choice ...
The next weekend I will do distortion measurements.
I hope the V-Fet equipped side is only out of balance (different gain of the halves) and after some adjustments, the sound of both channels comes to the same level.

YB-2,

thanks you, that helps !
 
I think you might have to replace a healthy V-FET in the other channel so the design mirrors, otherwise the sound will not be the same. Also remember the burn-in effect; I guess that the sound in the "new" channel is more open.

A caravan of test/trial units, interesting.
 
Yamaha B-2 said:
Amazing. Two new small VFETs in the last year. Will it never end?

C-2 has only two FETs. Both are in the power supply and are TR401 & TR402 (both halves of the power supply). They are 2SK30AY transistors. I'll google them and see what comes up.

Seems the 2SK30A-Y is a mil-spec n-channel JFET. Nothing particularly special. No other FETs in the C-2. But, has a very interesting phono stage design. Looks as if it would be a great candidate to update with modern caps and quieter/faster transistors. If it wasn't for those peaky NE06001/NE06003 ICs. Probably a pair of the Yamaha specific ICs.
My service manual says the MC ics are fet input op amps made by sony or Yamaha, so there is a CX.... ic that crosses. The C2a replaced these ics with 4 paralled bipolar transistors. You might consider replacing the C2s stage with a similar circuit.
 
The 2SK30 is used as a constant current source for the voltage reference...it provides stability.
If I am understanding correctly Unda Maris has a B-1 repaired by substituting something for a v-fet???? I'd like to research this. The B1 is designed never to allow damage to occur to the v-fets, even in the driver stages. The most common B-1 failure is in the -200 volt power supply.
My C-1 has been sitting at work , waiting for a switch and connector overhaul. A Naktalker in NJ is sending me his C-1, so I will just have to schedule a week to repair both! Sound wise the C-1 is not as fast sounding as my C-80, but I think the fastness is really a distortion added to the treble. It has been a long time though and I'd like to sit and listen to the C1 again.
 
Last edited:
Thank you Nakdoc for your insights !

The failure mechanism I found in my B-1 is quite complicated, but should appear in many B-1 units after so many years.

First the negative charge capacitor in the -85V power supply failed in some way and completely dryred out. The impedance, which was left was some kOhms, so the supply rail saw a voltage, which atlernated between -85V and 0V.

After that, two transistors (2SA810) failed in the driver board and the V-Fets switched on. This shouldn´t be a problem, since there are electronic fuses with a tyristor like circuit firing them, but here the powerswitches were reset by this silly alternating supply rail at a rate of 100 Hz ... That was obviously too much and meant the death of the bipolar switches ... and the V-Fets ...

How to avoid this: Exchange the two 2SA810 transistors against more reliable 2N5416 and recap the main power supply, before a cap fails ...

There is another failure mode of the V-Fet devices in the B-1: They don´t have any SOA protection, so Infinitys or even worse Quad ESLs with their overvoltage protected inputs are pure poison for the B-1. The preowner had a pair of Quad ESLs, which died once every year ...

Anders:

I normally would equip the other channel also with replacements, but I still want to compare them by distortion measurements. If I had two B-1´s, the things would be much easier.

I´m sure there is no burn in phase, because the bias current to date did not drifted away. Perhaps the original working V-Fets are worn out (to many shorts by Quad ESLs or thelike). I´m sure there is an aging mechanism in V-Fets, which let´s them die after many years without any external reason. I just lost a freshly refurbished Wega V4810 on a 8 Ohm load at normal listening levels ... The mechanism could be something like electrodiffusion or electromigration in the power V-Fets, which could change the shape of their depletion area around the gate grid. This could lead to dropping bias currents and increasing leakage currents over time and if you realign the bias current from time to time, there could be hot spots on the chip, because certain regions are switched off and others are always on ... That´s only a rough idea, but could be the case. So probably every V-Fet will die, if it is used regularily ...

From my personal view, the natural place for a V-Fet now is a vault (I like to collect them) and using a replacement in the amplifier instead.
 
Unda that is not the right spirit, you only use replacement if you can't get any V-FETs. The only sound they make in a safe is if you shake it and that is not a good sport.

I did note that with the 2 B-1's :scratch2:
 
Using V-FETs is an expensive sport !

A Sony V-Fet costs about 20€. In my V-4810 are now 4 dead ones.
It was freshly repaired and recapped.
80€ for only about three hours of listening, that´s not fair ... :tears:

I sweared to preserve some V-FETs for later generations.
So the best way to do that is to put them into a safe ...

And what is the myth around the original V-FET sound if the replacement sounds better ? Why is the best sounding V-FET amplifier (The TA-N7B) only a half one ?
I´m really confused about all that V-FET business ... :scratch2:

What did you note with the second B-1 ?? :scratch2:
Do you want to sell your´s ? :scratch2:
 
Unda Maris said:
Anders:

I normally would equip the other channel also with replacements, but I still want to compare them by distortion measurements. If I had two B-1´s, the things would be much easier.

...

From my personal view, the natural place for a V-Fet now is a vault (I like to collect them) and using a replacement in the amplifier instead.
Unda, what are you using to replace the V-Fet's?
 
Unda Maris said:
If I had two B-1´s, the things would be much easier.

Kaiserslautern is a long way from Copenhagen (925 km) and it has to be fix'ed first. But you never know, it would be nice to make a shoot out between an org. and the fix.

No, not for sale :)

80€ that is about the same as a trip to the opera, also takes about 3 hours :D but no, it is not fair.
 
Mark B said:
Unda, what are you using to replace the V-Fet's?

This stays my little secret, but it has something to do with power transistors ... :D :smoke:

the-real-mandak, that would be indeed really nice ...

And you are right, I should go more often to the Opera ... :D
 
Yes, that way you don't burn those precious V-FETs. Using them for something stupid as playing music tsk tsk tsk... Unda what were you thinking :nono:

:)
 
Back
Top Bottom