• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Replacing VFETs with MOSFETs

The additional supply is an ugly kludge but I need it to cut off the laterals. With the HEXFETs it wasn't needed but then I got the thermal issues. They might have been easily compensated by putting a 10k NTC across the bases of Q357 and Q359 or even coupling the two diodes with the heat sink. Also, it may just be that the parasitics I got in the pics above are caused by the MOSFET not being cut off. It then has gain. I will know tonite (EU time)...

Tjerk, 9ZZ
 
Looks like I've got a winner! Distortion 0.01% at 2 W and slightly higher at 20 W. I only have R-Ch but L-Ch is expected to have similar performance. Let's screw everything together and then do the VFET-MOSFET shootout!

Tjerk, 9ZZ
 
Very cool! Hopefully this thread will be some help to others. I've seen those threads where people just give up once they loose a VFET and can't find any.
 
Oh! Rejoiced too soon... The issue in R_Ch is still present. This has nothing to do with the MOSFETs but is caused by something else. L-Ch behaves really nice. Tomorrow I'm going to swap components. This is getting ridiculous. But then again, "It's a Sony".

Tjerk, 9ZZ
 
Bummer! Perhaps the original issue. If so, I'm absolutely really very extremely ultimately curious about what the hell that was!
 
Right now I'm thinking output relay or feedback circuit. Since the gate drive of both channels are more or less similar (except for bias) it must be in the feedback loop. That doesn't exclude the relay because the headphone output, which I'm using for the distortion analyzer, is also behind it. And if I switch off the speaker connectors the distortion disappears. Will do some more tapping (!) this PM.

Tjerk, 9ZZ
 
Right, does not at all sound implausible to me.

Both TA-4650's had the same issue. The VFET one even had it after the extensive restoration work performed on it, still existing before I brought it to you.
Now since I most likely blew up the protection transistors while probing the relay connections it might as well be that repairing that also solved the initial issue.

To a certain extent I even hope so, since that would explain the ridiculous long fault finding issue in the VFET amp.
 
PS: Another thing. Remember I told you another Dutch guy (passed away meanwhile) already did the trick of exchanging VFETs voor MOSFETs?

Well, I slept a lot of peaceful nights since that moment only to wake up in a nightmare this night realizing I did not tell you the truth...

Yes, he did bring a TA-4650 back to live, but he did not use MOSFETs, he used BJT's.

So, if you succeed in getting a truly well performing MOSFET solution to a broken VFET amp, you might be the first on earth to do that...

:)
 
No, the VFET amp had one leg of the relay shorted to ground. That's why I never found the fault in the first place. Once the relay got energized it shorted the output to GND and the overload protection circuit did its job.

About the distortion that I'm facing now: it is very different from what I pictured above. If I can't find the issue quickly tonite I'll post a pic of it.

Tjerk, 9ZZ
 
That's strange, I did not bring it with that behaviour to KT88. Well, he said he worked on it and could not find the problem, perhaps he introduced the fault, he was not at his best when he was working on it (unfortunately now still in hospital after more than 4 weeks... )
 
I'm having a Loekie de Lul (Asjemenou) moment. Tapped the board with the relay and HP socket on it, and the distortion completely disappeared! At 2 W L_Ch now does 0.010% and R_Ch 0.011%. Cause: connector not pushed fully home. Relay is not at fault. As the amp has to work harder the distortion increases to about 1/4% at the onset of clipping. I do have to fix the 2.5 V supplies because right now it's starved of current and unable to bias the gates at lower mains voltages without ripple. That is bad. At the same time can fix the mess one of the "restorers" made in the power supply. After not replacing the bad power caps he replaced the brass connection tabs with (gasp!) solder wick. Since the screws are gone anyway and the caps now have snap-in pins I'll replace it with thick copper wire.

"Asjemenou" indeed...

Tjerk, 9ZZ
 
Right, so this one;

8l9Z4YK.jpg
 
Lavane said:
Very cool! Hopefully this thread will be some help to others. I've seen those threads where people just give up once they loose a VFET and can't find any
Click to expand...
Sigh.

This was me a couple of months ago. I lost a VFET in a 4650 and sold the amplifier for the princely sum of $1NZD.
 
I just did a shootout between the VFETs and laterals:

ta-4650_shootout.jpg


The big volume control is connected to REC OUT of my trusty DFA888 and splits the signal to the inputs of both power amps. The big A/B box at the rear is my speaker switchbox that I can toggle between the two amps. Do that a few times and I won't know which amp I listen to. Then I tried to hear differences while toggling the amps. I did not hear any difference whatsoever. As was to be expected. Both amps sound equally sweet on my DIY speakers.

Now I only have to do the writeup...

Tjerk
 
You must log in or register to reply here.
Back
Top Bottom