DIY Sony VFET Builders thread

Dam it hot

Bias is 20.02 and 20.03 going to turn it off. probably won’t be able to look at it till Monday or Tuesday any ideas what might be wrong.
 

Attachments

  • 01F8213A-C827-419E-B1B9-E4078164532E.jpg
    01F8213A-C827-419E-B1B9-E4078164532E.jpg
    1 MB · Views: 382
Yes, the T-bracket should not be at 105°C. The bias current seems to be correct. That leaves possible issue with the thermal connection between the transistors and T-bracket, or T-bracket to main heatsink. Or the IR thermometer, but I don't recommend a finger touch if they are really that hot.
I found that I was able to tighten the transistor mounting screws in my amp a little more by hand. If you have a torque driver, try 8 inch-lbs. It looks like you already have two washers under the bracket mounting screws, so try three. We want to see a good amount of thermal goop squish out from under the bracket when the screws are tight.
 
Bias is 20.02 and 20.03 going to turn it off. probably won’t be able to look at it till Monday or Tuesday any ideas what might be wrong.

Well, the reading clearly marks 105° C!!! So yeah, other than Papa’s suggestion to point at your forehead and read something like 36C, I agree that powering off, let it cool, and then turning back on and keeping your finger. Remove as soon as you think it’s getting out of hand. Touching a 105 surface is NOT a good idea.

Does your heat sinks heat up at all? Did you put the thermal paste on the L bracket? If there is some sort of dissipation other than the L bracket, you should not see those temps.

But with those temps, I question if you would even feel the insane heat with your hand where you have it. Can you feel it hotter than boiling water?

Hope you figure things out! Good luck.
 
Vds of 20.02V and 20.03V as well as Iq of 1.22V/0.75R=1.63A and 1.23V/0.75R=1.64A are right on the money.

However looking at the picture, it seems to me that the T-bracket is not snug to the heat sink. There appears to be a small gap between the bracket and heat sink. Add another washer if the screws are already bottomed in the holes, to increase the clamping pressure to close the gap, to improve the heat transfer between bracket and heat sink.
 

Attachments

  • 01F8213A-C827-419E-B1B9-E4078164532E.jpg
    01F8213A-C827-419E-B1B9-E4078164532E.jpg
    976.2 KB · Views: 431
Either the T-bracket was not torqued enough or not enough thermal heat sink paste. I never had problems with having to double up the washers where the bolt were too long. I went with the hex bolts that came with the kit as you will get more torque pressure than bolts having phillips heads.

I hope my photos don't go distorted below:

pnOj9QRh.jpg



y2X7gwHh.jpg
 
I may be wrong, but I thought that the "Heat transfer paste" of whatever type chosen, was primarily to fill the (tiny, tiny, tiny!) voids left by any very slightly uneven pairs of mating surfaces, whilst still allowing metal to metal contact of those two surfaces over all other areas of those surfaces. (The void infill material having better heat transfer properties than air). This "granular" stuff, unless a better conductor than the parent metals seems a disadvantage, to me at least as it is distancing the two parts.
 
I may be wrong, but I thought that the "Heat transfer paste" of whatever type chosen, was primarily to fill the (tiny, tiny, tiny!) voids left by any very slightly uneven pairs of mating surfaces, whilst still allowing metal to metal contact of those two surfaces over all other areas of those surfaces. (The void infill material having better heat transfer properties than air). This "granular" stuff, unless a better conductor than the parent metals seems a disadvantage, to me at least as it is distancing the two parts.

Yes that's what I gleaned from reading up on this "paste",very thin to fill the imperfections only.In addition there are cpu paste tests that show some pastes claims are way out to lunch,the one supplied with the kit is one of them.
 
Also, there are only four screws holding the aluminium "T" piece, thick goops spread unevenly will not be able to be compressed enough to allow the best heat transfer and will just introduce gaps and voids. This may be possible to be reduced by "Wringing" the surfaces together as done when joining gauge blocks. But this is surly taking things much too far 🙂
Just put less in of a thinner material.
For those who have a "Sheldon" like mind..

Wringing Gauge Blocks - YouTube

If that doesn't sent you to sleep, nothing will!
 
Last edited:
A GOOD IDEA THAT WASN'T ONE

Back on tweaks...

I have purchased switches etc. to enable my very own FE boards: the non existing one. I usualy don't need that much gain, hence the idea to have a bypass switch to take the FE boards out of the loop when not needed.

Sure that could only enhence the sound. Sure, with a tranny in the signal path, and a big cap (eventhough I bypassed it with a PPP caps), thes FE boards must rob a lot of dynamic, details, cap the frequency response on both ends...

Well, I was taught another lesson by Papa, yet again. These FE boards aren't cheap add-on to get more gain.

We bypassed the FE boards, add 14dB gain and invert polarity at the source. We listen as hard as we could and... could hear NO difference.

OK, PERHAPS, there is a VERY MARGINAL plus in transparency, completely irrelevant given the initial situation (no lack of) and again, I am not even sure. Truth is we could'nt hear a real difference. OK, perhaps listening for hours would have revealed some minor bits, who knows... but point is: it is not worth the hassle.

The FE boards that came with the VFET amps do the amp justice. I can't even say they complement the sound of the VFET amp as they don't correct it: what you hear with (or of course without) the FE boards is just... the sound of the OS.

Papa has done a splendid job, don't get fooled by the ingredients or the simplicity of the FEs, Papa is an excellent cook... but then I should have known, stupid me

Moving on to the next mod... as if bypassing the FE boards isn't really worth doing (it can't harm either BTW), the VFET amp still lacks somewhat in my set up and in my ears speed / rising front / slam (call it what you want). In that regard, going for a "fast" FE that would impact the sonic signature could be a plus, and they are enough FEs to chose from. For the moment I would like to keep the amp as is, relaxed, but add speed... spoiler, that's the next mod

Claude
 
I admit ZM, I admit completely. 🙂))

I have no chance whatsoever to beat the Master, I hoped that I could perhaps take a set or two, not even that. Papa is Djoko...

I went the trouble to try to get this excellent little help, so consistant in its sound, even closer to my liking. It is very personal, and TBH there isn't much to tune anyway, more me playing again.

But...Now ZM, just for you, thinking again and considering your post... in fact perhaps my FEs are so good because... I bypassed a cap and made them hence perfect!!!! GLOL!

Nah, still learning from the Master, at least the lesson came for 10$ and somewhere I must confess that on second thought I am delighted it turned out that way, with Papa making so many things right for us... out of the box. Lucky us...

Enjoy music

Claude