DIY Sony VFETs OS2 Official thread

I note that the Lottery Sony VFET amp hook-up wire was Silver plated copper wire with PTFP insulation. Is there a source that I can purchase some in a small quantity to hook up the amp. I can find 25’ rolls, but buying a variety of colors in 25’ rolls in two wire gauges gets expensive quick. I am assuming that one uses 18ga. and 22ga. wire for hook-up wire. Mouser requires part numbers etc. etc. Or is there another good option?
 
While I wait for some appropriate hook up wires to get started on the VFET boards I decided to practice soldering on a pair of SMPS DC P0892B Filter boards. Boy they are tiny! I am planning to use these on my ACA 1.8 mono blocks that I built in 2020. At the moment I am awaiting for the arrival of the power jacks from Amazon. I have some thin plywood that I will utilize as a housing for the two boards and try them externally between the wall warts and the ACAs. If they improve the sound as others have documented, I plan to see if I can fit them into the ACA cabinets
This evening I will put together the Scourge Front End for the Sony VFET build. I want to practice some more soldering before attempting the VFET build.
 
Well, I finished the basic amp after many misadventures that I wonder if I am a walking curse. It is my first amplifier build, so I guess at every stage I expected that something would fail, but I put out final product that works, after many emergencies and gnashing of teeth.

The preamp section of the TA 4650 is just too exhausted and shows its feeble age when put into a high def situation, so I will remove that power supply. Fortunately, the Pass Korg preamp is more than up to the task of adding some gain. I hooked the amp first to a full range ribbon@3.6 ohms@87db or so efficiency, so it was probably putting out 6 watts at best, but still sounded good within volume limitations but less than ideal for the device all considered. It didn't self destruct.

Heat seems quite reasonable at the top case grill, maybe a skosh more (but maybe not) than a properly functioning and biased standard TA 4650 amp in original condition. I can keep my hand on the case over the heat sinks as long as I want, so there at least it's not too hot at the casing.

I put it into the main system to play into the 300 hz to 7Khz ribbon @6 ohms@ maybe 95 db efficiency. Hooked up with the Pass Korg preamp, the performance is quite remarkably good. I didn't really expect it to sound so good, even without the noise reduction input cards (which I blew somehow because they didn't work, I connected SMPS directly). In this context, it's up there with Tokin SISSYSIT, with perhaps a bit more leading edge, but imaging is fantastic. Spookee, but perhaps a bit more voluptuous than what I remember of the SIT 1 and SIT 2 amps at audio shows. It hasn't even broken in yet.

My skills resemble audio butchery, but it still worked out so far. NP won't be hiring me for his factory any time soon, and this thing could still blow up, I suppose, but I'll see. I still have to re-bias it and Pass Korg, too. I am AMAZED that I got through this odd project to a successful result. If it lasts, I'll get a PASS DIY sticker for it later.
 

Attachments

  • Screenshot 2022-11-16 at 12.56.35 AM.png
    Screenshot 2022-11-16 at 12.56.35 AM.png
    199.7 KB · Views: 87
Last edited:
  • Like
Reactions: 1 user
I listened to the Pass DIY 4650 last night into the wee hours. Paired with the Pass Korg unit I built as my first DIY project, the sound is exceptionally good within its power and impedance guard rails. I'm almost a little mad at it for challenging my other great amps in this fascia.

It has 'detail with all the clothes on' and startling transient attack. Nuance, PRAT, subtle shadings, they are all there and they don't need cranked decibels to sound complete. I am a bit in awe that the little Sony device had this potential as a single ended device with MOSFET current assist. It hasn't got the slightest harshness and invites listening and curiosity about everything that comes next.

As I said, it sounds nearly as detailed as what I remember of SIT 1 and 2 at shows, but extra tonality with the Korg Pass gain stage, which is good.

I don't have the

SMPS DC FILTER P089ZB cards attached yet.​

I put them in backwards because I was confused by the card nomenclature, but I tested them '*** backwards' and the two work fine. However, I am wondering if they are redundant in this setup because the startup thump is minimal and I'm not really hearing any significant noise.
 
Last edited:
I have been working on the Missouri input card and made a big error. I soldered the TI dual op amp and the Instrumentation Amp to the board in the place of the Dip-8 sockets. OOps. So I desoldered them and put the Dip-8 sockets there in places U1 and U2. That went fine, except of course I destroyed the two amps doing it. No worries I have extras. But I also compromised the pin behind Pin 1 on the U1 placement. About 3/4 of the pre tin on the board got scraped away. But I soldered it up anyway and it seems to be OK, ie I get a connection between that socket and Pin 1 of the Instrumention Amp in socket U1. I have enough parts to make up another board, but do you think I should try this one first? I have included a couple of pictured os the pin.
 

Attachments

  • 735C6A3D-C187-42F2-BC22-C2F3A7EB72DF.jpeg
    735C6A3D-C187-42F2-BC22-C2F3A7EB72DF.jpeg
    229.9 KB · Views: 74
  • 11520032-7ED4-4377-A94B-14352C0C3C4D.jpeg
    11520032-7ED4-4377-A94B-14352C0C3C4D.jpeg
    346.9 KB · Views: 63
Member
Joined 2011
Paid Member
I don't quite know which one is "the pin behind Pin 1". Please take a look at these images. I've labeled three PCB solder-points (where component leads are soldered to the board) on the Missouri schematic, marked A, B, and C. I've also marked their positions on your photographs.

B is pin 1 of integrated circuit U1.

A is the ground pin of supply bypass capacitor C18.

C is the VTOP pin of resistor R7.

Are any of these, the pin which you are afraid was damaged during rework?

_
 

Attachments

  • missouri0.jpg
    missouri0.jpg
    65.1 KB · Views: 60
  • missouri1.jpg
    missouri1.jpg
    68 KB · Views: 68
  • missouri_schematic.png
    missouri_schematic.png
    11.2 KB · Views: 61
”A” in both photo’s is one leg of C18 capacitor…a 0.1 film capacitor next to the LME49720. The ”B” pin is the #1 pin of the LME49720 TI dual opamp, and the pin directly to the left of it in the first photo,is the pin in question. In the second photo the pin to the right of ”B” is the pin in question. Pin #1 of the LME49720 is the one with the square solder pattern on the Circuit board. In photo one you can see that most of the solder joint is missing on the board. There is about a 1/4 of the joint soldered, but I think the joint on the front side of the board is fine. It is hard to see that joint tho as it is under the DIP-8 socket.
 
Member
Joined 2011
Paid Member
So all this time you actually meant "pin 2 of the IC is what worries me" ??

The schematic shows an easy way to test this and if it's bad, the schematic also shows an easy way to fix it. When I drew the schematic and laid out the PCB, I never imagined that anyone would ever need to perform this kind of surgery. But luckily it's all right there, close together and easy peasy.
 
Yea, Pin #2 of the IC is what worried me. I wasn’t sure how the IC pins were numbered. Sorry about the long yak. I see now that. Pin #1 and #2 are connected and they both go to THAT1200….
I can deal with this.
Thanks for your patience. This is why I ordered extra parts.
I am moving along tho. this was my first serious whoops. I have saved the Output stage for last so as to hone my knowledge and soldering skills. l appreciate the time and efforts that you all have put into this.
image.jpg
 
My Sony TA4650 chassis hot rodding has been operating very well for a couple of weeks. The internal preamp is a more than adequate gain stage, powered by a small SMPS internal 20V. I think it must have at least 16db of gain. From circuit diagram, the preamp looks like a 'flat amp' FET with an attached feedback BJT transistor operating as a 'source follower' feeding another FET/BJT local feedback combo as common source for gain (I am strictly amateur at reading these things). The only potential 'problem' with the internal preamp is an output impedance of 3k (high?). The preamp output and power amp input makes substituting external gain stages straightforward.

I got the bulwark cards, but so far, I don't see their necessity unless the internal preamp goes kaputsky. The heat stability seems adequate. Cutting off the third prong from the power cords lifts the ground and the whole thing is completely quiet as far as I can tell, even without the noise cancelling cards, so I will wait to consider putting those in, too. I have re-biased it and will use it for a while to determine reliability.

Sound of one VFET clapping (2SK60) is remarkable. Specular highlights, attack intensity, imaging and nuance belie the modest power rating. I'm glad it is the single die rather than the double die paralleled 2SJ28 varietal VFET just out of audio superstition. Where it differs from flea powered tube amps is its ability to frame dynamic contrasts.
 
I have the output boards stuffed and am ready to attach the boards to the alum. brackets and mount the VFETs and MOSFETs.
If I attach the 2SJ18’s to the tester they read as a diode one at uf 629mv, c=682, and the other uf 618mv, c=708pf. Does this sound reasonable for these 2SJ18’S? Am I testing these correctly? image.jpg
 
No SMPS DC FILTER P089ZB cards attached yet.
However, I am wondering if they are redundant in this setup because the startup thump is minimal and I'm not really hearing any significant noise.
It's not noise I can hear that I'm keen to eliminate, more a veil that masks the micro details that bring music to life. The 'black background' that some reviews mention can only be achieved with an (almost) perfectly quiet psu.
 
I purchased the VFET’s off of the DIY market place from Robby Wijaya in Indonesia. So it would be nice to know if they are good before I install them and assemble the the amplifier. They look new, but…..
So I attach the case to the tester with a jumper to get the third leg? What kind of values am I looking for?

Thanks