Sony vFET Illustrated build guide

Hi Guys,

Thanks for all of your advice. For this build I did measure every component before installation, and all of the board measurements were in alignment with Nelsons specs. When I reflowed the board I used new flux and solder, then cleaned the board with a tooth brush and IPA followed with some windex, I used a strong magnifying glass to make sure that there were no solder bridges. The pics I posted don't really show the bottom very well, I'll try posting some better close ups.

When I initially powered up the board I remember that it was more finickey then the other one and was harder to set up, and after reflowing the board the values had changed on the voltages. I'm hoping that this is an indication of having fixed a cold/bad solder joint. My problem now is that I have to reset all of the pots correctly with all of the components installed.

My path forward will be to follow wined's advice on the adjustments and see where I land. I'll post my results when I'm done.

Paul
 
Cascode feedback in the front end circuit?

In the VFET Amplifier Part 2 thread (here and here), Rodeodave was kind enough to show a P2P sub circuit that replaces the resistors that bias the cascode front end transistors (Q3 and Q4) in the DIYAudio VFET kit with the TL431 shunt regulator circuit. However, there is an absence of the cascode feedback circuit (seen in the VFET Amplifier Part 2 pdf as parts R29 and R30). Is there any reason or reasons for not implementing cascode feedback in the DIYAudio VFET kit design? And if there isn't, can one still use a 10K resistor between the drain of the P MOSFET (Q5) and the drain of the N JFET (Q1); and similarly a 10K resistor between the drains of Q6 and Q2?

Thank you!

Anand.
 
I finally got around to taking some more pics and measuring the TL431's ref legs to gnd.

bad channel Q11 = 2.48 v, good channel Q11 = 2.48 v
bad channel Q12 = -10.32 v, good channel Q12 = -8.40 v
bad channel Q13 = 2.47 v, good channel Q13 = 2.48 v
bad channel Q14 = -21.32 v, good channel Q14 = -21.22 v

I took these measurements after I had biased up the front end, it seemed to bias up more easily and stably this time after reflowing, but that might be because I have done it several time now. The readings between the two boards are pretty close with the exception of Q12, and I'm not sure if it is significant.
the pictures were the best that I could do with the camera that I have the specks that look like dust are artifacts of poor lighting and high compression. I looked carefully with a magnifying glass and went between all close legs with a fine blade and didn't find any bridges. I'll wait until I hear back on the opinion of the experts here about Q12, before I attempt to bias up the back end and give it a listen.

I appreciate all your help, thanks,

Paul
 

Attachments

  • image0001.jpg
    image0001.jpg
    215.1 KB · Views: 635
  • image0002.jpg
    image0002.jpg
    239.3 KB · Views: 634
  • image0003.jpg
    image0003.jpg
    203 KB · Views: 629
  • image0004.jpg
    image0004.jpg
    194.3 KB · Views: 626
  • image0005.jpg
    image0005.jpg
    187 KB · Views: 618
  • image0006.jpg
    image0006.jpg
    168.7 KB · Views: 206
Hello All,
my bad channel saga continues. I biased up the back end to 100 mV across R32 by shorting T18 to gnd and adjusting P1 & P2, so I got 100 mV with ~ 5mV DC offset on the output. I Gave things a listen tonight and I'm still getting the same problem. At very low volumes the right channel is fine but as soon as I raise the volume a bit the channel fades out to silence, if I lower the volume again the channel reappears.

So now I've got to assume that I have either a bad board, or a bad component somewhere on the board. A bad board is a long shot, so I'm betting on a bad component. Since the VFETs are not replaceable and were measured by Nelson I'm going to assume that they're fine. I already replaced the JFETs and still have the problem so I'm assuming that they are fine. I carefully measured every resistor before and after installation and compared the in values to the good board and they all agree so I'm assuming that they are good, and I measured and compared the TL431 regulators with the good board with which they agreed, so that leaves the mosfets and transistors as potential culprits. Does anyone have an idea which ones could cause such symptoms. Otherwise I'm out of ideas and I'm thinking just to replace them all.

Thanks,

Paul
 
With one good board you should be able to find the problem in the bad board. Raise the volume to where is fades out and check the voltages at all points between the good board and bad board. There is going to be a big difference somewhere. Just checking resistance without being powered up probably is not going to tell you where the bad part or circuit is since it only starts showing up when you increase the volume. It sounds like the input signal is being shorted to ground when the volume is turned up, possible.
 
Hi Paul,

I might have missed this, but did you try swapping the left and right channel
inputs?

Cheers,
Dennis

Hi Dennis, yes, I tried switching inputs as well as preamps. The preamp that I'm using now I use with my aleph j with no issues, in fact I just substituted the vfet into the aleph j system. I'm going to measure voltages with music playing with the volume up enough to cause the right channel to fade out as advised by wdecho. Then compare the readings between boards and see what stands out. Hopefully it will point to the problem. If nothing else this build is an exercise in patience.

Paul
 
Hi Guys,

I took some voltages on both boards while the amp was on and playing music, I turned up the volume just until the right channel faded out to take the readings. I got interesting results that seem to point to a problem in the front end. I've attached a table with all of the readings. When I turned the volume up until the right channel faded out the readings across R5, R6, T16 - gnd, and R32 ( 0.814 v, 0.642 v, 25.6 v, 57 mV)went wonky.

So then I decreased the volume until the right channel began to play and the readings all started to come back into line (1.51 v across R5, 1.427 across R6, 14.88 v T16 - gnd).

So I'm guessing that Q5/Q6 might be the problem. So I'm going to order new Q3,4,5,6, and replace them and see what happens. If there are any other components that I should replace please let me know.

Paul
 

Attachments

  • VFET voltage test 6-20-17.JPG
    VFET voltage test 6-20-17.JPG
    108.1 KB · Views: 431
Check those resistors around those transistors especially R5 and R6 and be sure they are 47.5 and not 47.5K, common mistake I am guilty of making in a number of builds when you have resistors like those. I would also check R7 R8 R9 and R10 as well for correct value as well as R34 and R35. Just pull one lead and check ohms. Doubtful a transistor is bad but can and does happen. Bjt's are cheap and I am going to assume you used FQP for Q5 and Q6 and they also are cheap. Other resistors in this circuit one has to be careful about are the 2.21K and 221 ohm ones. Easy to install wrong values in wrong hole. Pull one lead and check ohms.
 
Last edited:
Member
Joined 2004
Paid Member
You still need to adjust the right channel front end. T18 to ground should be less than 300mv warmed up. Keeping the output D.C. to less than 20mv on the speaker.

I don't get to concerned with measuring t8-t9. Put a meter on t18 to Ground and another on t6-t7. If you get this correct then t8-t9 will be correct by default.

Remember small changes make a big difference.

Keep at it, you will get there.
 
Last edited:
My money was on your preamp, but this seems not to be the issue. Another thing to try is to make sure you've got the trim pots properly adjusted. I remember bringing my Aleph-X amps up for the first time and thinking that something was drastically wrong until I tweaked the pots some more... Sometimes the adjustment goes smoothly and then all of a sudden it's just way out in left field. I was sure I had broken it, but an adjustment proved otherwise.