Sony vFET Illustrated build guide

[Dennis Hui]

What is your new transformer's secondary voltage?

 

[Jim50hertz]

2 x 22vac

 

[Dennis Hui]

Should be fine.

 

[Jim50hertz]

Power supply is rewired with the new power transformer.

I'm reading 62v between +ve and -ve each side, which would give me +ve31..0..-ve31v correct?

This is unloaded, should I be concerned that this is a touch high?

Also, should I continue to test each channel separately?

 

[Ben Mah]

It's in the ballpark. If you measure V+ and V- relative to ground, you should get +31v and -31V.

Definitely test each channel separately.

Also highly recommended to use a dim bulb tester for the initial power-up of each phase of testing.

 

[Jim50hertz]

Thanks Ben,

Useful advice re the bulb limiter, should I do this even though I’ve powered up already? Albeit with half the required voltage?

 

[Ben Mah]

It's probably ok but it's best to be safe. I would take the few extra minutes to do it rather than risk blowing some jfets.

 

[Jim50hertz]

Good call. I’ll do that then, thanks again.

 

[Jim50hertz]

Initial tests of both channels. Lamp limiter inline.

Rail voltages look better

Other voltages are out but I assume this is because the mosfets are switched on?

Should I be overly concerned that the rail voltages are around +/- 2.5v above expected?

Also, G-T18 is quite out of spec

Good to go ahead with biasing the front end do you think?

Sorry my pics are huge.

 

[Jim50hertz]

A more general question on the VFets from the DIY Audio kit. Are these matched?

Both my 2SK82s are marked 91, and one of my 2SJ82s is marked 100 but the number has rubbed off the other one.

Is it safe to assume the vgs of the unmarked 2SJ82 is 100 also?

 

[Dennis Hui]

I believe the vfets were matched in pairs in the kit. But you can always be safe and just
crank up the initial bias voltage for the 2sj82 to get the current level lower.

 

[Zen Mod]

Papa selected them, personally

 

[Dennis Hui]

OK. For G-T16, try adjusting T1 to see if you can get it above say 10V.

G-T18 is high because your mosfet bias currents are different. I think you can start the
bias procedure. I'm guessing your P3 and P4 pots are at midpoint and with
the R34/R35 change the mosfets are now really pulling some current.

Please start by adjusting P3 and P4 to try equalize the T6-T7(ie, across R5)
and T8-T9 (across R6) voltages and monitor G-T18. This offset should drop
as you start getting those two voltages to match closely. The target is
1.5V across both when they warmed up with G-T18 low (say below 200mV)

Please do this with one channel powered at a time.

Also, no worry about the rails at about +/-30V. That will drop a bit with
the output current draw.

Have fun.

 

[Jim50hertz]

Ok thanks Dennis and ZenMod, that's reassuring!

I started out adjusting P3 to get offset close to 0v and went from there.

I stabilised my voltages around 1.2v with offset very close to 0v and the voltages from t16 and 17 to G bang on. I need to walk away for now so have powered down but will come back.

So I need 1.5v across r5 and r6?

Seems very encouraging so far.

 

[Jim50hertz]

Also, in Papa’s notes he suggests a 1a line fuse for first power up with VFets installed.

Would this be a different value in Europe, using a lamp limiter (i don’t have a variac)?

 

[Zen Mod]

in Eu, if mains fuse is what's in case, half that amperage

 

[Dennis Hui]

The schematics shows 1.5V across R5 and R6 so I would stick to that as the
target with things warmed up. Don't worry about getting exactly that value
though you shouldn't have trouble getting it fairly close.

 

[bajjisw]

Also, when you're ready, double check voltages without the lamp limiter. In my case, there was a difference.

 

[Jim50hertz]

Thanks I did! Initial voltages with the limiter inline, then calibrated without, double checking voltages. Again, good call.

 

[Jim50hertz]

and now i have twins any my brother has his own vfet baby.

second time biasing was easier, best tip was to first bring t6/t7 and t8/t9 up to 1.4v with p3/p4, offest t18 around +-100mv and warm it up, wait 10min.

readjust p3/p4 for offset t18 around +-100mv then bring r32 up to 80mv with p1/p2, output offset +-50mv. wait 15 min.

now adjusting all 4 pots p1/p2 - p3/4 in cycle up to 1.45v at t6/t7 and t8/t9 and r32 to 90mv (check first t18), againg offset around t18 +-75mv and output offset +-50mv. baby steps needed or it run away fast (even more with one turn pots). let it heat up then for at last 30 min.

after this time r32 was a little bit under 100mv. go for 1.5v at t6/t7 and t8/9 and r32 to 100mv, check offest's. readjust again in cycle all 4 pots for the target bias and offsets. now with baby baby steps . wait after every adjustment some seconds, you get the feeling when the circuit get into a stable state. dont touch the small heatsinks, the offest over t18 acts at once. i done the final fine adjustment via p1/p2.

at the end i had a bias of 101 mv, offest t18 floating around +-50mv at t18 (1.51v) and output offsets +- 20mv.

remember everything up to +-200 mv over t18 is ok and +-50mv output offset is ok, +-50mv over t18 and +-20mv output offset is perfect for me. the amp needs 1 hour to get stable. floating around a bit is normal.

this way was a lot easier then put t6/t7 and t8/t9 direct to 1.5v, hope that helps.

don't rush .

I found this in the other thread, does it seem like a good idea to work with 1.4v initially, then install VFets then calibrate further from there?

Also, (being paranoid about killing the VFets here) when I install them and power up with a lamp limiter in-line, should I kill the power after 5 seconds or so? And should I install a lower, fast blow fuse with this, as the lamp limiter is limiting current? Should I initially be measuring for less than 100mv across r32?

Any tips here gratefully received!