R7 & Heat Sink bolts
Check for R7: #240 page 24, #241 page 25, #253 page 26, & #259
Heat Sink bolt tightness: #257 page 26
Check for R7: #240 page 24, #241 page 25, #253 page 26, & #259
Heat Sink bolt tightness: #257 page 26
more details...
When I decided to build a linear power supply for my VFET amp, I chose to keep it a bit smaller and simpler than my usual amount of DIY overkill. I also wanted to keep its size in a nice proportion to the chassis that was designed especially for this amp. This led me to purchase a pair of Triad Magnetics 160VA toroidal power transformers, VPT30-5330, on which to base the unit. These are just the right size for the Sony VFET amp. After a few hours of being left on, they are warm to the touch, but not hot.
To realize the best potential of the amplifier, I knew the PSU needed to be a full dual-mono design, and have very low noise and ripple. My success with the "SLB" capacitance-multiplier supply in my F6 indicated that the single-rail version would probably be a good way to go. This is a relatively compact PCB that fit well in the smaller chassis that I wanted to use. As it uses the LT4320 synchronous rectifier, it also makes the most of the power which the smaller toroids will deliver. As I did with the supplies in my F6, I made some different choices for the output transistors in the Szklai pair, intended to keep them in a more linear region of operation while allowing for a lower dropout voltage. Eventually I also had to re-tune the RCRC control network at the base of the control transistor to allow easier adjustment of the output voltage at 1.7A current delivery.
The power supply has its own warm up time before it reaches it final voltage, but happily that is less than the warm up time of the amplifier. The two running together need an hour or so before making the final adjustments. That is pretty typical of other FW amps with linear supplies, so no problem.
So far I haven't added any bulk capacitance to the output of the SLBs. With my F6, I found the extra bulk capacitance was necessary to get the very best performance. With the Sony VFET, I find its overall tonal balance to be just right with the amp as-is. If I were to make a change, it would be to lower the value of the signal coupling capacitor in the front end of the amp from 1000 uF to 440 uF, or even 220 uF. Then I would add a Kemet ALS70 series 24,000 uF, 63V electrolytic for each power rail inside the VFET amp chassis. This combination would have the dominant pole of the amp set by the supply capacitance, not the signal coupling cap.
The inside of the VFET amp was only altered to remove the SMPS ripple filter that is common to both channels. I left the Panasonic 1000 uF caps in place on the individual output stage boards. Nothing else in the amp signal chain has been altered as yet.
When I decided to build a linear power supply for my VFET amp, I chose to keep it a bit smaller and simpler than my usual amount of DIY overkill. I also wanted to keep its size in a nice proportion to the chassis that was designed especially for this amp. This led me to purchase a pair of Triad Magnetics 160VA toroidal power transformers, VPT30-5330, on which to base the unit. These are just the right size for the Sony VFET amp. After a few hours of being left on, they are warm to the touch, but not hot.
To realize the best potential of the amplifier, I knew the PSU needed to be a full dual-mono design, and have very low noise and ripple. My success with the "SLB" capacitance-multiplier supply in my F6 indicated that the single-rail version would probably be a good way to go. This is a relatively compact PCB that fit well in the smaller chassis that I wanted to use. As it uses the LT4320 synchronous rectifier, it also makes the most of the power which the smaller toroids will deliver. As I did with the supplies in my F6, I made some different choices for the output transistors in the Szklai pair, intended to keep them in a more linear region of operation while allowing for a lower dropout voltage. Eventually I also had to re-tune the RCRC control network at the base of the control transistor to allow easier adjustment of the output voltage at 1.7A current delivery.
The power supply has its own warm up time before it reaches it final voltage, but happily that is less than the warm up time of the amplifier. The two running together need an hour or so before making the final adjustments. That is pretty typical of other FW amps with linear supplies, so no problem.
So far I haven't added any bulk capacitance to the output of the SLBs. With my F6, I found the extra bulk capacitance was necessary to get the very best performance. With the Sony VFET, I find its overall tonal balance to be just right with the amp as-is. If I were to make a change, it would be to lower the value of the signal coupling capacitor in the front end of the amp from 1000 uF to 440 uF, or even 220 uF. Then I would add a Kemet ALS70 series 24,000 uF, 63V electrolytic for each power rail inside the VFET amp chassis. This combination would have the dominant pole of the amp set by the supply capacitance, not the signal coupling cap.
The inside of the VFET amp was only altered to remove the SMPS ripple filter that is common to both channels. I left the Panasonic 1000 uF caps in place on the individual output stage boards. Nothing else in the amp signal chain has been altered as yet.
Attachments
That is something to drool over, for sure! Great news for a lot of builders! Thanks so much for all this incredible endeavor Papa Nelson!The second run with the N channel VFETs is nearing completion.
Here is a photo of the output stages, partially assembled for testing.
Rafa.
The second run with the N channel VFETs is nearing completion.
Here is a photo of the output stages, partially assembled for testing.
They look great! Fantastic news! Fingers crossed (again)!
You should subscribe to the
“DIY Sony VFET lottery discussion”
thread, when the second round is started, it will show up on that thread..
The second run with the N channel VFETs is nearing completion.
Here is a photo of the output stages, partially assembled for testing.
Unfair looks so much better than the first run!
Due to the demand, there won't be any mass email sent for round 2.
However there will be a second entry window quietly advertised in the VFET lottery discussion thread for people who are very keen and following along with the active discussions.
The window date and time will be advertised with at least 3 days notice, in that thread, once we know everything is ready to ship.
Pass DIY Addict
Joined 2000
Paid Member