DIY Sony VFET pt 1

6L6

Moderator
Joined 2010
Paid Member
If my amp is any basis for performance, and I believe it is, you’ll not be able to make a better PSU than what’s provided.

It is the quietest amplifier I have ever heard, period. There is no sound at the speakers and it’s indistinguishable between the power on or off.

That said, since you will need 50V caps, 15,000uF x4 or thereabouts will probably be a good place to start. Single rail, you can split the store PCB and go dual-mono, or just use one half.
 
SRMcGee,

For a low noise linear power supply, you might consider looking up the SLB power supply which uses the LT4320 active rectifier and a cap-mx circuit. Folks who have built it say that it is much quieter vs a "standard" linear power supply. There is both dual-rail and single rail versions.

Having said that, it will be hard to beat 6L6's experience obtained using the recommended SMPS.
 
Founder of XSA-Labs
Joined 2012
Paid Member
SMPS is the future for Class A amps. They are indeed inherently quieter because they shift mains frequency of 50/60Hz into the ultrasonic range where the switching frequency is typically 400kHz or greater where very steep filters to block it, and also, our ears and speaker transducers simply don’t work in that region. A linear supply on the other hand has to work very very hard to suppress a signal in the audio band by more than -120dB to reach the equivalent noise floor of a good SMPS in the audible hearing band. It’s very hard to make anything filter a peak by -120dB. Typically, the best linear supplies can manage maybe -80dB or -90dB. This still leaves a very faint mains “hum” that can be heard with ears pressed to the speaker cone if the PSSR of the amp is not high. Many SE Class A amps have much lower PSRR compared to push pull Class AB amps. But if one must go linear, every trick in the book can be thrown at the PSU, low noise active rectification, a CRC, a capacitance multiplier and even a maybe a CLC thrown in for good measure. With all of this, 100mV ripple of a standard CRC PSU running in Class A at a couple of amps continuously, can get down to about 1mV ripple.

I recently came across a commercial SMPS (Micro-Audio) that doesn’t mind being connected to a large 80,000uF cap bank (manufacturer provides a bare cap bank PCB for user in fact) and don’t hiccup upon startup. The topology doesn’t use regulation so it varies with mains voltage - but the result is more like a linear PSU but devoid of mains noise peaks.
 
Member
Joined 2004
Paid Member
Jim, zman01 and xrk971:

I hear you: one step at a time. This doesn't have to be rushed, and the Meanwell may be a fantastic solution.

ZM:

I hear you as well, and the idea of building a custom power supply remains compelling for me. I just need to try the amp first before deciding whether to start messing with it.

Many thanks,
Scott
 
Mileage may vary.

It will be interesting with the stock off the shelf option moving forward. Version 1 or version 2 better? Measurements and debates to follow.

Then the 4 front-end options coming soon afterwards (and gain choices from transformer versus transformerless) being made available. Discussion of suitable pre-amps and will it be glass or sand in the path. There will be a plethora of member designed power options and swapping out of stock components to high-end audiophile grade cylinders that will be shoehorned into the case connected with the best silver money can buy.

The diyaudio member will wake up, bleary eyed from reading all of ZM's posts late into the night, stare into the mirror and as, "purist build or hotrodded and modded build; which one are you?"

Whatever the choices and what works for you, all I know is there are going to be some nice set-ups pictured and nice music being heard.
 
I wasn't a big fan of SMPS at all, but was given a lesson when trying to better the B1K PS. The low ripple SMPS, associated with Mark's excellent SMPS filter (for smaller devices!), plus some increased PS caps are giving me such a good result I doubt I can better it soundwise.

Class A is quite good in fact for these regulated SMPS, as it turns out they quite like constant draw / load but seem much less happy re ripple & Co with PS modulation / rail load fluctuations, say Class A/B etc.

My sole concern is the following: what can we do to protect this fabulous amp from SMPS failures?

SMPS in my experience can fail with desastrous consequences, and whereas they are easy to replace and the amp's output per se DC blocking cap should save the LS, I would love some kind of protection to absolutely preserve this VFET amp from any SMPS failure.

I don't know if this safety feature is provided somewhere in the system, be it at SMPS or or elsewhere...

Claude
 
Sadly no clue... whenever they went, it was on computers and small devices as I avoided them like the pest elsewhere on sensitive or expensive devices. When they failed on computers, I bought a new one and hoped for the best when plugging it. Worked fine with portable units, probably the battery / load device protecting them, less good on older computers or small devices where it spelled the absolute end. Can only imagine a big voltage and current surge, but don't know TBH... never bothered as not many $$$ devices at home with them, except perhaps computers.

Would feel better though here with some kind of protection, but perhaps today's wallwart SMPS include that already? Or protection diodes could help, perhaps?
 
frugal-phile™
Joined 2001
Paid Member
Spatial Audio M4 TM

Coudn’t find n M4 impedance, the M3 suggests they pay little attention to flat impedance response.

iu


The M4 with an additional woofer will likely be wilder.

Nominal 12Ω average would not be ideal for an amplifier that it seems is happiest at 8Ω. Vut i suspect that won’t be the big issue.

dave
 
frugal-phile™
Joined 2001
Paid Member
One member asked if he could return his winning ticket for the first batch and then take his chance on the 2nd draw (N-channel)... This is perturbing me since.. but I am easily perturbed anyway:confused:

As i understood the rules, one does not have the choice to choose so if you forgo your P-Channel i do not believe any of your entries will have survived to Round 2.

dave