DIY Sony VFET pt 1

Member
Joined 2020
Paid Member
Biggest advantage of external SMPS “brick” is that it has been safety certified and approved by almost every possible organization. CE, UL, CSA, FCC, etc...

The amp itself is inherently safe from electrical-mains shock hazard. It allows a product development cycle to save a lot of $ in CE/UL certification fees.

As a side benefit, it’s cost effective and contributes less radiated noise when outside.

A 36v 4.44A capable SMPS really can do a lot of things. Imagine two in series would be a great +/-36v PSU for a multitude of 50w Class AB amps.

More than a side benefit is noise performance benefit. Here is the FFT of the noise floor of an amp I recently built using a SMPS from Micro-Audio. This is with the inputs to the preamp stage shorted. The preamp used a small Mean Well PCB mounted DCDC converter (another tiny SMPS) to generate +/-15v from the 12v auxiliary output of the 51V SMPS. The SMPS is capable of supplying 600w continuously. I don’t know about you guys, but I have never been able to get the noise floor of any linear PSU to be this clean. Full scale output is referenced to -10dB. So the noise floor (not including the little blips) is very very low - inaudible. Cannot tell the amp us turned on - a similar comment I heard 6L6 say about the VFET - and that’s a great thing to say about any amp. So both preamp and main power amp were powered by SMPS’s. Result is absolutely quiet amp.

FFT of amp output into 10ohm load, inputs to preamp shorted. Ref to max signal at -10dB so noise floor is -130dB:


With my ear pressed (touching the driver), I cannot tell the amp is turned on.

This is a very interesting matter. One of the things I was planing to build was a PC-high end dac, and was wondering if a high end PC power supply could work well with a dac. After all PC psu´s gives us 12v 5v and 3.3v lines, that can be usefull for audio.

Also there is the fact that GAN-fet are begining to be seen on the highest end power supplies, and those are supposed to get very very little ripple.
 
Member
Joined 2002
Paid Member
Normally building a good linear PSU for a power amp is expensive, time consuming and makes the amp very heavy (and large).
Maybe a good designed SMPS with good filtering is perfect for "constant current" amps like KL. A amps.
For class A/B a linear PSU with a large cap. bank is probably needed to deliver fast peak currents.....

Heavy? My latest 50W class A adventure broke 50+kg(110lbs) and my back! :eek:

Regards,
Dan
 
I have complete faith in the SMPS power supply - especially if Nelson does......

I built an amazing low-noise MC phono stage, designed by Wyn Palmer over at Audiokarma - and it was super-quiet. It used either this Meanwell SMPS supply (or one very similar). Wyn was impressed with the specs of this supply enough to use it for such an application
 
Founder of XSA-Labs
Joined 2012
Paid Member
What about regulated? Maybe similar to what you've designed?

A linear PSU that is either regulated or has a capacitance multiplier will have about -50dB less 60Hz noise than a typical CRC. So that’s about —100dB to -110dB. Even shunt regulated supplies cannot get this low. I have not seen a regulated linear supply that can handle tens of amps have -130dB peak at 60Hz.
 
I anticipate a bunch of surprised builders who suddenly discover that a narrow chassis makes it a little more difficult to get both hands inside, and a lot more difficult to get a soldering iron, a piece of solder wire, and both hands inside.

This had slipped my mind... completely, I wonder why... :)
 

Attachments

  • ACA's .jpg
    ACA's .jpg
    607.2 KB · Views: 489
The new VFET amp has features in place to perform at a high level with a single SMPS power supply. In addition to the DC inlet filter, the input cards and the output stages have local 1000 uF power decoupling capacitors to improve channel separation.
That doesn’t mean that I won’t eventually build a dual mono linear PSU for it. I’m just funny that way. It does mean that the external linear supply will need to be a top of the line implementation to have a fair chance of letting the amp sound better. So, big and heavy, and full of AC line voltage. Something very similar to what I did for my F6. ;)
 
It's done,all buttons pushed

I am one of the undeserving but lucky few from the bottom of the barrel, and trust me I could not be more surprised when I saw that my number was picked.

But before everyone more deserving start loosing it, I have donated to the HiFi community for the better part of 30 years, and I think it is safe to say that Peter Walker owns my left kidney and Leif Ernstsen the right one.
So fear not, this amp will live a rich and full life together with other Pass designs, quad amps, carver, dynamic precision etc.

As a teaser for myself I tinkered with some shelf ware and put together a TL speaker that I think will fit this little beauty perfect, aiming for +95dB in real life with AMT in the top, Audax PR17M0 in the middle and most likely Eminence deltalite dealing with the low end.

Anyhow, this was mostly written to spread some hope that even lurkers stand a chance, and at least this one knows what a special opportunity this is and will make the most of it.

Best of all is that fedex promisied that I get my hands on the chassie before end of business Monday, what a great start to a week.

Thanks to Nelson, Jason and everyone else whom made this possible.

Chris:wave::wave:
 
Question from a non-technical type person.

Will playing a signal through the amp before the VFets have reached their operating temperature shorten the life the VFets, or harm them in any way?

My amp will be doing double duty, for music and for TV for my kid.

When doing TV duty, there's no getting a 4 year old to wait for more than a minute before the show has to start.

TIA
 
Founder of XSA-Labs
Joined 2012
Paid Member
So for +/- you hook the + terminal from one to the - terminal of the other?

Seems that you would hook the 2 - terminals together ....

This only works if the negative terminal is isolated from the earth ground. Which I think it is. I can check the Mean Well 24v brick I have. I have done this with the 24v 5A LED lighting SMPS on the M2. It worked like a champ. Very low noise and the hum went away.

Plus to minus forms the common 0v (analog ground).

Just checked. The earth ground is indeed isolated from the negative output (and positive output of course). This is a Mean Well GST60A24-P1J. 24v 2.5A SMPS “brick”.
 
Last edited: