Pass DIY Addict
Joined 2000
Paid Member
re: its silence. I've used SMPSs (clearly) in a bunch of "smaller" applications previously for audio, but this is my first power amp. All potential misgivings have been eliminated. It is dead silent in my system, and the amp is just amazing IMO.
I have to say, this amp project has made a convert out of me as well
I've used a number of SMPS over the years and have always been disappointed in one way or another. I've even warned others here against using them because of my own experiences. Seems that we've turned a corner with this technology, at least with the well-constructed ones (I'll still not recommend the $10 ones from ebay). I am surprised just how heavy that little Meanwell unit is!
I have been out for a week on travel and my VFET packages were sitting there at home, waiting patiently. I hope to be able to build it soon. It has been great to live vicariously through all the folks here who have been building their VFET amps. Super job everyone!
A well designed SMPS has a much better chance of reducing or eliminating noise and mains hum. It pushes the modulation noise out of the audible band (50/60Hz) up way past the ultrasonic range (100’s of kHz) where our ears don’t work and where the speakers cannot respond. Linear supplies especially affect an amp with an unshielded signal transformer as there is potential for the Edcor to pick up radiated magnetic fields from a large linear transformer if it were inside the chassis. This was my experience with the M2/M2X. Once I shielded the Edcor with mu metal, and moved the linear transformer outside of the chassis or switched to a SMPS, the hum was significantly reduced, but never silent.
My only experiences in a silent SE Class A amp using a linear PSU always required use of a capacitance multiplier after the linear supply and making the supplies dual monoblock. The sharing of a common linear supply was another typical source of low level noise or hum from ground loops in the common connection at the source left and right channels from the source RCA negative/shield.
More companies should make dual monoblock DAC sources.
A well designed SMPS has a much better chance of reducing or eliminating noise and mains hum. It pushes the modulation noise out of the audible band (50/60Hz) up way past the ultrasonic range (100’s of kHz) where our ears don’t work and where the speakers cannot respond. Linear supplies especially affect an amp with an unshielded signal transformer as there is potential for the Edcor to pick up radiated magnetic fields from a large linear transformer if it were inside the chassis. This was my experience with the M2/M2X. Once I shielded the Edcor with mu metal, and moved the linear transformer outside of the chassis or switched to a SMPS, the hum was significantly reduced, but never silent.
My only experiences in a silent SE Class A amp using a linear PSU always required use of a capacitance multiplier after the linear supply and making the supplies dual monoblock. The sharing of a common linear supply was another typical source of low level noise or hum from ground loops in the common connection at the source left and right channels from the source RCA negative/shield.
More companies should make dual monoblock DAC sources.
I have to say, this amp project has made a convert out of me as well
Eric, you'll be pleased to see that the diyAudio Store is selling a PCB + kit_of_all_parts for an SMPS filter, aimed at line level applications: preamps, dacs, linestages, headphone amplifiers, etc. Max DC voltage is 48V and max current is 3 amperes. Just right for "wall wart" powered devices -- put the filter in series between the wart and the audio gear.
Members have been experimenting with them on the Korg NuTube preamp, the ACP+ preamp/headphone amp, numerous DACs, and even Raspberry Pi boxes. So far the results have been very positive.
And yes, before you ask, people have performed the experiment of putting "N" filters in series (N>1) for additional filter-osity. Gilding the lily costs ten buxx, why not? They report that using one filter is tremendously beneficial, and using two filters in seris is slightly better but only slightly. Perceptible, beneficial, but diminishing returns.
diyAudio Store link
_
I’ve had resounding success with SMPS in two of my recent class AB builds:
Rebuild of Hafler DH220 with Bob Cordell redesign
Hafler DH-200/220 Mods
vzaichenko TubSuMo monoblocks
Ultra-high performance, yet rather simple - hybrid and more!
They are very quiet, but not as quiet as the VFET. Hoping for a higher-power SMPS filter concept from Mr. Johnson...
Rebuild of Hafler DH220 with Bob Cordell redesign
Hafler DH-200/220 Mods
vzaichenko TubSuMo monoblocks
Ultra-high performance, yet rather simple - hybrid and more!
They are very quiet, but not as quiet as the VFET. Hoping for a higher-power SMPS filter concept from Mr. Johnson...
Last edited:
I have been out for a week on travel and my VFET packages were sitting there at home, waiting patiently. I hope to be able to build it soon. It has been great to live vicariously through all the folks here who have been building their VFET amps. Super job everyone!
A well designed SMPS has a much better chance of reducing or eliminating noise and mains hum. It pushes the modulation noise out of the audible band (50/60Hz) up way past the ultrasonic range (100’s of kHz) where our ears don’t work and where the speakers cannot respond. Linear supplies especially affect an amp with an unshielded signal transformer as there is potential for the Edcor to pick up radiated magnetic fields from a large linear transformer if it were inside the chassis. This was my experience with the M2/M2X. Once I shielded the Edcor with mu metal, and moved the linear transformer outside of the chassis or switched to a SMPS, the hum was significantly reduced, but never silent.
My only experiences in a silent SE Class A amp using a linear PSU always required use of a capacitance multiplier after the linear supply and making the supplies dual monoblock. The sharing of a common linear supply was another typical source of low level noise or hum from ground loops in the common connection at the source left and right channels from the source RCA negative/shield.
More companies should make dual monoblock DAC sources.
That's great news ! And with the Mark Johnson additionnal filter, the Audio Nervosa is definetly cure concerning SMPS.
No, but only because I haven't collected the parts to make a 2nd version yet... when the Mosfet or Bipolar output stage becomes available there will be lots and lots of mix n' match to try.
My experiments have been with MoFo and F5. Everything so far points to absolutely wonderful results, I'm looking forward to how it all comes together.
My experiments have been with MoFo and F5. Everything so far points to absolutely wonderful results, I'm looking forward to how it all comes together.
And no matter how well corporations design and build better containers to keep all that sand and smoke inside, I come along to challenge them as the better idiot.
Hmmm, can you maybe provide a link or similar to Nelsons explanation?
If an isolated supply was used, what's the point of isolating the chassis from audio GND using a thermistor?
Also, if a supply without isolation is used, the audio GND is connected to mains earth, which may cause ground loop troubles. It would be better to go with an isolated unit.