So, hey guys, Im still in the testing and prototypeing phase but i finally finished testing all power source solutions for my Aleph-3 build. The best configuration that makes this amplifier sound absolutely magical is one transformer, two rectifiers, and two capacitor banks. This type of Class A amplifier is incredibly influenced by the power supply, and you can wire it in the most common three ways:
1.A single transformer, one rectifier, and a big capacitor bank: This setup will give you the worst bass possible, a very flat center image, no layering, and absolutely no body in the sound.
2.Two transformers, two rectifiers, two capacitor banks: This setup is very problematic because if one of the transformers is not on the same phase, the sound quality is absolutely gone, and it won't sound better than a cheap '80s or '90s amp. But if they are in phase, the sound quality is decent but not great. The soundstage is there but not precise, you can't pinpoint details, layering is really weak, and the body is still missing.
3.The overall WINNER: one transformer, two rectifiers, and two capacitor banks: This way, the sound is absolutely magical. Precise, pinpoint accurate imaging, fabulous layering, and incredible body to all sounds. It simply becomes that amp that you can't turn off. This configuration also has some downsides: the rectifiers get hot, and the transformer needs to be big. Oh, and under no circumstances should you use a CLC or CRC filter in this configuration. I find it pointless and, in some songs, it ruins the guitar sound magic (e.g., Buckethead songs).I can go into more details, but if you are a real DIY person at the beginning of building one of these, just do it as I say and be amazed. Of course for the final build I will use super fast diodes and decent capacitors .Any comments are greatly appreciated. Peace!
1.A single transformer, one rectifier, and a big capacitor bank: This setup will give you the worst bass possible, a very flat center image, no layering, and absolutely no body in the sound.
2.Two transformers, two rectifiers, two capacitor banks: This setup is very problematic because if one of the transformers is not on the same phase, the sound quality is absolutely gone, and it won't sound better than a cheap '80s or '90s amp. But if they are in phase, the sound quality is decent but not great. The soundstage is there but not precise, you can't pinpoint details, layering is really weak, and the body is still missing.
3.The overall WINNER: one transformer, two rectifiers, and two capacitor banks: This way, the sound is absolutely magical. Precise, pinpoint accurate imaging, fabulous layering, and incredible body to all sounds. It simply becomes that amp that you can't turn off. This configuration also has some downsides: the rectifiers get hot, and the transformer needs to be big. Oh, and under no circumstances should you use a CLC or CRC filter in this configuration. I find it pointless and, in some songs, it ruins the guitar sound magic (e.g., Buckethead songs).I can go into more details, but if you are a real DIY person at the beginning of building one of these, just do it as I say and be amazed. Of course for the final build I will use super fast diodes and decent capacitors .Any comments are greatly appreciated. Peace!
Attachments
Last edited:
I really liked both your comment/description and your photos.
Well done!
Keep posting your stuff...
And I welcome you to the Forum just here. 😉
Well done!
Keep posting your stuff...
And I welcome you to the Forum just here. 😉
It would be one fine day, if your sonic observations could be explained in engineering parlance, regarding the power supply topologies.
Not at all doubting what you've heard, just "why"? And, how would one go about grabbing that gold ring on the very first try, using those engineering principles apparently found via experimentation? Or is audio engineering purely a matter of try this, nope, try that, nope - try this other way - and keep iterating until the magic button gets pressed?
Not at all doubting what you've heard, just "why"? And, how would one go about grabbing that gold ring on the very first try, using those engineering principles apparently found via experimentation? Or is audio engineering purely a matter of try this, nope, try that, nope - try this other way - and keep iterating until the magic button gets pressed?
IMHO
For what it's worth, a deep belief of mine (just like other people's) is that the power supply design of any piece of audio equipment plays a crucial role in its SQ.
I myself have had similar experiences and It doesn't surprise me at all, quite the contrary, when someone manages to describe simply how effectively.
There is not yet a measurement or measurement methodology that can tell us how a piece of equipment sounds.
And look what happens with loudspeakers, which as transducers make things even more complicated, but not impossible to manage in some way.
I trust what I hear with my ears exactly as much as I trust a person when I speak to them in person, looking into her eyes and hearing the tone of her voice.
I can be wrong, sure, that's obviously not in dispute, but it also doesn't mean that I can't perceive differences in sound when I create differences in the design of a device or speakers.
So yes, it would be nice if that day came soon, if it ever comes.
But I'm not at all sure that all the engineers will like it.
For what it's worth, a deep belief of mine (just like other people's) is that the power supply design of any piece of audio equipment plays a crucial role in its SQ.
I myself have had similar experiences and It doesn't surprise me at all, quite the contrary, when someone manages to describe simply how effectively.
There is not yet a measurement or measurement methodology that can tell us how a piece of equipment sounds.
And look what happens with loudspeakers, which as transducers make things even more complicated, but not impossible to manage in some way.
I trust what I hear with my ears exactly as much as I trust a person when I speak to them in person, looking into her eyes and hearing the tone of her voice.
I can be wrong, sure, that's obviously not in dispute, but it also doesn't mean that I can't perceive differences in sound when I create differences in the design of a device or speakers.
So yes, it would be nice if that day came soon, if it ever comes.
But I'm not at all sure that all the engineers will like it.
So I can go into detail, of course, I can talk for days about audio stuf, scientifically and esoterically .
1.Single Transformer, One Rectifier, Big Capacitor Bank:A single rectifier setup with one capacitor bank can struggle to deliver the necessary current cleanly, leading to increased ripple at the amp board terminals and poorer performance, especially in a high-current Class A amplifier.
2.Two Transformers, Two Rectifiers, Two Capacitor Banks:
The issue with phase alignment is critical. If the transformers are out of phase, they can introduce noise and imbalance, severely degrading sound quality. When in phase, this setup can work well but may still face challenges in delivering perfectly balanced power, affecting sound precision and imaging.
3.One Transformer, Two Rectifiers, Two Capacitor Banks:
This setup combines the benefits of having independent rectification and filtering stages while maintaining a single, well-regulated power source. The downside is increased heat from the rectifiers and the need for a larger transformer.
A single transformer might experience more significant voltage drops and ripple under heavy load, causing the rectifiers to work harder to maintain steady DC output. This increased effort can generate more heat in the rectifiers.
For the esoteric part :
After years of working with HIFI I found that numbers don't always add up . And trial and error are the only way to go to be absolutely sure . For example I know the phizicks behind the way that electricity travels through a wire at different frequency's but after trial and error I can tell you that it doesn't matter at all because after testing I can tell you for sure that a breaded cable sounds better and that a tick breaded power cable makes an amp sound better. And I can go on with examples .
Now pls don't get me wrong I'm not trying to be a smart ***, know it all . I'm just saying that, when a small thing like letting you speaker cables on the ground can influence the sound, one can't possibly think of everything and if one has the possibility to test all ways or most ways to get to the same outcome, one must test and research . This is why we are DIY people and **** up the whole living room with tools and parts just to test a theory. We do it because it's our fun that gives us the fulfilment of success.
Attachments
Currently I use Warfdale 505.2. Found them at a sale old new stock mint condition . I had to replace te ferofluid in the tweeter the rest is stock . Never really liked Warfdale"s but these are great . Yest there is a small hum but I think it's because of my terrible wiring.
Attachments
Last edited:
I totally agree with everything you just said . I see that the quest for great sound also pushed you beyond engineering 😉 .
With only a rectifier and capacitor, there will always be some residual hum.
The usual series R or L and another C can knock it down quite a bit (20dB or so), at the expense of load regulation.
Although perhaps you can't see them perfectly because the wires are thick and short, I did practically the same with the wiring of one of my Class D power amps. 😉
I searched for a long time for a way to do that 4-strand braid, and I finally found it on YT
www.youtube.com/watch?v=vdvugjyP6ZI
It is the third of the three ways, with 4 wires (they seem like only two wires, but instead there are four ones).
As I mentioned a CRC or CLC makes the sound more lean decreasing the atac in some parts of the sound, not by much, but it dose, you can especially hear it in guitar strums, the guitar strums dunt start and finish they don't have that transient that starts high or low and changes as the string settles.
I see a class A amp as a very tensed up rope so tensed that you can walk on but, if there is a soft place along the rope its a spot that will give away .
I have tried more variants of the power supply than described.
To eliminate or reduce the hum more capacitors will be added to the final build . I currently have 17000uf / rail, + 44000/rail will be added.
For now my biggest problem is finding a transformer that will be sufficient im thinking 500/600 VA 21-0-21 .
I see a class A amp as a very tensed up rope so tensed that you can walk on but, if there is a soft place along the rope its a spot that will give away .
I have tried more variants of the power supply than described.
To eliminate or reduce the hum more capacitors will be added to the final build . I currently have 17000uf / rail, + 44000/rail will be added.
For now my biggest problem is finding a transformer that will be sufficient im thinking 500/600 VA 21-0-21 .
Attachments
Have been building Randall Thatcher's Aleph 30 Clone kits for 3 years now and agree with your assessment "you can't turn [this amp] off" completely. He prescribes the same PSU design you describe and attached please see a set I just completed using Gianlucca's fine monoblock chassis.
Have developed a system where I can switch between 300B, Class A, SEPP (=Aleph), and A/B amps quickly in order to match the amp type with the media source: tube amp for vinyl, A/B for digital, and so forth. I call this "active amp listening."
The Aleph is so much better than any of these other amp types I keep it on as the goto most of the time.
Notice that your schematic calls for a wide voltage range of +/- 35 volts. Am not a technical DIYer so wonder if you can tell me what advantage there is to this by comparison to the +/- 22 volts that I have been using? Does a wider voltage span produce more dynamic range?
Also, wondering about regulation. None of the Aleph kits I have built are regulated (as preamp PSUs normally are) and wonder if a regulation to a specific voltage would create any sonic improvement. Am now using separate PSUs for each L/R circuit board at the PhonoPre and Linestage levels -- both levels regulated to +/- 18vdc. With this VDC operating "upstream" to the Aleph would it be of any sonic advantage to having the Aleph PSU also sending out +/- 18 volts?
Looking forward to hearing of your final build results and, especially, your audio analysis.
Congrats!
Attachments
Yes it was.
However, I wasted some wire in lenght since I made the braids much longer to be able to handle them, after fixing the 4 ends on a chair with cable ties.
Then I shortened them and crimped the ferrules on the ends.
Dual rectifiers - makes sense. I like the ground break with opposing diodes, resistor, cap. There are many ways to do this. CL-60 is another way.
From an old Nelson post:
"Most customers want total silence from the amplifier, including
mechanical noise. If there is not complete matching between the
secondary coils and only 1 rectifier bridge, any net DC imbalance
between the current of the + supply and the - will tend to
saturate the core of the transformer and create noise. This is
seen for quite low current differences and can also show up with
low frequency output. Using two bridges eliminates the problem.
In the eyes of a manufacturer, any other subtle differences
pale in comparison to the cost of having to replace a transformer
in the field due to mechanical noise."
Have you tried 2 transformers, one per channel with 2 rectifiers per transformer? In other words - dual mono / mono blocks with 2 rectifiers per channel.
What Craig is showing is my dual mono V8 boards. CCRCRC filtering. 2 rectifiers per transformer, plus snubber circuits. The reason I went for the overkill filtering is because I run JBL horns, and I don't want to hear any hum. There are many happy builders of Aleph 30's, Aleph 60's, and a few Aleph 2's with those boards in them.
What rail voltages are you running? For Aleph 3 / 30 (and most First Watt Dual Rail amps) you can easily go for:
Stereo (one donut) - 300VA 18+18V Secondaries
Dual Mono / Monoblock (2 donuts) - 200VA 18+18V Secondaries per channel
You could also go to 20+20V secondaries. You'll get a little more power output (not much) and higher rail voltage.
Using 20+20V donut means you need 35V caps, not 25V caps.
From an old Nelson post:
"Most customers want total silence from the amplifier, including
mechanical noise. If there is not complete matching between the
secondary coils and only 1 rectifier bridge, any net DC imbalance
between the current of the + supply and the - will tend to
saturate the core of the transformer and create noise. This is
seen for quite low current differences and can also show up with
low frequency output. Using two bridges eliminates the problem.
In the eyes of a manufacturer, any other subtle differences
pale in comparison to the cost of having to replace a transformer
in the field due to mechanical noise."
Have you tried 2 transformers, one per channel with 2 rectifiers per transformer? In other words - dual mono / mono blocks with 2 rectifiers per channel.
What Craig is showing is my dual mono V8 boards. CCRCRC filtering. 2 rectifiers per transformer, plus snubber circuits. The reason I went for the overkill filtering is because I run JBL horns, and I don't want to hear any hum. There are many happy builders of Aleph 30's, Aleph 60's, and a few Aleph 2's with those boards in them.
What rail voltages are you running? For Aleph 3 / 30 (and most First Watt Dual Rail amps) you can easily go for:
Stereo (one donut) - 300VA 18+18V Secondaries
Dual Mono / Monoblock (2 donuts) - 200VA 18+18V Secondaries per channel
You could also go to 20+20V secondaries. You'll get a little more power output (not much) and higher rail voltage.
Using 20+20V donut means you need 35V caps, not 25V caps.
Also - if you're running nominal 35V rails you're in Aleph 5 / 60 territory.
2 output + 2 current source MOSFETs will be absolutely cooking with 35V rails
Are you running IRFP240,IRFP150, or something else?
2 output + 2 current source MOSFETs will be absolutely cooking with 35V rails
Are you running IRFP240,IRFP150, or something else?
Fabulous job on your builds! I was also considering using a large aluminum heatsink case, but I want to create my own take on the amp. My plan is to separate the PSU from the amp section and design it in a way that allows air to flow from the bottom to the top using perforated plates.
As for your inquiries: My board uses a maximum of ±25V, with a current draw of about 3 amps per board. I tested voltages between ±18V up to ±23.5V and did not notice any difference in sound quality at normal to mid listening levels. I never intended to try a regulated PSU, as all my experiments with regulated PSUs ended up being converted linear PSUs. While regulated PSUs sound very good in amplifiers, they can sound almost too good, and during long listening sessions, I tend to get bored and want to turn it off. So, I haven't used a regulated PSU for a long time now.
The last part about sending ±18V upstream from the same PSU is an interesting idea that I will test later in the build because it might have a huge sonic impact. Great idea, Craig!
Attachments
"CCRCRC filtering. 2 rectifiers per transformer, plus snubber circuits" is actualy great , and in Audio nothing is overkill 🙂 .
My curent transformer is a 300VA 18-0-18 but its to small and it heats up . Also you have to take into consideration the voltage drop for example my transformer gives me 20 volts after rectifi and filtering with 2 boards . So im thinking of a 560VA, prim: 240V / sec: 21V-0-21V i do think that at 4 ohm impedence it will help alot .
I'm running IRFP140s but can't comment on them or any of the other models you listed. All I can say is that the clone I'm using is highly regarded as one of the best because of some modifications they did. It sounds great but heats up more. It's the Aleph-3 by Mofi Audio China.
You might think I'm a fraud, but I don't know all that much about FETs. I can talk for days about BJTs, but I have little knowledge of FETs. For example, in the differential pair, I want to change the Zener diode to a couple of series LEDs, but I don't know how it will react because I never played around with FETs. So, asking me about this topology is mostly pointless. However, it does sound fabulous.
Attachments
I'm just interested in source of your Weed, Man .........
