Power supply that plays music

The Alfa and Omega

The beginning and the End ....
The Alpha and the Omega ....
.... of every Hiend Audio device, ( be it an amplifier, a CD player, a DAC, or whatever else you use out there ) ...

is the *** POWER SUPPLY ***

OK, so this might be a shock to some folks out there, who shall probably consider such musings as heresy, but NO.

It is NOT your "amplifier" that is playing the music.

It is the Power Supply that does.

What is actually an "amplifier" ?

Basically, in cynical terms, that thing, what we commonly refer to as an "amplifier" .... is nothing more than a modulated power supply,
providing power to your speakers.

DIY ....
So this is that what this section is called ?
Is that what I am "supposed" to be writing about ?
If indeed - where is a good starting point ?
I suppose that the best place to start ... is to start at the beginning.

By the end of this article, you will be executing your first DIY project.
You will be investing a total amount of ONE EUR and purchasing.
Purchasing one or five capacitors, which you will be soldering into your equipment.
Even in your most expensive piece of hardware, since, ... these capacitors may simply not be there ....

At the very beginning - we have a power supply.
And guess what - at the very beginning of the very beginning - there is a ..... YES !! correct. ..... There is a rectifier.

Obviously, here I could start rambling about all the possible topologies, the full wave, the half wave,
a "this" topology or a "that" topology. Let's leave the rambling to the teachers. But we want some action out here !

The plain simple truth is, that the most common application of a rectifier,
statistically in about 99% of the cases of any and all audio devices ....
... is a simple, plain vanilla, four legged silicon bridge rectifier.

Is this not shocking ? I mean that at the very beginning of the power supply,
the supply that is the Alpha and the Omega of any means of modulating the flow of power into your speakers,
there is something so trivial and boring as a Graetz rectifier bridge ?

I can already see those waves of protest rising already, readers screaming:
"... HEY! Mark DOWN. There is absolutely NOTHING interesting, in terms of DIY, within a bridge rectifier !!!! ....."

Or is there ?

This is a simplified view of a common circuit that is present in almost 99,99% of all audio devices out there. Be it cheap low market stuff, or even the most expensive devices - they all tend to use some form (identical or "similar") of this circuit. Boring, you say.
"It is a typical circuit".
But guess what.
A typical circuit means typical problems and typical SOUND.

Have a closer look between the "thin" lines again.

And now confront those thin lines with a simple state of fact:
Did you know, that these diodes DO NOT CONDUCT AT ALL ? Statistically speaking ?

YES. This IS true.

Statistically, they do conduct no longer than 1% or maybe 3% of the available cycle time.
FOR THE REST rest of the time, THEY DO NOTHING.


For 97% of the time, your amplifier is running from the energy stored in the filter capacitors.
And the more "beefy" setup you have, with more and more capacitors in the filter, with probably zillions of micro-farads ...
.... the longer is the time that the diodes DO NOT CONDUCT at all and you run solely from the caps.

This is due to the simple fact, that the more capacitors there are in that filter,
the lower the voltage drop on the caps between cycles,
so the percentage of time where the voltage from the transformer and rectifier is "higher" that that on the caps is smaller.

Now, consider this simple mathematical exercise.
Imagine, that your beefy amplifier is pulling a constant current of say 1 Ampere for all of the time from the power supply.

When I say "all of the time" ... I obviously mean 100% of the time.
Now, imagine that you are the diode(s). Since the output of the power supply needs to be pumping 1 Ampere during 100% of the time .... How much current do you need to PUSH INTO those capacitors, if your time-frame is no longer than 3%, or even 1% of the total time available ?
Guess what, as uncomprehendable as it may seem, your calculations are correct.

In the former case, your diodes are pumping ( 100 / 3 ) * 1 = 33,3 Amperes.
In the latter case, your diodes are pumping ( 100 / 1 ) * 1 = 100 Amperes.

Ha! Now I got you interested, have I not ?

As unbelievable as it may seem, the math is unforgiving. You may not like this revelation, but you will now have to live with it.
this is a visualization of the 100 Amperes route (or 33,3 Amperes, if you have a smaller bank of capacitors ....)

Since we have to live with this unbelievable fact, .... maybe this is the right moment to take your screwdriver, open up that cover of your expensive amp, and check the diameter of all those cables in red, that I signaled. The red path is the high current path, encompassing the whole circle route: the secondary winding of the transformer, the first rectifier diode in the bridge, the FIRST (and most important) capacitor within the filter, and then the second (return) diode of the bridge, and then back to the secondary winding, to complete the circuit.

Apart from changing those feeble cables that you find in there (in your expensive amplifier, I mean) .... there is not much you can probably do in terms of "shortening" or "minimizing" the resistance of the route between the secondary winding and the rectifier (unless you are ready to move the rectifier closer to the transformer, which I would recommend).

But apart from changing the diameter of those cables into a totally absurdly thick one (fully justified !!!) ... there is one other thing that you can take under your kind consideration.

Consider SHORTENING the route between the rectifier bridge and the FIRST capacitor of the Filter Bank.
How about shortening it to ZERO CENTIMETERS ? Sounds nice ?

Impossible ? Well, that depends. How about taking an extra electrolytic capacitor, of non-trivial capacity and of appropriate voltage rating, ...
... and soldering it on DIRECTLY on top of the "+" and "-" pins of the Bridge Rectifier ? Kind of like making them an inseparable entity ?
The cap should be essentially SITTING on the PINS of the RECTIFIER.

Every little bit helps. And in ways, that are not fully comprehendable. YES. There is another side of this coin.

It is called NOISE, and electromagnetic pollution. when you shorten the red route from rectifier bridge to the first capacitor, you essentially shortening the loop of the antennae that is emitting into the insides of your amplifier big amounts of electromagnetic radio frequency noise. You might be skeptical and ask: But where does this "noise" supposedly come from ?
The mains frequency is only but 50 (60) Hz.

Correct. The base frequency is as quoted. But .... If you jump into the role of that diode again ... If you are constantly switching between conduction, and non-conduction, ... there is no such thing as an ideal diode. Upon such switchover, there is a small amount of time while the diode actually conducts in the reverse direction. This is the time it takes to recharge and re-polarize the p-n junction within the diode ... And then, ... suddenly ... BANG ! the current being conducted (in reverse) suddenly hits a wall of a closed p-n junction. It is like racing your Ferrari, full of electrons, and banging head-on into a thick concrete wall. The shock is instantaneous. The arising Impulse is nearly vertical on the graph of the time domain, as observed on the oscilloscope. The amount of the resulting high frequency harmonics - is ... meaningful. And harmful. So the less emission you achieve from a "shorter" loop of your 100 Amperes antennae, the better for you. That is also the reason, why you would wish to consider soldering that electrolytic DIRECTLY on top of the output pins of the rectifier bridge.

So, we now learn that the diode is not an "ideal" one way conducting device.
Can we do anything about it ?
We can invest an amount of about ONE EUR, or ONE USD, depending on location, and purchase our remedy: four polipropylene (MKP) or polistyrene (MKT) capacitors, of a value 47~100nF, and solder them on directly in paralel to each of the diodes within the rectifier bridge. Like this:

ATTENTION: the voltage rating of these capacitors need to be at least THREE times larger than the output voltage of the rectifier filter capacitor.
This is important. Do not use voltage underrated capacitors in this location !

So, just to summarize: in this very first DIY project of yours, you have essentially:
a). inspected the diameter of those cables going from your secondary to your bridge rectifier;
b). came to the conclusion that they are too thin so you have exchanged them;
c). added a beefy, High Ripple current rated electrolytic capacitor and soldered it DIRECTLY on the output pins (plus and minus - please observe the proper polarity!!!) of your bridge rectifier.
d). added the small little MKP / MKT capacitors, foil capacitors, in parallel to each of the diodes in the bridge, as a means of dampening the devastating impact of those nasty current spikes that take place during switching.

So, now that you have finished this small little DIY modification project, I suggest you open a bottle of wine, listen to the music, and ... enjoy. Enjoy a sound that now shall have much less "hisssing" Ssssssssssssssybilants. The Diana Krall .... I mean the girl in the other room ... shall exhibit a totally different "s" in every word she sings, ... provided that there is an "s" in that word.

You will simply fall in love in those new "Esses". It shall be an "s" like you have never heard it before. Enjoy.

Best Regards,


P.S. The article as above contains certain mental shortcuts and purposeful (and brutal) simplifications, so as to better "visualize" certain concepts, problems and issues. In my humble opinion, this simplified representation of physical phenomena is cruel, but justified, for the sake of clarity and keeping it "short".
This is not an attempt to provide classes on the theory of entry level electronics, but a layman belletristic visualization of some of the effects arising thereof,
and how we can get some "action" and quick results without the boring maths.
Alas, it always is a good thing to read a book or two on electronics and stuff, just to be certain that you know what you are soldering. Have Fun !

Matej Isak. Mono and Stereo ultra high end audio magazine. All rights reserved. 2006-2013. www.monoandstereo.com. ..:: None of the original text, pictures, that were taken by me, links or my original files can be re-printed or used in any way without prior permission! ::..