AC/DC ... or a Highway to Hell


Your wall socket outlet, what's in there ?
Is it AC or DC ?

Did I hear you say: "AC" ?

Sure ?
Quite sure ?
Are you REALLY sure ? ( I mean: really, really, really SURE ? )

But what if you are wrong ? What if this is just a case of mainstream wishful thinking ?

What if I were to tell you that your wall socket delivers AC/DC ?
That it delivers BOTH ?

That your wall socket, that you use to power your audio equipment, provides you with lots of AC, and just a small little tiny tiny touch of DC ?

Let's say that you live in Europe, and your wall socket provides you with 230 Volts of AC ..... and on top of that, it provides you with 0.275 Volts of DC ?
Peanuts ? Is that what I heard you say ?
That I am splitting hairs ?
That such 0,275 Volts of DC is negligible ?

And what if I were to tell you that such a minuscule amount of DC, when entering into your audio mains transformer,
embarks your equipment on a Highway to Hell ?

At this point, I wish to express a great amount of thanks, to a great mind - to Mr. Rod Elliot, of Elliot Sound Products.
His site contains an ocean full of DIY projects, but ... let's just concentrate on the AC/DC issue at hand.
For a full, somewhat technical version, please check out the details under this source link:

For the sake of DIY efficiency, let's just say that I shall be skimming just the tip of the iceberg on this.
I suppose that I can give you a "short" layman's version of the problem,
of the remedy, and of the implications that this imposes on the sound of your system.

The bad news is: That 0,275 Volts of DC, that may pop up in your wall socket can potentially raise havoc
in the efficiency and performance of your mains power, audio grade transformer.

Within that source link as above, there is a Table no. 1, where it shows the conditions and the performance characteristics of
ONE AND THE SAME toroid transformer, that is powering a hiend-audio device.
The left column - shows typical operating conditions, when the toroid is working of an idealistic (i.e. not existing) AC mains input voltage.
The right column - shows the operating conditions of the same toroid, when it is fed from a contaminated AC mains input voltage.
Contaminated by a small DC voltage offset, of "ONLY" 264 millivolts.

But mind you, this small "little" DC offset voltage - it wrecks havoc in your hiend-audio mains transformer that powers your equipment.

The transformer magnetizing current (RMS - root mean square, as measured by any cheap multimeter) increases, in the presented example, from 15.7 mA up to 218 mA.
The Magnetizing current value, "peak-to-peak", increases from 50 mA to 1 full Ampere.
You can see, that the secondary Voltage starts sagging, and falls from 31.8 V RMS to 31 V RMS, and this with no load connected.
The no load amount of VA lost on this transformer rises from 3.77 VA up to 52.32 VA.
The primary Impedance falls down on its face - from a healthy 15.3 kilo-ohms to only 1.1 kilo-ohms.
The effective Inductance of the unit falls down from 48.7 Henry down to 3.5 Henry.

The data, as reported by Mr. Rod Elliot, is as measured on a typical 500VA toroid.

As you can see, is seemingly minuscule amount of DC voltage within your mains socket, a DC voltage that is a nuisance that is "superimposed" on your AC voltage,
causes a fairly dramatic complications.

Anyway, I'll say it another way: I have my own bad experiences with this phenomenon. As I may have mentioned, I have build an Output Transformer-Less vacuum tube amplifier. For it's construction, I use a 1200VA toroid (yes, that is ~1,2 kilowatts, give or take).

Now, I am seriously considering building this DC blocking circuit, as presented by Mr. Elliot of, and this is because that each time my woman switches on that damn hair dryer in the bathroom, and switches it on at half-power (as she always does), the toroid within my OTL amplifier in the living room starts groaning in agony.
In fact, it moans so terribly, as if it were short circuited - and believe me, I have seen many a short circuit in my life, with the fumes of melting and / or burning toroids that accompany that. For the sake of clarity - let's just say that I am "allergic" to the sounds of toroid transformers, that are moaning in agony.

The DC-Blocker schematic is TRIVIAL - it is comprised of only THREE elements: a high power rectifier bridge, and Two big electrolytic capacitors.
If you decide to do a full symmetrical version of the circuit, the count of elements shall increase to SIX. That is all.

At this point, you may ask: Where is the CIRCUIT DIAGRAM ? A good question. I purposefully do NOT include it here.
The reason is as follows: Lets just say, that the circuit is TRIVIAL. But at the same time - it is very dangerous, and if you decide to build it,
please kindly refer to the link as signaled at the beginning of the article, and read the material in full - especially the SAFETY considerations.
Believe me - it is a very easy add on circuit to you system, your power supply distribution box, or whatever you have there.
But you have to be fully aware of the dangers. And of the inherent risk of a resonance circuit, that you could build by accident.
Whatever you do ... DO NOT even THINK of trying to reduce the capacity of those electrolytic capacitors. It is big. 2x 4700 uF / 35V at the least.
DON'T reduce that. (unless you want to see a highly efficient resonance circuit powered off mains, frying toroids and exploding high voltage electrolytic caps).

But, ... where was I ? Ah yes. When you are listening to the quieter sections of the music that is playing, this "moaning" of the toroid - it is simply irritating.

Ok, you say. So I got a "moaning" toroid. But does this have an impact upon the quality of the "sound" of my system ?

Actually, it does. If you use the DC-Blocking circuit, you shall observe a "better" performance of your system.

But what is the correlation between something so abstract as 0,275 volts of DC in the mains socket and the sound of my system ?

I think I can "try" to explain that.
It seems to me that any audio device that has a solid, strong, stiff power supply with low internal output impedance, and hence with a stable and 'non-sagging' voltage delivered to all the interested circuits, one that delivers a constant voltage - irrespective of the dramatically changing variety of possible load currents - such a system is bound to sound "good".
However, consider the same device as before, with the same mains power transformer, but suddenly, it 'magically' turns out to be a flimsy, frail power supply, one with a high output impedance and hence, with a voltage that "dives", or sags, on each account that the circuit that constitutes the load wishes to pull some above average current from such a power supply. Such a system, subjected to an inefficient and sagging power supply - it simply must sound worse.

If you look at the article of Mr. Rod Elliot at the beginning of this text - especially look upon that first table with the results of his measurements,
it is enlightening to note how one and the same transformer can have completely different performance characteristics and working conditions.

How terribly those parameters change (to worse), as soon as the smallest amount of minuscule DC offset appears in the wall socket.
This seemingly irrelevant amount of DC current, flowing through the primary winding - this results in such gruesome effects.

But if you think about it - it is very logical.

Please concentrate on, or calculated, the "Amper-Turns" - which is the result of multiplying the current flowing through the primary, by the amount of turns that constitute the primary winding. This physical quantity, the amper-turns, when further combined with the magnetic permeability of the iron core within that transformer, actually gives you an idea of the total magnetic flux that is flowing through the iron core. A magnetic flux that might be just a tad too high.

Take a small DC current, say 50 milliamperes (peak-to-peak) and multiply this by the number of turns in the primary winding.
But how much is that actually ? Well, from my experience when fiddling with extra windings added to an existing toroid transformer, I observed that each additional "turn" around the core of a toroid transformer gives me roughly about 0.1 ~ 0.15 extra AC Volts per turn. This means, that a primary supposed to work from 230VAC will probably be be constituted from something like 230 / 0.115 = 2000 turns, give or take.

So, now, returning to our calculation: the "miniscule" 50 miliamperes of DC offset current flowing through the primary winding, multiplied by 2000 turns, gives a whopping result of 100 Ampere-Turns. Now, throw into this the fairly high magnetic permeability of the iron core, and you embark upon an AC/DC Highway to Hell,
as the magnetic flux in your iron core is of very significant values, expressed in Gauss / Tesla, and may be dangerously approaching the core saturation levels.

I suppose that "close to saturation" operation conditions of the iron core are by no means optimal for such a toroid. Especially from the viewpoint of the secondary winding ,
which is trying to generate your secondary voltage and currents that are needed to feed the rectifier, the filter capacitors, and ultimately - your whole hiend audio device.

Thus, in magical terms, the one and the same power supply mains transformer may have times, where it is "Strong" and "Powerful", capable of hefty current capability,
but at other times, it suddenly becomes a flimsy, clumsy weakling, that falls down on it's face every time a meaningful current is demanded by the load. Every such meaningful demand essentially translates to a sagging voltage.

But these - are only my wild assumptions.

However, let me provide you with an insight as to my practical observations:

As soon as my woman turns on that damn hair dryer in the bathroom, with the 1/2 power switch activated (essentially constituting a silly rectifier diode, that unsymmetrically loads the mains sine wave), my 1400VA transformer in the OTL begins to complain. This coincides with the voltage readings on the output power tubes slightly sagging in value.
Not much, but it is a correlation that I observe time and again, so it is not a coincidence. The sound of my poor feeble 1400 Volt-Ampere toroid issometimes very much similar
to that of an overloaded welding machine. When I first heard it , it was such a shock and I was worried to such an extent, that I was about to pull the plug out of the socket.\
At that time I did not have the slightest idea of the reasons that are causing this phenomenon, and irritatingly, why is it "coming and going" - it suddenly appears, and then, for seemingly no obvious reasons, just disappears again, just as suddenly, as it appeared.

As mentioned, I am "allergic" to the the sound of agony and the moaning of a toroid. Any toroid, because it then seems that there is some fault in the secondary circuit, some short-circuit, and that the fumes shall shortly appear again. The more so, because it happened to me once before - I actually fried a toroid once, it melted the plastic insulations and the stinking fumes were abundant. The breakfast tray under it - that one melted also.

One more observation. When the damned, cheap hairdryer is shedding havoc, the music from my system is still playing.
But, providing that the process of hair drying lasts just a bit too long, I can actually observe, that the Power supply block of my OTL, ...
... it actually becomes much hotter, than in normal operating conditions (i.e. = with no hairdryer being used in the bathroom).

Happy soldering ! ... and PLEASE, do not get yourself killed, while you are at it.
Death - is permanent. And most of the time - it is also painful. Common sense - applies.

The capacitors need to be equipped with RADIATORS.
That essentially means, that they need to be of a BIG SIZE.
This is necessary, so that they can cope with disseminating a fairly large amount of heat from their body.
They need to be capable of SECURELY handling fairly high amounts of ripple current.
The only way to achieve that is by sheer volume.
That is the reason, why we pick the 63V versions for those electrolytic capacitors.
This is not due to the voltage, but a simple fact that electrolytic for higher voltage have a bigger volume,
and hence, a better heat radiating surface, and hence, a higher ripple current handling capability.

The bipolar capacitors are connected in series, with the "+" signs facing each other.
This is so as to "produce" a unipolar version capacitor, i.e. one that performs equally good in both directions of the incomming DC voltage.
We must assume that the mains DC offset may be of any polarity. That is why we need to "unipolarize" the bipolar caps.
Just to be extra safe, I added three parallel capacitor pairs. Simply to provide more heat dissipation capability,
and to be on the safe side.

As for the Rectifier Bridge: this needs to be absolutely FAIL-PROOF.
The diodes are not allowed to failed. Therefore, we over-dimension them, to the border of sanity:
Take the 35 Amperes, 1000V version of a bridge. At best the one in a cast metal screw-on body.

If you wish to create a totally symmetrical version of the circuit, one that does not depend on the "LIVE" to always be on the "Live" side,
you can simply duplicate the whole contraption and make a similar one on the "NEUTRAL" lead.

Enclose the whole thing in a closed box or something, so that if a cap wishes to explode, it has the freedom to do so and not harm anybody.

Happy soldering.
Matej Isak. Mono and Stereo ultra high end audio magazine. All rights reserved. 2006-2013. ..:: 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! ::..