Mains Conditioners

Hi Adrian.

 

I personally don't bother with a mains conditioner and in fact if you use our products with a PSU1 we recommend that you plug it straight into the wall socket as the conditioners can cause a problem with phase.

 

I believe so strongly in what my ears tell me, not to mention my lab equipment, so strongly in fact that I have been designing, building and selling power line filters for seven years now ( http://www.dezorel.com ).

 

The benefits of a good power filter are many and across the board. From lower noise floors, to better detail rendition to sometimes spectacular spatial info deliverence.

 

The problems of power line filters are their necessarily high prices (since you are dealing with kilowatts of power) and their less than foreseeable results. One does know what one can expect in general, but since the overall results still do depend on specific system and location factors, one cannot reliably predict the scale and range of effects. This of course means that one might fail the cost/benefit ratio minimum.

 

In case of Graham's Solo headphones amps, specifically my own filters will literally amaze. I have had it happen to me, and over the years, I have retried the test many a time and it never failed.

 

Obviously, I cannot speak from Graham's point of view, I have no idea what kind of devices Graham had dealings with, so it's best Mr G tells us himself.

 

I have had a few interesting Close Encounters myself. My best example are very much hih end products from a well respected European manufacturer, naturally, all tube. Playing them on their own, I got a good sound, but a sound I felt was way below their price tag, one I could pay 1/5 or 1/4 of the price and have it. This was especially true of the fine inner detail, which was smeared to kingdom come. Then I connected one of my own power line filters and was amazed indeed - the sound volume went up a bit, but the sound quality went down quite a bit. After much experimenting, I concluded that the designer was not merely well aware of the incoming line noise, but in fact used it to mask certain circuit shortcomings.

 

I double and triple checked with similar products from other sources, and obtained broadly the same end reslt, albeit with considerable detail difference.

 

The point being that it can come to pass that line noise is actively used in the design, ridiculous as it may seem at first glance.

 

Overall, I'll be the first to admit this is a rare case, in my experience, unique, but where there's one, there are likely to be more. In most cases, this is hardly so, in most cases you have typically two problems inherent to power conditioners.

 

One is that of the dynamics - it is said that power conditioners limit dynamics. While hardly a truth carved in stone, it is nevertheless true in far too many cases to be overlooked as malicious gossip. This is generaly because of the sonic signature of the circuitry; no matter which components you use, each and every one will leave its sonic imprint, period. The trick is not not to have it, because there's no way you can avoid it, the trick is to have as little of it as possible, and then not in concentrated, but rather is spread-out form as much as you can. When dynamics are a problem, generally there are two reasons why this happens in most cases: 1) component quality in power filter is simply not good enough, and 2) there is too much magnetic saturation in the power conditioner transformers, never mind which type they are (E-core, toroidal, etc).

 

The second problem is extraneoius to the conditioner, but is nevertheless most important to its operation, and that is the ground potential. One way or another, most power conditioners use the ground to dump into it whatever garbage they want to remove. When designing a filter, one must start for a set of known factors at least in part, and the ground is assumed to be a known factor.

 

When in fact it is anything but known. Electrical safety standards assume ground to have a potential of 0.8 V or less, with an impedance of 100 Ohms or less. A friend from Paris, himself a most talented designer, reported that the ground in his apartment measured a potential of 80V !!! How is immaterial in this instance, but when your design assumes 0.8V and your product faces 80V, your design is 100 TIMES off. Then, ANYTHING can happen, but one thing is for sure - that filter certainly won't filter much.

 

And in between, a lot can happen. All of which brings us back to what I keep saying: the greatest problem with power line filters is that you don't know what the real effects will be like until you have it in your system. Sad, and working against me, but true.

 

On a purely technical level, Graham's power supply is being grossly misused. The filtering capacitor, for example, is being asked to filter out grit in the 10-80 kHz range, where about 98% of all line junk lives, which it really can't do at all, because at the same time, it is asked to do its regular job as well, which is to clean up residual 50/60 Hz leftovers after rectification and to act as an energy storage reservoir, all at the same time. If an external power filter can clean up the incomig muck before it even reaches the transformer, that transformer will hum less, will heat up less and will provide more of its nominal power because it is offloaded. That electrolytic cap will clean up 50/60 Hz residuals better because it is offloaded up in the range, and it should be subjectively faster beacuse it has no high frequency garbage impeding it.

 

Now, I cannot comment on the stability issues because I have not even seen, much less used, what Graham calls "green PSU". If this is related to switchmode power supplies, all I can say is that all my PCs and about a half of my lab instruments use switchmode power supplies, all are 100% of the time given fully filtered power and none has ever done anything untoward. In fact, they run somewhat (7...10%) cooler, or if you like, they are somewhat (7...10%) greener.

 

Graham, can you please comment on this, I am very interested? You obviously tested something with something, and the results were not pleasing?