Line Conditioners or Not and Why

Richard Dane posted:

Paristhea posted:

Richard Dane could perhaps give me some reply on the topic, if available.  I would really appreciate this.

I can't give you technical insight - I'd leave that to someone like Roy George or Steve Sells at Naim HQ (they will talk about impedance or some such thing) - but I can tell you of my subjective experiences.

The first was when we had a few mains conditioning units at the factory to try out.  All bar one were highly recommended by the various forums, magazines, the odd one out was a special lab unit that had come from the States and had been recommended by a few "in the know".  All made the music sound flat and uninteresting to my ears. One or two of them managed to do this by just being plugged in to the mains next to the system, rather than having the system itself plugged into the unit. 

Next experience was at a dealer event in Sweden.  I had spent the late afternoon clearing the main dem room of extraneous kit, ensuring all bar the Naim system (CDX2, Supernait) was plugged in and setting up the Fraim stand and speakers.  Once all was just so, we sat down to try out some of the CDs and tracks that I always use and are now so familiar to me - this way it's easy to know when the system is doing things right or not.  It was good in many ways, but it was also just a bit "flat" and wasn't soaring the way I knew it should have done.  Around the room were curtains that hid flush shelving.  I had left some items on the shelves because there was nowhere else to put them.  Nothing else looked like it was  switched on.  However, I noticed what looked like a big silver fronted power amp and it had the logo of a well known brand of very expensive mains conditioners on its front.  I pulled it out from the shelf and lo and behold, it was still plugged in to a mains socket.  I unplugged it and put it back in place on the shelf. I drew the curtain and we all sat down again to listen...

Suddenly the system had life.  It soared.  It was quite a change - hard to pin down but undoubtedly the music seemed freer and less constrained, with better flow and much more "real".  It was so profound a change that we all just sat smiling, laughing, slightly stunned that it should have made such a difference to the system.  

So, that has been my experience.  

Thank you Richard.  That is why i had asked for your input in this subject.  

Basically and in simple terms, i understand that there is no clear cut reasoning to add a line conditioner in a hifi system.  Why some times we think that it adds something to the sound quality, is it a misguided perception/impression?  In my case i had definitely noticed/observed that the background gets quieter, but i can not say with certainty that the sound quality was any better with the turntable as a source, i think i would prefer the TT sound without the line conditioner.  With the CD as a source, i could say with certainly that the bass became firmer and i have a perception that the sound quality was better, i do not know why.

But i am also wondering, and please accept my apology for what i am about to say, whether hifi manufacturers choose to say NO to line conditioners, in order to sell their own PSU's to their customers?  I assume that a PSU does incorporate an element of line conditioning, or mains filters at least?  Or am i completely off the mark here about PSU's?

Regards,

Paristhea, you are off the mark about PSUs.  Linear PSUs are normally built without line conditioning elements built in.  The mains interference rejection comes from good engineering design of the entire system (PSU, amplifiers, DACs, etc.) not just the PSU.

Mains conditioners are likely to have a negative impact for reasons I stated previously.  The real question though, is whether those negative effects are worse than the presence of uncontrolled mains born interference.  In most places the mains interference isn't too bad, and the mains conditioner will degrade the sound more than the interference.  However, if the mains interference level is particularly bad, then the mains interference can have a worse effect than a mains conditioner; in this case a mains conditioner may improve things (but that's still not guaranteed).

Line conditioners and PSUs for hifi equipment are completely different things, so please do not confuse the two.  Every item in the system (unless passive speaker, clockwork or wind-up) needs a power supply.  The quality of that power supply will have a very profound effect on the performance of the item in question; in general terms, and all else being equal, the better the power supply, the better the performance.  Some come outboard, from wall warts all the way up to PSUs like the Naim 555PS et al.. while  others are inboard, most of which are integrated within the item and cannot be easily upgraded.

There's no clear cut reason to add a mains power conditioner to any system.  In the context of a Naim system I would say that there is plenty of reason to avoid such a thing altogether.  When I asked Roy about what kind of power conditioner might work best he opined that he felt such a thing was possible but that it would need to be about the size of a house, or at least a small power station.  He talked about mains impedance and such things, at which point I felt decidedly out of my depth, so please excuse my lack of expertise or ability to explain things further.

Thanks Huge, good to know this about PSU's.

I therefore conclude that the best method is to go by one's ear as well as common sense approach, about the expenditure on the line conditioner, if at all.

How can a line conditioner degrade the SQ though, it beats me.  The signal happens after the line conditioner, downstream so to speak.  Hence how is it possible that the signal gets affected since it occurs after the line conditioner has done its part?

Hi Paristhea, as Richard says, it's mostly about impedance.  To simplify (but without loosing all gignificant meaning!) if a power supply has a high impedance, it won't be able to deliver DC current to the amp's capacitors fast enough to keep up with demand.  The DC voltage will drop briefly and alter the way the amp performs.  It's possible to hear an amp struggling to keep up like this.  Similarly if there's a higher impedance in the mains supply, the same thing happens (but somewhat less noticeable).  If the mains voltage has ripples on it (for instance from the resonance of a mains conditioner), than a similar thing happens and this will also affect the DC voltage supplied to the amplifiers.  These small changes of DC do slightly break through into the output signal of the amp and affect the sound.

Hence mains conditioners do affect the sound of an amp.  Mains interference also affect the sound of an amp, even though the designers go to great lengths to minimise the effect, some small effect always remains.  It takes extreme mains interference to cause as much degradation as an average mains conditioner.  So as most people don't know the type of, and extent of, any mains interference they actually have, there's no clear cut reason to add one.  Even if you have extreme interference, there are usually down sides to the mains conditioner as well as some possible gains, so even here the case isn't clear cut.

Sorry I can't give a simple absolute yes/no answer, except to say that yes is only very rarely the right answer and then it has to be checked, justified and re-checked before believing it!

Paristhea posted:

How can a line conditioner degrade the SQ though, it beats me.  The signal happens after the line conditioner, downstream so to speak.  Hence how is it possible that the signal gets affected since it occurs after the line conditioner has done its part?

One theory is the X&Y capacitors affect the sinusoidal wave form & the distortion is carried through to the transformer secondary winding.   It is possible that it causes higher than normal DC ripple after rectification.  

I did say theory w.r.t. the subject we are discussing.  However I have seen it in real life with an industrial application & incorrectly spec'd voltage converters

Line shampoo anyone?

Thanks again Richard.  Although i am an engineer, i would too not understand the fine detail either, as i am not electrical.  But as i asked Huge just before, i can not understand how the SQ can be affected by a line conditioner, since its work is done before the sound signal is even generated?  I would assume this could happen because the sinewave is altered in such a way by the line conditioner, which in turn affects the performance of the electronics downstream of the line conditioner, maybe!!!

One think i observed at my home in Cyprus, was as follows: I bought an LP12 with a Lingo PS and Ittok LVII in December and was delivered just before Xmas.  Once i connected it and switched it on there was a strong and annoying roaring noise coming through the speakers, only when i had the input selector on the analogue/turntable.  I had at first thought it was something wrong with the LP12 or the Lingo, then the Ittok, then the 5-pin connector, then the phono stage settings.  None of that.  After long and exhaustive trials i finally discovered that it was the DSL router and range extender (they are on the same power extension) that happens to be in the same room, far way on another socket, which most probably happens to be on the same ring circuit. When i switched it off, the roaring completely disappeared.  Now i have it connected on a power socket from another room.  So yes, funny things happen from the mains power supply, which is definitely beyond my understanding.

Anyway, having considered all options, i decided not buy a line conditioner.  In any case, and if i was to buy one it would have been the Isotek Gii Sigmas (second hand), but i can save the 1,000 Euros or so.

Call me old school - but power line conditioners to me have always been about correcting power factor - which can be a curse on AC lines and cause inefficiency and therefore heating - because as we should know the current sine wave and voltage sine wave occur at different points in time and can be different in shape.. from memory impedance doesn't change this relationship - but it does the regulation.

Simon, you're old school 

And technically correct 

phew I thought I was loosing it

Don't worry I lost it years ago - now where did I leave that box of marbles?

Hi Paristhea, are you a mechanical engineer?

If so I may be able to provide a mechanical analogy as to how the mains voltage can affect the output of the PSU, and how that in turn affects the sound of an amp.

Huge, I am not a mechanical engineer as you know, but I would interested to read your analogy

Huge posted:

Hi Paristhea, as Richard says, it's mostly about impedance.  To simplify (but without loosing all gignificant meaning!) if a power supply has a high impedance, it won't be able to deliver DC current to the amp's capacitors fast enough to keep up with demand.  The DC voltage will drop briefly and alter the way the amp performs.  It's possible to hear an amp struggling to keep up like this.  Similarly if there's a higher impedance in the mains supply, the same thing happens (but somewhat less noticeable).  If the mains voltage has ripples on it (for instance from the resonance of a mains conditioner), than a similar thing happens and this will also affect the DC voltage supplied to the amplifiers.  These small changes of DC do slightly break through into the output signal of the amp and affect the sound.

Hence mains conditioners do affect the sound of an amp.  Mains interference also affect the sound of an amp, even though the designers go to great lengths to minimise the effect, some small effect always remains.  It takes extreme mains interference to cause as much degradation as an average mains conditioner.  So as most people don't know the type of, and extent of, any mains interference they actually have, there's no clear cut reason to add one.  Even if you have extreme interference, there are usually down sides to the mains conditioner as well as some possible gains, so even here the case isn't clear cut.

Sorry I can't give a simple absolute yes/no answer, except to say that yes is only very rarely the right answer and then it has to be checked, justified and re-checked before believing it!

Hugo thanks so much.  Fantastic reply, i understand the technical issues much better now.  I had to go to wikipedia and other sources to refresh my memory as to what exactly impedance is, but having done that i am happier now.  So when you see a lot of coils inside a line conditioner, it is not a good sign in principle.  I am sure it is not so simple as these companies go to great lengths to design and test them, but at least i understand the reasons why not to have them.

And no i do not have any particular interference at my house.  I had a problem, as i explained in my previous post, but that was simply caused by either the router or the range extender, and it is now solved 100%.

Thanks again.

Simon, I'm shortly off for a games night with friends, so it'll have to wait 'til tomorrow, but I will post it.

Huge posted:

Hi Paristhea, are you a mechanical engineer?

If so I may be able to provide a mechanical analogy as to how the mains voltage can affect the output of the PSU, and how that in turn affects the sound of an amp.

Yes i am a mechanical engineer, although i am now a consulting engineer for building services

OK, the Analogy (warning MEGA length post)

The Power Supply

Consider a device where there is a circular disk mounted eccentrically on a shaft and the shaft rotates.  The rotating disk (a cam) pushes against two weights at exactly opposite sides of the shaft, that are drawn back inwards by two springs.  If the disk rotates fast enough, the springs won’t manage to keep the weights in contact with the disk.  Faster still and on each rotation the weights will just be nudged back out again and there’ll be almost constant tension on the springs.  This is an analogy for the power supply.  The disk is analogous to the mains 50Hz, the nudging of the weights is the rectifier and the tension in the springs is the DC +/- voltage rails.

The Power Amp

Moving on to a power amplifier, instead of the springs of the ‘power supply’ being connected to a fixed point, each one is connect to a gas strut, and the two gas struts are connected together in the middle.  There is also a distribution valve that controls the pressure in each of the two gas struts.  This is arranged so that as the distribution valve moves it lets more pressure into one gas strut and releases pressure from the other.  In the neutral position the pressure in both struts remains the same.  The position of the distribution valve is primarily controlled by the position of the ‘input’ lever connected via a spring; however there is also another control mechanism superimposed on this.  The position of the junction of the gas struts operates the feedback lever, the other end of the feedback lever moves by 1/10 of the change in position of the junction of the gas struts, but in the opposite direction.  This is also connected to an equally strong spring that’s also connected to the distribution valve.  (***)

As the input lever is moved, the distribution valve moves and one strut gets more pressure, the other less pressure.  The junction between them moves, but this alters the position of the feedback lever and that then starts to balance out the effect of the input lever.  When the gas struts have moved 10 times the distance that the input lever has moved, the distribution valve returns to the neutral position and the gas struts stop moving.  As the input lever is returned to the neutral position the reverse happens until the gas struts also balance out at the neutral position.  If the cycle is repeated in the opposite direction, the gas struts move the other way.  In each case the movement from the gas struts is 10 times greater than the movement of the input lever.

Unfortunately for the electronic version it doesn't end there, the PSU springs provide the driving force for each moving element, so they are weakly connected to the input lever, and the diverter valve, and they provide the pressure to drive the gas rams (don’t ask how, this is just an analogy!).

Mains interference

If the bearings on the shaft are worn, then the movement of the disk will become erratic and jerky, rather than the smooth eccentric action of before.  This roughness and jerkiness will be transferred to the springs and from the springs to the gas struts.  As a result the final position of the junction of the gas struts won’t just reflect the position of the input lever but will also be tugged about but the errors in the movement of the eccentric disk.  This effect is made stronger as all the stages are tugged around by the PSU springs, and most significantly they provide the pressure for the gas rams.  In other words, the PSU provides the energy to make the whole thing work, so it is linked to and influences every single bit of it.

(*** At this point some people some people will recognise the long tailed pair, the driver transistors, the output stage and the negative feedback loop.)

For those who want to take the analogy a little further and understand automotive mechanics, here's an extension to include the pre amp.

In addition to powering the power amp, there's another weight and spring that powers a smaller vacuum pump connected to a vacuum flask.

In the case of an integrated amp with just one power supply this is driven by the same eccentric disk as the power amp.  In the case of a separate power supply for the preamp (Flatcap, Hicap, Supercap or 555) there is a separate eccentric disk driving this weight.

The vacuum flask has a pressure regulator which is 99.9% efficient at regulating the pressure.  However it isn't perfect, it still lets through some variation, particularly if there's a sudden demand on the vacuum or there's a shock wave from the PSU.

This vacuum flask is connected to the brake servo from a car, and the output from the brake servo instead of being connected to the master cylinder, is connected to the input lever of the power amp.  The 'normal' position for the 'break pedal' operating the servo is half way depressed, and that puts the input level of the power amp in the neutral position.  moving the 'break pedal' up and down sends a signal to the power amp.

So far so good.

But this eccentric also has worn bearings, and that sends pressure variations down the vacuum line and a small proportion of these gets past the pressure regulator causing the break servo to twitch.  This twitch is then passed on to the power amp.

The DR regulator is faster and more accurate than the integrated regulator it replaces, so it attenuates shock waves more than the previous regulator, so sounds cleaner.

The advantage of a separate power supply for the pre amp is decoupling.  In the situation where both are driven from the same eccentric, if the power amp demands more energy from the PSU springs, more load is placed on the one eccentric and it will have to work harder to keep up,  This means that there's slightly less energy available in the eccentric to drive the power supply for the pre-amp and a small proportion of this reduction in energy is passed to the preamp itself.  So demand on the power amp (i.e. loud music passages) also make the pre amp less efficient and the system has a less dynamic sound.

Brilliant Huge.  

I have no idea what any of it means but it all sounds most impressive and I'd love to see it in action - if you built it out of say, Lego or Meccano, there's probably a place reserved for it on Youtube or some such site.

Astonishingly clear. I am neither an electronics nor a mechanical engineer, but Huge's analogy is the best and most accessible explanation I've ever read. Thank you. 

Hi Richard, the trick would be to drive the whole thing using a 78, as they don't have RIAA equalisation.

Huge, very good, you had me laughing.. Babbage's theoretical design for his Difference Engine sprang to mind 

I had a puzzling muse on your powersupply analogy, those weights would equate to the capacitance of reservoir capacitors would they not?

Also the RFI where it was external could equate to vibration affecting the stand supporting your contraption perhaps?

With an imagination like that you might have missed your vocation as a children's technical book illustrator..   I have a good publisher friend who has discussed this sort of idea for illustrated children's book to explain physics and make accessible for kids... 

Huge posted:

Hi Richard, the trick would be to drive the whole thing using a 78, as they don't have RIAA equalisation.

Huge you may like to have a look at this non-cmmercial link concerning the EQ used for 78 recordings.

http://www.mil-media.com/pdf/M...078%20EQ%20Chart.pdf

ATB from George 

Simon, the capacitors are represented by the weights and springs.  Without the springs no energy would be stored.

An alternative would be a dual cam system using gravity, but that looses the view of the + and - rails being in opposite directions and self balancing (or rather approximately so - hence some of the non-ideal behaviour when the mains isn't perfect).