Tesla puts it together

Now we’re getting to the core of understanding how power in our hi-fi system is transferred from the wall and in the next few posts we’ll start to understand how we convert this power into music. First let’s wrap up our little history lesson on why we use AC power at all.

If you’ll remember, Tesla was getting ready to change the world from DC to a new form of DC called AC, something we use in our homes to power our stereo systems to this day.

Nikola Tesla was poised to build his AC power system based on the idea that each home or block of homes would connect to a very high voltage AC main line through a transformer – the transformer reducing the high voltage to a lower and more usable voltage for the home.

As we’ll remember, transformers only work when the power that feeds them is moving between plus and minus rapidly – something we call Alternating Current or AC. But Tesla had a problem and in a matter of only a few days he solved that problem and created the modern power grid still in use today.

The problem he had was generating the AC and once generated using it directly to make a motor work. In those days, dynamos or generators as well as motors were all generating and running off of DC. The dynamos that generated the DC and the motors than ran from DC were essentially the same things – each used in reverse of the other. This was sort of the same thing we’ve seen before with coils and magnets – if you spin a motor it makes electricity and if you put electricity into a motor it spins.

It was known even by Michael Faraday a decade earlier how to generate AC – one simply used a physical magnet connected to a hand crank in close proximity to a coil of wire. When you spun the magnet (a long straight bar of metal turning end to end) it would change the magnetic field from plus to minus and what came out of the coil was AC. But no one knew how to make a practical design for that and, perhaps more importantly, no one knew how to use that AC to make a motor.

Over the course of only a few weeks, Tesla saw the entire system in his head and in one of the great strokes of genius figured out the modern power grid, invented the AC motor, the AC dynamo, the efficient AC power transformer, and the system that would make it all work. Tesla’s grasp on the fundamentals of coils, electricity, magnetism and the efficient means to use all those elements together is unmatched since the genius of Michael Faraday who set all this in motion.

Within but a few years time, Westinghouse and Tesla’s new power system took over the modern world and sent Edison home to work on other projects like the phonograph, the motion picture camera and the thousands of innovative products he invented or brought into the world – but power was not to be his fame – it was Tesla who would make the leap necessary to build the world’s power grid – still in use today with but few changes.

Edison never gave up the fight for DC, even inventing the electric chair to prove just how dangerous AC power was. Edison personally electrocuted hundreds of animals and more than his fair share of condemned prisoners in public showings around the country – all in an effort to prove to the world AC was not to be used. But in the end AC survived even to this day.

Here’s a picture of Edison personally strapping in the first victim of the electric chair he invented to prove how bad AC was.

So, now you understand why AC is what comes into our homes and why.

Paul McGowan – PS Audio, Intl.

Artist tools

A few readers have asked me to back off on being uber technical and get back to writing more about sound and high end audio from a listener’s perspective rather than a designer’s and I’ll definitely do that but I did want to finish up on our little series about DACS.

One thing to remember is that while most of us just want to use our equipment as black boxes and not really care what’s inside, I think it’s always good to have some level of understanding about their workings to help make buying decisions. After all, an informed buyer is usually better equipped to make a choice that will bring greater musical joy than one without any understanding at all.

The first choice a designer has to make when it comes to the output analog stage of a DAC concerns gain: how loud the output of the DAC needs to be. This is an interesting dilemma for a number of reasons and one of the primary ones concerns what I just wrote about: making smart choices when it comes to purchasing your audio equipment.

Loud sounds better than soft. Sorry, it’s just human nature and those of us having been around long enough have watched the audition process too many times to ignore it. Most users will always pick the louder DAC than an identical sounding one with less gain. I have heard this described as sounding “more powerful” and “bigger” by those that don’t gain match their audition pieces.

Secondly, the designer also has to determine use case: does the DAC have an internal volume control and if so is it likely to be used with or without a preamplifier? This is a tougher choice as an increasing number of DACS have built in volume controls.

If the user is going directly into a power amplifier, as we ask people to do with our PWD, then that unit must have enough gain to drive the amp without turning the DAC all the way up. Why? For two reasons: the first and most important is that you have enough volume to play all recordings at maximum level with any given amplifier. The second is the age old misconception that when a volume control is nearing the end of its range the sound can become “strained” as if it were a gas pedal on a car and taxing the engine. Like it or not, as designers, we have to give enough gain so people don’t panic that they’re running out “room” and straining the system – despite the fact they are not.

As I have written in past posts, this is completely wrong but the ideas remain for new users and older ones as well. Think of the volume control more like the car’s brake, as opposed to its gas pedal.

Once the designer has figured out what gain they want, the next step is how will the output stage work and sound – in other words, what kind of configuration will it be? We’ll cover some of that tomorrow.

The one thing you might want to take away from this article is the following: whenever you audition a new piece of gear, do your best to first gain match as closely as you can to whatever your reference is. Don’t be fooled by a “louder is better” decision.

Paul McGowan – PS Audio, Intl.

Smaller is better

Boy, that sounds like a no-win claim. 🙂 In the case of a transformer, however, bigger is always better relative to what the designer is trying to accomplish from the viewpoint of power delivery at a certain frequency – at least from a sonic standpoint that’s true.

Years ago we pioneered the use of oversize power transformers in audio equipment and to this day, our power transformers are easily 50% bigger than most other company’s transformers in source and control equipment. Bigger transformer have lower output impedance because of their heavier gauge wires in the coil and the sonic differences we hear between smaller and larger are significant.

But when we deal with a Switch Mode Power Supply (SMPS) the landscape is very different for a number of good reasons. Here, our traditional thinking about transformers, power supplies and what makes something sound a certain way are all turned upside down and we need to apply a different set of rules.

In yesterday’s post we covered the fact that the size of a power transformer is directly related to the frequency it is working at: the lower the frequency of the wall AC the bigger the transformer has to be to deliver power. This also means the opposite is true and this is where it gets interesting for SMPS.

A 1000 watt SMPS can be built in a form factor smaller than a CD case. In fact, advanced version have been built not too much larger than a pack of cigarettes. A linear supply of the same size would be huge – just picture a big power amplifier weighing in at 50 or so pounds – most of that being the power supply.

So what’s the difference? Mainly the size of the power transformer. On a SMPS the power transformer has been shrunk to almost nothing, yet it still manages to isolate and deliver huge amounts of power. How does that happen? Higher frequency.

If you’ll remember yesterday we saw that a transformer running at 50Hz is approximately 25% larger than one running at 60Hz. Now imagine that same scale in reverse: run you power transformer at 100,000 Hz and the size drops down to almost nothing.

How in the heck do they do that? Sounds like a line out of Modern Marvels, eh?

Tomorrow we shall see.

Paul McGowan – PS Audio, Intl.

Transformers

In yesterday’s post we covered the reason why we need a power supply: getting the correct voltage to our equipment and providing safety isolation from the wall socket. Today I want to cover transformers and their size requirements: understanding this will be key to getting a grip on how a Switch Mode Power Supply works.

The size of a transformer is dependent on two main factors: how much power you want to get through it and the lowest frequency you operate it at.

If you’ll recall the transformer has two coils of wire: an input coil that attaches to the wall socket and an output coil that connects to our equipment. These two coils sit close to each other but do not have any electrical connection between them – instead they share a magnetic connection. When we want to get some power through our transformer we are generating a magnetic field and the size of that field is dependent on how much power we want: bigger field equals more power, smaller field less power. We control the size of the field by the size of the wire and the amount of steel in the transformer: heavier wire and more steel give more power but make for a physically larger transformer. This is why a power amp transformer is really big while a preamp or DAC transformer is really small.

The second factor in transformer size is frequency. I know this sounds technical but it’s actually simple. The wall voltage is AC which means it is constantly changing its direction: positive to negative and back again. This is the same as would happen if you had a battery and you simply flipped the battery around so many times a second.

Homes in Europe and Asia have the voltage flipping around 50 times a second and here in the US it’s 60 times per second. To transfer energy between the two coils of our transformer, the magnetic field needs to be moving back and fourth as described. If you just put plus voltage into the first transformer coil, it becomes a permanent magnet – but it won’t transfer energy into the second coil. Only the movement back and fourth of the voltage creates a changing magnetic filed and it is this change that makes it possible to transfer the energy from one to another.

The faster you change the magnetic field, the easier it is to transfer the energy – the more efficient the process is. This makes sense when you think about it: no movement means no energy transfer, slow movement means a little transfer and lots of movement equals lots of transfer. This is super important to understand so read again if you don’t quite get it.

So in the case where we have slower movement, 50 times a second relative to 60 times a second, we have less energy transfer ability in our transformer. How do we solve this? Make a bigger transformer with more steel which will help make a bigger magnetic field to compensate for the slower movement.

Here’s what’s interesting – the size of the transformer goes up dramatically with even a small change in frequency. 50Hz transformers are sometimes 25% larger than 60Hz transformers depending on the manufacturer.

Since most hi fi manufacturers like PS Audio sell all over the world and do not want to stock multiple transformers for each product, we simply specify all our transformers to work at the lower frequency of 50Hz. Then, when you put it on 60Hz, the transformer just loafs.

Tomorrow we get smaller.

Paul McGowan – PS Audio, Intl

SMPS

Switch Mode Power Supplies, or SMPS, just sounds nasty. No, not the supply itself but the word “Switch Mode”. Shades of choppy, digital, not linear, radiating, messy. Definitely something you want to stay away from.

A SMPS is a linear power supply only it doesn’t have a big power transformer like what we think of when we mention a linear supply.

Audiophile wisdom suggests that linear supplies are better, especially in power amplifiers, than SMPS. If you try and nail down why this thought prevails the arguments quickly fall by the wayside because, well, we don’t actually know why. It’s just that a digital power supply seems so much nastier than one that isn’t – in the same way that a digital audio product seems so much more messy than an analog music system – or an automobile electronic ignition system vs. a simple mechanical distributor.

I thought it might be helpful to spend some time examining what each of these two types of supplies actually are, how they work, why they are different and how they affect the products we listen to.

To start off let’s ask the question: “why do we need a power supply at all”?

What comes out of our wall sockets at home is pretty useless to us for hi-fi. 120 or 230 volts moving between + and – 60 or 50 times a second isn’t something we really want to try and make music with. Instead, we want a steady state + and – voltage, like that from a battery, at much lower levels: perhaps 5 to 30 volts for a solid state preamp or DAC and as high as 100 volts for a tube or power amplifier. None of that comes out of our wall sockets.

To further complicate the issue we have safety concerns to deal with as well. We don’t want to have our music systems capable of connecting us directly up to the wall socket – like we can with an ordinary appliance like a hair dryer or toaster. We’re a bit more cautious and concerned with the machines that make our music relative to the machines that dry our hair and make our toast – which makes sense because in the case of our stereo system the output is an electrical connection (interconnect or speaker cable) and in the toaster’s case, a piece of non conducting bread and hot air for the dryer.

So the job of the power supply is to reduce the voltage from the wall, change it from AC to DC and protect us from possible electrocution. The main element that makes this possible, in both the linear and SMPS, is the transformer.

We’ve already spent a great deal of time on transformers, their importance to us and how they work, in these posts but I know it’s probably helpful to review.

A transformer is nothing more than two coils of wire, one hooked up to your wall socket and the other connected to your equipment’s power supply (there’s also a whole bunch of steel to help the coils work better). When the AC power moves back and fourth between + and -, the coil of wire produces a magnetic field – meaning it becomes a magnet – and so the power in your wall socket is converted to magnetism. The second coil of wire picks up this magnetic field and reverses the process to make electricity again.

The amount of wire in each coil determines the output voltage of the transformer – meaning if there’s more wire in the wall socket side of the transformer (called the primary) than in the output side (the secondary) the transformer puts out less voltage than what goes in. The protection occurs because the two coils of wire don’t actually touch each other and the only connection between them is magnetic, not physical.

So in each case, the linear and the SMPS, we have a power transformer that connects the wall socket AC to our equipment safely and at the proper voltage.

Tomorrow we’ll start to understand what’s different between the two.

Paul McGowan – PS AUdio, Intl.

Audio Clubs

Saturday night in mid February we will be hosting the Colorado Audio Society as they hold their annual elections.

The CAS group are the same folks that put on the Rocky Mountain Audio Show and a terrific group of Audiophiles and music lovers.

The evening will be fun, members walk through PS Audio, listen in our sound rooms, talk with our team, see some new stuff and have a good time. Last year we had a load of hand crafted pizzas for the group, we have to figure out what we’ll do this year.

I bring this up in my posts because if any of you reading this have ever considered joining an audio society or starting one, I would like to encourage you to do so. The camaraderie , sharing of ideas and general good cheer shared by people who love music and audio is a very positive side of what we all love to do.

We’re proud to be a sponsor and open our doors to CAS and would encourage anyone reading this post to consider engaging in such an activity in your local area. It’s fun.

Paul McGowan – PS Audio, Intl.

A day in Sea Cliff

Following yesterday’s post about output impedance I was reminded of a day in Sea Cliff New York where HP (Harry Pearson) the founder of TAS lives.

This was many years ago and Arnie Nudell and I were running Genesis Technologies, a high-end loudspeaker company based here in Colorado. We had arrived at HP’s and setup a new pair of loudspeakers we wanted him to review.

All Genesis high-end loudspeakers had built in subwoofers, a trait started by Arnie while at Infinity. Arnie believed (as do I) that loudspeakers should be full range to the best of their abilities – adding extra boxes to make them reproduce music properly was a wrong headed idea that placed additional burdens on customers they didn’t need.

The built in subs had two means of sending the audio signal to them: an internal feed from the main speakers or an external feed from the preamplifier. The external feed from the preamplifier was the preferred method since the first method took the output of your power amplifier and necked it down to get the gains correct: the preamp feed therefore being the cleaner of the two.

When we go to visit a reviewer we rely on their existing setup: in the case of a speaker manufacturer the reviewer’s electronic chain is used and, in the case of an electronics manufacturer, the reviewer’s loudspeakers are used. We, being a loudspeaker manufacturer, relied on HP’s system of the day which, if I remember correctly, was some sort of tube preamplifier and power amplifier: perhaps Conrad Johnson. The preamp and sources were located well away from the loudspeakers, the power amp near them.

Arnie’s routine for speaker setup always begins with the subs off. This is a good idea because if you methodically setup the midrange and tweeters sections for best tonal balance and imaging without the subs on, then you can easily focus on just the mids and tweets contribution without also wrangling the woofers around.

He spent the better part of the morning tuning and tweaking and when he was happy, asked me to connect the subs and then we would go to lunch – dialing in the rest of the system upon our return.

Following our return from a delightful meal at one of the many great restaurants the town has to offer, we returned and started listening again, this time with the subs active. It sounded awful, very much unlike when we left for lunch. Arnie glared at HP’s setup people and asked if they had “messed” with anything. ”Nope”.

Retracing our steps I turned off the subwoofers via their power switch – no dice, still sounded like crap. Then one of us got the bright idea to start unraveling our steps, one at a time. So the first thing I did was disconnect the long XLR cables from the preamp that was feeding our subs to complete the reverse process. Bingo! The system jumped back to life.

Puzzled, I repeated the steps and added back the XLR cables from the preamp to the subwoofer input and the same collapse of the musical life took place. The preamp, a tube, simply did not have the current capabilities to drive two long cables – one feeding the main power amp, the other feeding our subwoofer. Our subwoofer input was something I designed and I knew it well – it’s input impedance a mere 100k Ohm so it wasn’t dragging down the preamp – it was the cable itself.

Without saying anything to HP we simply switched over to the internal subwoofer connection, finished setting up and at the end of the day went to dinner and then home. We got an excellent review.

The point of this story is to show just how critical output impedance and drive capability is on a piece of equipment. Tube preamps are by far the worst for driving long runs of cables if they are not designed with that in mind – but all products can have the same problem.

Bottom line: test your assumptions. Don’t just assume the designer has done his or her work in a way that suits your particular needs. Try it with and without, in and out, and make your own determination.

Paul McGowan – PS udio, Intl

Having enough juice

One of the hotly debated issues amongst designers is current capabilities of an output stage on a DAC or preamplifier. It’s number 8 on our list of the duties of an output stage.

On the one hand it can be argued that modern interconnects placed between the DAC and the preamplifier are an easy load – after all, they are just a wire. That wire is connecting the relatively low output impedance of our stage with the higher input impedance of the preamplifier or power amp. One shouldn’t have to pay much attention to the drive capabilities in order to do that. Of course cables shouldn’t sound differently either, but …..

What we’ve found seems to be quite the opposite of this conventional wisdom. In fact, PS Audio’s output stages are designed in the same way that a power amplifier’s output stage is – a complimentary pair of transistors with a constant bias running through them and extremely low output resistors – all to be able to properly drive a pair of interconnect cables between two products.

The whole notion of high output drive capability for preamps, phono stages and DACS entered into my view years ago when we were first investigating the audible effects of feedback in a line stage. The effects of feedback on the audio signal has been (and still is) a hotly debated topic amongst designers. I would venture to say the prevailing opinion is that judicious use of feedback is probably quite acceptable and even desirable in a high end product – but back in the late 1970′s not many were being judicious with its use and even fewer of us understood the ramifications of its use. It was a time of experimentation and learning.

Part of the experimentation we were involved in was comparing zero feedback circuits with “normal” feedback circuits of the same gain and topology to hear the differences. We noticed two major changes between the two: zero feedback was much more open sounding yet wimpy in the bass and slam than that of the feedback version. What would be awesome is if we could maintain the openness while gaining back the bass and slam.

One of the side benefits of feedback is lower output impedance which, as we’ve discussed before, equates to better drive capabilities at the output. Lower output impedance means that whatever you attach to the output of your amplifier stage has less impact on that stage’s performance – the lower the impedance, the fewer problems driving a cable. So, for example, if you use a long interconnect cable with all its challenges, the circuit with lots of feedback won’t be bothered by the cable and the opposite would be true for the zero feedback circuit.

In our setup at the time we were running a 15 foot cable between the test amplifier circuits and our power amplifier driving the loudspeakers. So obvious was the zero feedback stage’s loss of bass and slam that it was hard for any of us to imagine we couldn’t actually measure something causing this loss. We placed a scope on the output of our stage and noticed a change in the waveforms when we connected the cable or disconnected the cable. Aha!

To fix this problem we designed a simple power amplifier-like output stage that could deliver plenty of current and had low output impedance on its own without the use of feedback. We attached this to the output of our zero feedback amplifier stage – verified the connection of the interconnect cable had no effect – and gave it another listen. Bingo, bass and slam were back in spades and the openness we were hoping to maintain remained intact.

This was some thirty years ago but the principals we learned back then about drive capabilities, managing feedback to respect openness and musicality while giving us much needed qualities difficult to attain any other way remain to this day.

You have to make sure there’s plenty of juice available, even if the chore seems really easy.

Paul McGowan – PS Audio, Intl

Are you sure?

Lars, one of our newest readers, responded to the January PS Audio Newsletter’s request for suggestions on what you’d like me to write about by asking the following question:

“You are walking down the street perfectly happy after a good lunch or so, when suddenly from nowhere you hear music. It could be a sound two blocks away, poor acoustics, noisy background etc. Maybe it’s a street musician, but the thing is that your brain can detect this as live music as opposed to reproduced within fractions of a second, despite the conditions around you.

We are trying hard to achieve the very same illusion of real music in front of our beloved music system, but never seems to reach the same fidelity even in “perfect” acoustic conditions. So, sitting in sweet spot, not having any background noise, speakers optimized to the room etc. still learns us we have a bit to go I think.”

That’s a great question and one our long time readers will recognize as a consistent theme in Paul’s Posts – because it also amazes me as well. Just the other day my family and I were in a pub enjoying a cold brew and dinner when, from a completely different room, someone started playing a flute. All 5 of us at the table immediately turned in the direction of the music but were unable to see anyone playing as they were in the next room. So convinced were we that it was live, one of my sons volunteered to go investigate just to make sure. Of course, it was live, two rooms away from where we were in a very crowded and noisy environment.

Perhaps you’ve had the same experience.

I would suggest to those out there that believe they do not possess golden ears or are convinced they cannot hear differences in audio equipment to reflect for a moment on the fact that all of us have the ability to recognize live sound vs. recorded and reproduced sound. The only real trick is to try and differentiate within a reproduced environment whether one presentation is more live than another because the ear/brain understands none of what you’re listening to is live.

I believe the basic problem is the loudspeaker itself and less so the electronic chain and microphones used to pickup the music.

This doesn’t answer the equally tantalizing question of why we can also identify a live electric guitar playing through a loudspeaker either, but hey, we have to start somewhere.

Paul McGowan – PS Audio, Intl

Single or double

Number 7 in our list of design challenges for a DAC or a preamplifier output stage is providing a single ended or balanced output.

As in everything we designers do there are multiple ways to achieve this: some simple others more complex and each with different audible results.

First let’s remember why we would want to have these choices available and what they do for us.

A single ended output is the most common output on source and control equipment – it’s what we normally call the RCA out because it uses a RCA connector (also known as a coaxial connector). This output has two wires, a ground and a signal or hot wire that the music travels down. Sounds good but has no ability to reject common mode noises.

A balanced output is less common but certainly available on a number of products. Most PS Audio products support balanced inputs and outputs because there are certain advantages to using them relative to the single ended ones. This output uses an XLR connector and has three wires: two signal wires and one ground. Balanced is the best sounding option and can remove any noises entering into the connecting cable. If you’d like to learn more about the balanced connections, you can refer back to the start on this series to the post Balanced.

From the output side of building a balanced or single ended output it’s relatively easy: simply take the single ended output and run it through another IC op amp to invert the signal (flip the polarity upside down) and this gives you the second signal wire necessary to sending a balanced output. This is how many manufacturers do this – but it’s not a great idea.

Running the output signal through another amplifier stage to flip the phase 180 degrees means that 1/2 the signal going into your power amplifier or preamplifier is different than the original – because it’s going through a second amplifier stage – and since a true balanced input amplifies only the differences and not what’s the same on both signals, the balanced output of this compromised circuit would probably sound worse than just using single ended output.

If the audio designer really cares about sound quality he will design a fully balanced output stage from end to end. This means that there are twice the number of components in the output stage because you basically have two identical output stages – one for each of the two signals needed for a balanced output.

The beauty of this approach is that the signal is identical in the length and sound quality of the chain for both halves of the signal. This is the proper method of building a true balanced output stage and it’s what every PS product with a balanced output has. Unfortunately, we are among the minority when it comes to this topology – after all it is twice the work, twice the parts to get it right.

Whenever the designer chooses to take the easy path in audio design he had better consider all the variables and how they affect the sound before making those decisions.

So the next time you want to consider a new DAC or preamplifier that has a balanced output, ask the company the right question. ”Is this a true fully balanced design or a single ended one with a phase splitter at the end?” You might be surprised.

Paul McGowan – PS Audio, Intl.