Yet another short article that will be expanded later.....really! No, I mean it!!!

I had an epiphany today. Many of the technical problems we talk about in audio amps, including tube amps, are problems that only exist due to cost to manufacturers. The pressure to make an ever-increasing profit...for corporations (we really need to do something about corporations - they have all the rights of individuals, and virtually none of the accountability or liability each of us have)...or the intense pressure to just stay alive, for the privately held amp-maker garage-setup company...creates devil's choices of how to get good sound without spending so much on components and fabrication that profit goes right out the window.

These problems do not need to be an issue for DIY'ers. Yeah, we have cost issues, all right, and a favorite topic on tube forums is the "junk box" amp...it's FUN to take a bunch of junk that's piled up over the years and make a half-decent amp out of it.

But we don't pretend...I hope not...that the junk-box amp is GREAT. It's just reasonably good, and cheap, and a great way to make old stuff useful. One can always use another tube amp. Laundry room, garage, outhouse....

As DIY'ers, we have the wonderful opportunities to burn-up (or cut in half) "straw-men" all over the place. In this article, as I accumulate them, I'll be listing them here. Submissions are welcome; I like to attribute other authors, so you're welcome to submit, I won't steal your work, I'll publically thank you.

The first straw-man, which I've made a BIG DEAL about in other parts of this blog, is the matter of "blocking". (See the article "What Transformation Audio is All About", considerably far down the page; I don't want to copy it all here, because it would just slow you down here if you've already read it...later I'll put a link here about "blocking" if you don't know what it is). And I talk about direct-driving the output stage, with direct coupling, using a cathode-follower stage or mosfet-follower stage or TubeLab's PowerDriver circuit (see links on right)...but looking at Max's superb EL-34 power amp at AngelFire (see link near top on right)...I had this epiphany.

"Wait a sec....." I sez to myself, I sez, "Max drives the output tubes with cathode-followers (good)...but the cathode followers are capacitor coupled to the output EL-34's....why not direct-couple?....wouldn't blocking be a potential problem?"...and then it hit me: "Ahah!!!!!!!!! Blocking is only a problem in amps where you're trying to save cost by beating the crap out of a single output tube, instead of paralleling enough of 'em to avoid the problem altogether!!"

"Blocking" is only a problem if you're pushing the output stage tubes to the limit. If you parallel enough tubes (or enough mosfets, heh heh heh), you have so much reserve power - assuming you're running the amp at sane volume levels, of course - that the output stage never gets whacked to the ceiling or slammed to the floor. This means you can capacitor couple to the output stage without worrying about "blocking".

A beautiful example of Cutting the Gordion Knot! (see "Wonderful Links for LEARNING about tubes, electronics, mosfets & transistors").

Max wanted extremely low distortion in the output stage BEFORE negative feedback, so he wouldn't need to use much negative feedback (with his design, you could do without negative feedback altogether, if you want). How to do this? The same way you get low distortion with ANY tube: run the tube with only a small portion of it's potential signal-swing capability used. I'll put a diagram in here later to demonstrate what this means, you'll understand instantly when you see it. (I have to figure out how to do that in code; the blog-ware won't let me do it directly, but no big deal, just time).

So, Max used an ELEGANT solution (not a simple...as in "simple-minded"...solution...different than elegant): he used multiple output tubes in parallel, so he could get good power WITHOUT pushing the tubes near their limits. (He essentially made the output stage into more of a CURRENT device than a VOLTAGE device; by paralleling the tubes, he gets lower plate resistance, higher current, and a more efficient impedance match to the output transformer...I'll post an explanation of this, with math, later). Aaaaand how he doesn't need to push one tube to it's max....the quartet of EL-34's kinda loaf along, and transients, if they clip, are clipped softly, and don't cause "blocking". See? How cool, huh?

So, we now kinda have two legends/myths transmogrified (Calvin and Hobbes) into one: Cutting the Gordion Knot, and Straw Men: Cutting Straw-Men in Half with a Mighty Bronze Sword!! Cooooooooool.

More Straw-Men to come heh heh heh heh

Best, Charlie

The web, in my opinion....IF we can keep it FREE...free from censorship, free access for all......is the next step in the evolution of mankind.

It enables people to learn just about anything, and this is especially true of electronics. There are so many wonderful teachers of electronics on the web, who just want to share. Maybe they make a nickel or two from site visitors, and that's a good thing...I know for sure they're not getting rich, it really is a labor of love. So, I've chosen sites where you can learn everything you need to know to design and build good tube and mosfet and transistor amps just like the pro's. The essentials of a good amp were well known 70 years ago; the only thing that's changed is we have a lot more options now: Constant Current Devices (CCS's), far better transformers, mosfets, op amps, transistors, and so on.

All you really need for a great amp is good transformers, some decent tubes, and a good, solid topology. You do NOT need silver wire, gold plated switches, special tubes cryogenically treated, and super expensive capacitors. YOU DO NOT!!! (Although you DO want to avoid paper capacitors). The most important thing about an amp is THE TOPOLOGY (type of circuit). The second most important thing about an amp are the TRANSFORMERS (except for mosfet output amps, of course, heh heh heh). The third most important thing about an amp are APPROPRIATE TUBES FOR THE TOPOLOGY...AND THEY DON'T HAVE TO BE EXPENSIVE.

Do not believe the hype. And, if you pay attention to the tutorials, you will come to understand that biasing a cheap tube properly, driving it properly, and loading (or, really, unloading) the plate properly, and in some cases protecting it from vibrations that might cause microphonics, will give you far better sound than the most expensive tube in the world not set up correctly.

"Tube rollers" would have you believe that certain brands of the same tube are better than others, etc., and while there's SOME truth to that, don't get your knickers all twisted up about it....a cheap tube properly set up can still sound better than a ridiculously expensive NOS Mullard with black plates, etc., not set up just right. And then, when you add in the fakes on the market, expensive collector's items that are really junk from China relabeled, you start to understand that there's a lot of snobbery involved in tubes that you do NOT have to buy into.

After that, it gets subtle. Types of resistors, etc. Avoid paper caps, etc. So, after the educational links, I'm going to include great examples of circuits, circuits you could just go ahead and build, circuits you can learn from by applying your understanding to them. Every great amp circuit...and there are hundreds of them...teaches you something about electronics. Sometimes it's subtle, sometimes it's one of those things that hits you between the eyes and you go: "Doh!" - but either way, you've learned something.

I will also list here books - the best of the best - you can purchase to deepen your understanding, and which include projects you can build, also.

One word about projects: there are four main categories (these are kind of like tornado categories...be wary of a category 4!!):

Category 1: Decent amp projects that don't cost too much and aren't huge and heavy but which will give you a great deal of satisfaction and quite good sound;

Category 2: Really good amp projects that incorporate new ideas and have, I think, some conceptual breakthroughs in them...some complicated, some amazingly simple; some cheap, some expensive;

Category 3: Great amp projects that cost a lot and/or require much greater technical and electronic skills (NOT for beginners....REALLY!!);

and Category 4: Great amp projects that have a WOW factor but are really, for most people, simply impractical - way too big, way too high voltage, way too complicated. I'll list them so you can look at them and go WOW, and drool a little, but please don't build them - they really are impractical, they really are. Build some Category 1's before you even think about building a Category 2 or 3...and tell your spouse to STOP YOU if you start talking for real about building a Category 4...but they're still fun to dream about.....

SO, LET'S GET STARTED:

LINKS FOR LEARNING ABOUT ELECTRONICS AND TUBES, IN PRIORITY ORDER!!! DO #1 FIRST...REALLY!!!!

1. START HERE:
These links are at Max's AngelFire site...don't underestimate these pages, they give you foundational understanding that you must have. Once you've mastered these pages, then you can go on...but not before; if you jump ahead, you'll just cheat yourself.

To start at DC basic electronics, go here (this means you, unless you're already an EE (Electrical Engineer):

http://www.angelfire.com/electronic/funwithtubes/Basics_01_DC_Circs.html

(or you can go to the home page and see all the tutorials in one list):

http://www.angelfire.com/electronic/funwithtubes/index.html

either way, do them all, from start to finish. IF you already are an EE, but only got trained in transistors and need to get brushed up on tubes, you could start here:

http://www.angelfire.com/electronic/funwithtubes/Basics_03_Diodes.html

and then continue on to here, etc.:

http://www.angelfire.com/electronic/funwithtubes/Basics_04_Triodes.html


All done with AngelFire's tutorials? You DID read them all, right? You DIDN'T skip any, right? You at least tried to do some of the problems yourself first? Good for you! Now you have the basics you need. Let's move ahead to articles that give you a more global understanding of audio amplification...later, we'll come back to highly specific (and full of math, but math that makes sense) articles that focus in on particularly important topics.
2. NOW GO HERE:

http://sound.westhost.com/amp-sound.htm

This is Rod Elliot's site, one of the world's best. READ ALL THE PAPERS. The link above is where you start, but you can just go to the main page and click on the articles link to see the index of all of them, if you wish:

http://sound.westhost.com/index2.html

Don't worry that many of the papers are about SS amps. You may find that you'd just as soon go SS, and use Rod Elliot's PCB's or Kits. That's cool. Tubes are fun, they glow, they have a special sound, but they are also a downright hassle compared to SS. You might find it's easier to make your first amps SS using products such as Elliot's. You certainly couldn't go wrong.

I remember one of the finest, clearest, most incredibly realistic sound I ever heard was Art Garfunkle's first solo album (remember that? it had it's appeal...) through a Heathkit SS amp through Acoustic Research speakers...until I turned up the volume a hair too high, and the impedance dips in the AR speakers crossovers completely FRIED the ouput transistors of the Heathkit amp. What a bummer!

I doubt you'd have this problem with any designs by Max at AngelFire or Elliot at Rod Elliot's, design has come a long way in terms of protecting the amp from nasty impedance dips from speakers since the 1970's.

Have you read all of Rod Elliot's articles? You are way down the road now to being an expert DIY'er, and you haven't built anything yet! You COULD at this point choose a project from either of the two sites recommended above, just to get your feet wet, get something built, and feel the satisfaction. Don't worry if it's a small amp...later on, you'll like having a tiny amp in your bathroom, kitchen or by your computer...and then can move on to building a much bigger amp for your dedicated listening space.

You may notice some various small points of disagreement, or different philosophies, between Max's site and Elliot's site. Welcome to the world of being an audiophile (not audiophool)!

By the time you're done with the full set of links I'm going to put here, you will find some very BIG disagreements...which is cool, because, A: Build one of each, and then compare! Decide for yourself! B: As you learn more, you'll see that some of the disagreements are simply about the advantages and disadvantages of different topologies and components, and you'll find you can make distinctions from your own understanding and experience and feel great about going down a particular path. For example, the day I decided to give up on the idea of winding my own Output Transformers, a huge weight lifted off me. I had cut a "Gordion Knot"...and now I know precisely where I personally want to go with design: replacing the OPT with mosfet circuits. Now I'm happy. I can't afford good output iron. I can't afford the TIME to become a master DIY winder of my own OPT's. But I CAN figure out how to use mosfets instead of OPT's well. Now, if you want to become a master DIY winder of OPT's, go for it!!!! You'll be one of the rarest of audiophile DIY'ers. Later in this set of tutorial links I'll give you links to sites where you can learn how to wind the best of the best.

The Legend of the Gordion Knot is very important to understand, especially if you want to be an audiophile DIY'er. Otherwise, you might just go nuts..........I'll explain the Legend of the Gordion Knot, and what it means for helping each DIY'er come to peace with what they want to achieve, tomorrow...

Here's the legend. Now, note that it's not a myth....historians are pretty sure this really happened, or a reasonable version of it...so it's a legend, not a myth. A myth is  a story so old that the origins of the story are lost in time, and often it's so old the origins predate the written word. A myth also contains symbolic language or concepts that are sacred, or philosophical, or describe creation. A myth is neither false nor true...it's value is it's symbolic or sacred meaningfulness.

A legend is generally a true story that is heroic or awe-inspiring in some way. My favorite legend is the Legend of the Gordion Knot.

We go back in time to Alexander the Great, 333 BC. Wow, that was a long time ago. He set out, after being taught the history and philosophy of the Greeks by Aristotle, to conquer the known (and unkown) world. Why? For the same reason most conquerors give: to bring peace, prosperity and order. Of course, a lot of innocent people die along the way, so I'm not so thrilled about that part; that isn't what makes Alexander great to me.

But Alexander did something legendary that was of mythic proportion - meaning that it has a lesson for all of us. Now, Alexander, remember, was tutored by Aristotle. He was familiar with logic, with math, with solving problems, and he also knew about unsolvable problems. For example, Alexander had surely been taught by Aristotle that it was impossible, using the tools of geometry at the time - the compass, string, the right angle, and so on - to square a circle. That is, take a circle, and then find the square of it that would have precisely the same area. Aristotle could not have known that over 2,000 years later, that problem would remain unsolvable, by the most powerful computers and mathematicians available to mankind; he may have believed it really would be solvable some day. But I digress!

When Alexander was beginning his campaign to conquer all of Asia, to bring the wisdom of the Greeks to all the "barbarians" of the Persian Empire and the unknown-but-hinted-at lands beyond, he, with his army, came to the town of Phrygia in an outer province of Persia. Here was kept the Gordion Knot, the knot which could not be untied. It is theorized that this was a knot which had the two ends spliced together, so there were no free ends, and moreover which had been tied wet and then dried in the sun so that it shrunk, and was so tight that not only could you not manipulate any part of it, but that even if you did find a free end - which there was none - no part of it could be loosened enough to allow any untieing to take place.

In any event, Alexander was no dummy. He is said to have played with the knot for a brief time, until he had satisfied himself in his own mind that there was no point in attempting to untie it. Then came the moment of suspense. What would he do? Now, the seers of the time in that town - the priests of whatever religion the Persians were operating under, I'm sure whatever it was it was akin to Egyptian religion - strange, complex, allegorical, and ultimately unfathomable - had challenged Alexander to untie the knot, because the myth was that whoever could untie it would go on to conquer all of Asia. Since that was precisely what Alexander had in mind, and everyone knew it, the stakes were high. He HAD to untie that knot. He knew it. His soldiers knew it. The priests, laughing up their sleeves, knew it. But Alexander knew quite well that it was impossible to untie it as it was; it was a trap to defeat him before he got started.

So Alexander did the unexpected, and in doing so, performed one of the most brilliant strokes of genius in the history of humankind.

He drew his sword and cut the blasted thing in half. Voila! Lots of free ends, nicely loosened up, and now easy to unravel.

The allegory is this: we humans present ourselves with unsolvable problems all the time. We make mental barriers of all sorts in life, finding ourselves stuck in jobs we seemingly can't get out of, perhaps stuck in a relationship, perhaps a problem with a child that seems unsolvable. But the problem is not the job, not the relationship, not the child. The problem is our own belief that we cannot find a way to cut the knot in half!

I don't mean to preach. I do not underestimate the magnitude of the challenges we all have to face. But there is always a way to cut the knot in half, unravel it, and begin anew. Always.

I will give you one example I observed recently, and then move on, back to electronics and tubes. Just to illustrate what I mean. A co-worker at work has been having enormous difficulties with their teen-age daughter. And yet, as I've observed the drama of ever more-intense phone calls over the past several months, I was struck by how confrontational this person was. This person was constantly causing ruckuses with all sorts of co-workers. Not with me, because I can get along with just about anybody. I would draw the line at a Stalin or Hitler or Mussolini or Weird Al Yankovic (KIDDING), but basically I like people and do my best to treat everyone with equal respect and kindness.

Which is why I have no tolerance whatsoever for forums that do not delete insults. Insults are not acceptable to start with, and leaving them in a thread permanently is absolutely not acceptable. But, back to my story.

I then saw how this person was constantly causing confrontations with this person's daughter. Confrontations that did not need to happen, that treated the teen-ager as a child, that punished the teen-ager unnecessarily. No wonder the teen-ager was acting up! Gently, I talked to my co-worker, and suggested that this co-worker might consider getting help for themself, rather than the daughter, so they could then be a better parent. (Boy, is it hard to write in English without including gender!)

Luckily, this person did not take offense, and I was considerate and diplomatic. But this person also paid no attention to my advice. This person could not see that they were the problem. As long as this parent was part of the problem, they could not cut the knot in half, unravel it, throw the blasted thing away, and heal their relationship with their daughter. So, the drama continues, even today as I write this. So sad and unnecessary.

Alexander got this, you see. He understood that HE was the problem. HE could not untie that knot. But, he COULD cut it in half - so that's what he did. And that's what we're going to do, on this site.

Now, to electronics. There are lots of Gordion Knots in electronics, especially tube electronics. One of these, for example, is "blocking", where one stage is coupled by a capacitor to the next. A loud transient or crescendo can cause a capacitor-coupled stage to "block" (I go into the technical details of this with links later). Now, a lot of engineers have figured out ways to minimize blocking by tweaking the circuit, but keeping the capacitor. Why? Because commercial amps are absolutely driven by cost - and it is a heckuvalot cheaper to capacitor-couple stages to each other!!

But what would be the Alexandrian Solution to blocking? (Alexander's troops referred to his cutting of the Gordion Knot as the "Alexandrian Solution" - they thought it was a HOOT!!) Direct coupling. Get rid of the blasted capacitor!!

Is it necessary for all stages? No, usually just the coupling to the output stage. But some phono preamps are designed to be direct-coupled, and the designers claim that the sound is "faster", "clearer", "cleaner", more "exciting". And they're right. I've built Jeremy Epstein's direct coupled 6C45P (one-stage RIAA) phono preamp, and it is WONDERFUL. I plan to replace the phono stage in the Heathkit AA-151's I mod from now on with his circuit. I'll be posting my way of making his design later, when I start posting projcts. All in good time! But you don't have to wait for me...just go to his link. His direct coupled phono preamp is in the links on the right, just look for the one that says "Jeremy Epstein's 6C45P...etc."

As for getting rid of the coupling capacitor to the output stage, the output tubes? Simple as pie. Use cathode followers, or, better, mosfet followers. Or, even better yet, George's Power Driver mosfet circuit (see the link on the right to George's TubeLab site). Later on, below, I'll give you links to articles and circuits that show how to do this. Is it extra work and expense? Yes, but not much, and which would you rather? Spend an extra couple of hours direct coupling to the output stage, or listen to "blocking" for years and years until you build it over and do it right? If you look inside a McIntosh Amp, some of the best amps ever made, guess what? You'll find the output tubes driven by cathode followers, eliminating blocking. Right there is one of the reasons Mac's were...and are...so good. And you can do the same, piece of cake. Or pie, whichever you like. Or pi, for that matter, since this all started with squaring the circle, which includes the area of a circle, which is .... remember? pi * (radius(squared)). Just kidding.

What are some other Gordion Knots we can cut? Well, as I've already said, we can get rid of the Output Transformer by using mosfets. Or (shudder) transistors. Or even op amps, or chip amps!! For me, rather than trying to become an expert Output Transformer winder, because I really cannot afford really good output iron, and I really don't want to use global Negative Feedback to make up for the shortcomings of lousy output transformers, the "Alexandrian Solution" is just to get rid of the blasted things, and use mosfets. Which is precisely what Pete Millet did with his famous "Starving Student Headphone Amp" (look at the top link on the right, you'll see Pete Millet's site. BUILD THAT HEADPHONE AMP! YOU'LL BE GLAD YOU DID!!)

What's another Gordion Knot? There's lots of them. How about hum from filaments? Well, there are two ways to cut this knot. One "Alexandrian Solution" is to raise the bias on the filament 70 volts positive to the cathode (or, in some cases, 30 volts; see below). That stops the flow of electrons from the filament to the cathode, and very effectively prevents hum from going from the filament (on the electrons, or, rather, with them) to the cathode of the tube.

Whether you choose to raise the filament 30 volts or 70 volts above the cathode depends on several things: how much signal swing the control grid is going to be seeing, and what the filament to cathode voltage rating is. This applies to indirectly heated tubes only, not the old direct heated tubes (DHT's)...(actually, I think when people use the abbreviation they usually mean Direct Heated Triode's, which were the original audio tubes, but there were also direct heated pentodes, as well as direct heated hexodes and so on, so you have to take the context into account to know which meaning is meant if the write doesn't specify)...(DHT's are wonderful, by the way, but biasing the filament respective to the cathode doesn't apply to them).

If the tube is only seeing an input signal of a few volts, or less, such as the input tube for a phono preamp or an intermediate amplification stage in an amplifier, then you can probably bias the filament 70 volts positive to the cathode, so long as the filament to cathode voltage rating is at least +/- 100 volts. If the input signal voltage (peak to peak) is likely to exceed 30 volts, for a tube with a filament to cathode rating of +/- 100 volts, then you're better off biasing the filament 30 volts positive to the cathode, so you don't run the risk of the cathode swing exceeding the filament-to-cathode voltage rating. This is likely to be true for the output tubes, and possibly the driver stage of the output tubes. Now, some tubes have filament to cathode ratings of +/- 200 volts, which can usually be run at 70 volts filament positive to the cathode.

Why 30 volts, historically? Because at one time RCA had almost a monopoly on tubes, and certainly had a monopoly on making the tubes for companies that were using RCA's copywrited circuits...and RCA didn't want to have additional manufacturing costs associated with a filament to cathode voltage rating of greater than +/- 100 volts. RCA didn't make a distinction between input tubes, where the signal peak to peak voltage swing is tiny, and output tubes, where it can be as high as several hundred volts.

However, Tomer, in his classic book on tubes (title later), found that putting the filament 70 volts positive to the cathode has two advantages: it reduces hum (and noise from the filament) more than 30 volts difference will, and it gives the longest tube life - which is very important, in my book. Actually, I think it was more like 68 volts, I'll have to find my Tomer and look it up, and then I'll revise this to give you the precise voltage. If you're going to go to the bother of biasing the filament more than 30 volts, you might as well choose the precise value. I've been in the habit of just using 70 volts, but now that I write this and remember the value was slightly lower, I'll revise my own value for this! Doh!!! (The old saying, if you want to know something really well, teach it, is TRUE, as I just reaffirmed for myself!)

Another solution, which might feel like tying a new knot (!), is to supply the filaments with DC, rather than AC. If you stay with AC, which some audiophiles say make tubes sound better (I don't believe it, and I personally always DC my filaments...all I hear is a black background with incredible detail, so why I'd go to AC I cannot fathom...), but, anyway, if you stay with AC,  you have to route the filament wires carefully around the edges of the chassis, keeping them away from any signal wires, especially low-signal wires. Personally, I do both....DC my filaments, and raise them 70 volts (or 30 volts, depending) above the cathode. And route carefully, but I don't have to obsess about it. And I don't have hum in my amps, period. Later, you'll find articles on these topics, along with nice, simple circuits that get the job done just fine.

It's good to keep in mind that while it seemed a simple thing for Alexander to cleave that knot in two, it actually required a really good sword, a strong arm, a mighty swing, a lot of nerve, and good aim. The metallurgy and skilled craftsmanship that went into the making of Alexander's bronze sword was by no means trivial. So, an "Alexandrian Solution" is not necessarily simple....it is, however, elegant...the simplest solution that will WORK. And when I say "work", I mean "work to perfection".

Providing DC to filaments and raising the filaments 70 volts or 30 volts above the cathodes "works to perfection". Giving the filaments AC and having to route the wiring with great care does not. Some amp builders totally disagree with me about this. Fine. You're free to disagree too, it's fine with me. But put your ear up close to a speaker after I've modded a Heathkit AA-151, and tell me what you hear, with the volume all the way up and no signal? Nothing. Put your ear up to the speaker of an amp with AC to the filaments, and what will you hear? Hum. It may be slight, but it's there. You can argue that the music would be so loud that the hum couldn't possibly be a bother...and you're probably right...but you know what? It's still there. The knot remains tied.

MORE TO COME! BACK TO LINKS FOR LEARNING!!

This is a jump ahead, but I want to capture this link here while I have it:

http://www.youtube.com/watch?v=VC-sxvNzC8I

What more could you ask for? A lecture on how to design speaker systems to fit your listening room?? From one of the world's true EXPERTS?? WOWOW

The rest of the series you'll see on the right in a box at You Tube, but start with the link above, so you have foundational understanding for the rest of the series.

Cheers!!

Best, Charlie

(MORE TO COME...11-12-8...WILL ADD MORE SOON!)