VENUE: Haberfield Community Centre
78-80 Dalhousie Road, Haberfield
DATE : Sunday, 27th January 2008
TIME : 2.45pm to 6.00pm |
Rasmussen reveals …
It’s another year and so it will be my privilege to put on
the show for this month. The list of equipment will be as follows:
Integrated Triode Amplifier: The two previous articles describe the way this amplifier has been updated. Make sure you read them. This one is rated at 18 Watts per channel Triode Mode. It was sourced from Malaysia and is called Audio Frontier.
Elsinore Mark 3 Loudspeakers: Previously Mk1 and Mk2 versions have been demoed, but this time it will be in the final Mk3 guise as this is expected to be the final version (famous last words). They also look somewhat better than before, veneered and not a slap of black coat of paint. Even low powered amplifiers can drive them to loud levels or large rooms.
Harman Kardon HD970 CD Player: Upgraded with new Super-Low Power Supply for the clock, tuned to 1/1000th of a Hertz, rock steady clock signal, differential current converter in the output stage and with minimal filtering. Several power supply enhancements that lower noise floor and the output stage is SuperReg’d. This player has better sense of air than any other CD Player I know.
JLTi Phono Stage: This is the Oz version – the Swiss made version is sold in the U.S. for around USD $1600. Roger S. Gordon reviewed it for Positive Feedback – he bought it – and has declared that no phono stage under $6,000 betters it. Some of you may know that since Richard bought one, he is now playing far more LPs than he used to. They are available at a reduced price to ASoN members.
Townshend Elite Rock Turntable: Fitted with the rare Townshend Excalibur Tonearm. The turntable was recently upgraded and fitted with the Origin Live DC200 motor – better than their cheaper DC100 motor. A new DC power supply was designed based on the same power supply technology used in our clocks for digital players.
The Vacuum State Moving Coil Cartridge: The best cartridge made by Benz micro, ever! So far only ten made. Vacuum State went to Albert Lukashek of Benz Micro. He was asked what was the best cartridge he could make with what technology was currently available to him. He pointed to the top-of-the-line Benz Micro Ebony LP which sells for USD $4700. He said he would nude mount the cartridge, no ebony wood, and change the aluminium mounting plate with something better. It was decided to use a certain type of brass. Minimum order was ten. One is now fitted to my Rock turntable and I will bring it along. Please, no unauthorised operation! The price is the US will be $2500 and here in Australia A$3000. The Californian importer, Jay Kaufman from Audio Revelation: “This is wildly good and out of the box I think would compete with anything I've heard so far.”
Those are the highlights; we will add cables etc. We should have a good time. I will start setting up at 12 Noon. We will also try an experiment by changing the load value on the input of the JLTi Phono Stage and see what the audience think. Will they hear any changes and what preference might they have?
Finally, I think it should be mentioned and it isn’t just about me: In the future we hope others will volunteer and bring something to the meeting, large or small. Hopefully that might be you. We should conduct all meetings in such a way such as not to discourage participation – and also keep in mind that a meeting will always have some shortcomings and it is not always possible to please all. Our comments and any critique during meetings should reflect this.
I can’t wait…
Joe Rasmussen
At The Cross Roads
joeras@vacuumstate.com
BARGAIN TUBE AMPLIFIERS THAT CAN COMPETE WITH ANYTHING!
PART ONE
You cannot beat the Chinese at their own game. Look at the following picture and check the price of what a tube integrated amplifier can be bought for. I spoke to Elson and we both agreed we couldn’t even buy the parts for that price.
So if you cannot beat them, do you join them? In effect, that is what a lot are doing. Japanese brands are now mostly made in Asia because it’s cheaper. Increasingly anything remotely electronic is: Made in China.
The above picture shows an amp that was for sale on eBay via a local importer. The minimum asking price was $749 and it was passed in.
I think Peter Crisp and a friend did show the way at an ASoN meeting a while ago. If memory serves me right, they had bought two of these amplifiers directly from China. I have also had some of these samples shown to me. As Peter also realised, they had a number of problems, even though the basic bits were good. Also recall Peter saying that certain capacitors in the power supply were not rated high enough in voltage and that along other problems, if not fixed, would mean the amp would likely fail in shortish time.
I would expect the one that could have been had for $749 from importer here in Sydney means that they would have made sure that obvious flaws were fixed.
For another importer, this one in Brisbane, take a look at www.coemaudio.com.au especially the EL34 amp: www.coemaudio.com.au/products.php?product_id=119
This Coem MC-34A EL34 amp sells for $1200. It has Input Selector and Volume Control, so your CD Player connects directly in. Add some decent cables and speakers; you have the heart of a good system. But this amp is potentially superb.
The Chinese do not know tube circuitry; they copy what is freely available. These are classic Williamson type circuits or Dynas from the 1950s. You can poach these from many places on the ‘Net and that is what they use. Manufacturers of tube amps with proprietary circuits are very wary of the Chinese who have no scruples copying what does not belong to them. One manufacturer deliberately gets the Chinese to make a product, a tube amp for him, with a deliberate mistake. The Chinese thinks he is mad and compensates by substituting a couple of resistor values. But the last laugh is on them as the fault is easily corrected after delivery before being on sold. The real owner is now comfortable that his circuit isn’t going to be copied correctly.
But back to these EL34 amps, though basic their circuits are, the sheer potential of these amps needs to be tapped into. Let us take these amplifiers, use the basic circuit and then add elements that will take it from the 1950’s to the 21st Millennium.
- Current Source Technology
- New method of connecting outputs into Triode Mode
- Output Transformer Linear Enhancement Technology
- Decent Volume Control
This and some other more basic things can take these amps to another level. Let me show the basic circuit likely to be used by such an amplifier and then later show the changes:
We see three basic stages. The first Triode near the input is the Voltage Stage. Problem is that it often has too much gain – especially if we wish to eschew feedback. We will use a technique that allows us to adjust the gain without losing the operating point of the tube. The next two Triodes form a Driver Stage that also doubles up as Phase Splitter, as we need to feed the two output tubes out of phase. But the phase splitting is not very tight, the AC Balance poor. Finally we have the Output Stage. What is shown is an Ultra-Linear based on a pair of EL34 Pentodes. I hate U-L and I know Elson does too.
My strong suggestion is to source these amps with EL34 Pentodes fitted. There is something that is not generally known about the EL34 and also, when taken full advantage off, the EL34 according to legendary designer Richard (Dick) Sequerra, is the best Power Tube ever made. The results we are getting lately finally are bearing this out. Sequerra’s amps are USD $70,000 – so I have been told – and at that price he could have used any tube he wished; yet he chose the EL34. But you need to know how to get the best out of it and it is clear he does.
And so do we!
At the first ASoN meeting in January we will present and listen to one of these potential wonders. The following shows the above schematic changed to include a number of new features.
The basic circuit is much the same. What is highlighted in Red, these will be the changes implemented. The L.E.M. is an ultra-sonic module that excites the Output Transformer and enhances its resolution. The Current Sources are those funny Red Circles, these control gain and also improves the AC Balance of the Phase Splitting as well as enhancing power supply rejection and hence better stability. The Output Stage is now in Triode Mode but in a way you have never heard before. We can include technology that people like Sequerra are putting on multi-ten-kilo-buck amplifiers. The last little tweak is using my favourite PPS By-Pass caps.
Speaking to Richard, our current Presidenté, we will also arrange to have another amp of known quantity, to be used for comparison.
It is estimated the upgrade to cost around $600 – making this more than just outstanding value. This is an opportunity to get the sound of your dreams. It will not be a super powerful amp, but being Triode EL34, something around 20 Watts should be possible, and depending on the original design of output transformer and power supply, a little more than that is possible. But Triodes clips ever so much more softly than usual, so expect 20 Watts to sound more like 50-60 Watts. Not all ‘Watts’ are equal. But speakers with good sensitivity will have a definite advantage. Those into Full-range speakers such as Lowther etc, this will be perfect. I will use the Elsinore Loudspeakers in their latest Mk 3 guise. They supercharge most power amplifiers.
In Part 2 below, I discuss in more detail the use of L.E.M. and why it works as it does. It is based on a patent during World War II and then consequently forgotten – but now resurrected with the help of Menno Vanderveer, the famous Dutch transformer designer (designs for Plitron) whom I met in Europe last year. We continued to stay in touch and the L.E.M. is the result. I also discuss why Triode Mode can be dramatically improved simply by controlling the impedance seen by the electrodes in Pentode tubes like EL34. But we are not just limited to EL34 and we can use more powerful tubes like KT88 and 6550 etc. But the EL34 is magic.
PART TWO
The previous Part One was to have been published in November. But alas, due to personal circumstances of our esteemed editor, that did not happen. But a number of you would have received it in a separate email attachment from me directly.
So here we go straight into the new material, Part Two.
I must say that I am particular excited about recent developments in valve/tube technology, not the least because having a part in it. The conversion that I briefly described in Part One has now been done to two pairs of mono-block amp plus two integrated amps, one of which I am bringing to this month’s meeting on Sunday 25th. This also means that I have had quite a bit of feedback from those who have heard those amps. The basic conclusion is just how dynamic they sound, rich and powerful and a treble performance that is transformed. Valve amps are not noted for sounding bright – but this conversion makes the treble of most valve amps sound positively untidy and sibilant in voices, especially closed mic’ed female voice. There is at the same time more natural detail and air, but is done in a very under-stated way. It also helps focussing on the all-important midrange, which has great presence, vocals sound real and emotive. Finally the bass is full of power and devoid of smearing. Interestingly, with good speakers, the bass centre image has as much physical presence as the midrange, but you must seat in the middle. Away from the sweet spot it can sound a bit full – but this is not where it should be judged. Overall the sound is more like a superb single-ended tube amp.
(Before I go on, I would like to thank Bill Perkins for his invaluable help.)
Let us describe in greater detail what the new Triode mode does. Here is Pentode (drawn without the Suppressor Grid for simplicity):
The above electrodes are colour coded in this way:
Red is the Anode (Output)
Blue is the Screen (which causes the problems)
Green is the Grid (Input)
Black is the Cathode (the internal reference)
It is the Blue Screen that is the cause of our problem, what to do with our Screen electrode? If we could go inside the tube and rip it out, then we would have accomplished our purpose. But that is not an option. What we must do is hide the Screen electrode. But how?
Here is the usual way of connecting a Pentode as Triode. Connect the Blue Screen to the Red Anode. The resistor value should 100 Ohm at least. But that resistor is bad for sound. Why not connect the Screen directly to the Anode and eschew the resistor and that would make the Screen part of the Anode. Oh, why didn’t I think of that before? Problem solved.
Actually, it is not such a good idea. The tube will have a very short life span. The Screen is much more fragile than the Anode and while this idea would work, the tube might only last a minute, an hour or a day? Who knows? I am not going to find out. But once the Screen goes up in internal smoke, the tube is heading for the bin.
We have to find a better way, one that will protect the tube but still take that stupid Screen out of the way. The solution is not to make the Screen part of the Anode, but rather shadow the Anode. Do a Houdini – slight of hand – an illusion that actually works.
Let us use our colour coding to demonstrate our solution and present what AC and DC voltages will then present itself within a tube.
The Black line is the Cathode. This is always the internal reference of the tube. It is bedrock. Think of it as being zero volts, if you like. The input via the Green Grid is below the Cathode reference. So the input is negative with reference to the Cathode. Well above the both is the Red Anode, which is positive (relative to the Cathode) by many hundred volts, typically over 400V. What we need to do is put the Blue Screen below the Anode, let it see the same AC voltage but below it in DC terms. This is our solution.
Since the Grid is the input to the tube and the Anode is the output, note that the Green line is a) smaller in amplitude and b) the Anode output is in the opposite phase. Triode stages like this are said to be inverting. Because we want the Screen to shadow the Anode, the Screen also needs to be inverted. It must track the Anode but below the Anode in DC terms. The DC component between the Anode and Screen must always be the same, no matter what AC – music signal – it processes!.
If we go back to our usual example to achieve Triode mode using a resistor (bad for sound), the resistor cannot comply with our constant DC requirement. This means that effectively the input Grid now has to compete against another “grid” as the resistor causes the Screen to be current modulated. Effectively our Screen has been become a “Renegade Input Grid” and that means our real input Grid is not totally in charge of what our Triode is doing.
The Screen as a “Renegade Grid” will also act partially as an Anode and thus the current error (which has a DC equivalent) across resistors will be inverted.
So what happens if we have one grid moving in one direction and the other undesirable grid moving opposite, in a relative sense to both the real grid and the Cathode? It does not take a brainiac to work out this is a mechanism for dynamic compression. It’s like a brake! And it sounds that way, but you only realise after the brakes are taken off.
So our method is to tie the Screen to the Anode so they see the same AC signal, but space them apart in DC terms. This DC spacing will prevent the potential damage to the Screen and save the tube. The other requirement is that the AC impedance between the Anode and Screen must also be very low. This will finally mean that our input Grid is totally in charge as is normal with real Triodes, but never with Pentodes. The Grid does not have an undesirable Screen acting as a rival.
There are many other complexities at work here. We have kept things as simple as possible and therefore have not included the output transformer effects. All I will say that these are also minimised. For example, there is an invisible high pass filter at work between the Anode and the Primary winding of the output transformer (formed by the R of the Anode and the L of the Primary). This filter must be kept as stable as possible, especially in Push-Pull amplifiers. Our solution also stabilises the Anode impedance that is part of this filter and thus stabilises the filter dynamically. Otherwise it will constantly be modulated, shifting in frequency. Again, audibly better dynamics.
It is no wonder that listeners have been impressed by the extraordinary life that is imparted to the music. This is what we hear in live music but we go home and still miss. There is a physicality to live music. I am now also convinced that only tubes, when done just right, can do this. Forget solid-state transistor amps. There is another point to make, and some will take exception; single-ended power amplifiers are unable to achieve this. Even if they are Triodes, the lack of symmetrical rise times gives Push-Pull an enormous advantage. This leads on into the next topic, the job of the Output Transformer:
Permeability – The Ability to do the Extraordinary
Permeability is related to the exotic transformer characteristic called Flux Density. We will try not to get into lot of theory here (it’s going to be hard), but we need to explain the fundamental idea of Permeability. An output transformer has basically three parts. A winding that receives the signal, called the Primary. The output of the transformer consists of another winding, the Secondary, from which we derive the output signal. The third part is the magnetic core. Without this core, no significant signal will appear on the output Secondary. Signal in, nothing out!
Here the single-ended Triode has an inherent advantage over Push-Pull, but one it is about to loose. Single-Ended transformers are air gapped, which in this instance is very linear. Push-Pull tube amps do not make use of air gaps. Thus single-ended amps get a lot of Permeability for free. But not all is lost for Push-Pull, this is about the solution we have now.
In electromagnetism, Permeability is the degree of magnetization that responds linearly to an applied magnetic field. Mumbo jumbo to most people, I know. Don’t worry if you don’t understand it too well. Think of Permeability as the ability of the material chosen as the core to become magnetised enough to do the job. Different types of steel and other metals have different abilities to be charged magnetically.
But this core needs the Primary signal to produce the magnetic flux that Permeability of the core relies upon. The density of this flux is totally dependent on the Primary’s voltage across it, even if the core is easily magnetised. Yes, some cores respond better to this Primary voltage and thus we get better Permeability and our output transformer works. Here is the official theory:
“Magnetic flux is produced by the primary winding, and contained by the high Permeability [of the core], links the secondary winding. The mutual inductance between the two windings results in an induced voltage [on] the secondary side, whose magnitude is determined by the ratio of turns between the two windings” [Wikipedia].
So Permeability is the invisible link between the Primary and Secondary.
But not all is well. Problem is that at very low levels the Primary inductance collapses and the flux dies and so does the Perm. At low levels of audibility, the flux density of our transformer can become weaker than the Earth’s own magnetic field. It would be logical then that below that the Earth swamps the transformer, ouch! You cannot fight Mother Earth!
Mother Earth 1, Transformer 0.
So it is the Primary (with its high inductance) that gives us the flux density and that Permeability of the transformer relies upon. Permeability is the ability of the transformer to pass the signal through the transformer in tact and linearly. Lack of Perm produces distortion, or perhaps better said, lack of resolution.
Because of all this, the famous Dutch transformer designer, Menno Vanderveer, says that output transformers behave in a similar way to Digital to Analog Converters, the DACs used in out digital players. The lower the level, the lower the resolution and higher distortion. Higher levels are OK.
So we need to maintain enough Permeability at low voltage, low levels of music.
Usually the problem is partially fixed by clever designers and their clever use of materials making up the core. The penalty is the added cost, such as amorphous cores, which can near double the price – Lundahl Transformers now make amorphous cores if you want to pay extra. These High-Perm cores also suffer earlier saturation, about 33% less, which means their problems are now shifted to the other end of the spectrum, half the power rating. So this is no free lunch!
Alternative Solution: Inducing Permeability
Maybe we can deal with this in a different way and yet still play to the rules of physics. How? We can use a constant non-musical signal induced into the Primary that stops the transformer’s Permeability to drop too low and keep resolution always high. We can induce Permeability.
This graph is largely self-explanatory. The corrected Permeability is arrested in this case about –70dB. I tend to aim a little higher than. I asked Menno (with whom I spent three days with at ETF’06) what bandwidth he aimed for in his transformer design. I kinda had an idea already, so when he said 100KHz, it was clear to me. I suggested using that excess bandwidth and at several times higher than the range of the human ear to excite the transformer using a function generator. “Of course,” he said, “that will work.” When I returned to Australia we kept in touch via email and eventually this was the result:
The L.E.M. keeps the transformer constantly excited; it is always processing a Primary voltage. Our Primary inductance remains high, the core’s flux density retained and there will always be enough Perm to allow even the smallest signal to be linked through to the Secondary. We now have a highly linear output transformer that even high quality and costly amorphous cores cannot compete with. Oh, not to mention the dramatic lower cost?
The L.E.M. can be fitted and retrofitted to any existing amplifier using output transformers and transform the performance. You will hear further into the soundstage and hear all those tiny little details which now are not getting lost.
We are coming close to the end of our discussion. With correct Triode mode and the addition the L.E.M., the first gives us realistic dynamics whereas the latter, while also contributing to more correct dynamics, add the refinement necessary in spades.
Our discussion needs to finish where it began, like a conventional sonata, it is like a round trip. We do not need to make these amplifiers from scratch. We can take existing Chinese (or Asian) amplifiers and simply upgrade them with the above technology. The quality of the transformers is quite adequate for most parts. But these are exactly the type of transformer that will respond best to the L.E.M. The potential results are inexpensive amplifiers that can compare and most exceed the quality of any amplifier on the market, irrespective the asking price.
This is the Markhill EL34 Integrated Amplifier; it can be had for A$749 plus A$30 shipping to a Sydney address. Can we compete? No, but we can take advantage! I hope to see one of these soon on my workbench. Could end up killer amp for less than A$1500 total.
The amplifier I will be demonstrating at this month meeting will be the Malaysian Audio Frontier also fitted with EL34. It will be 18 Watts per channel, real Triode mode and L.E.M.
Eventually when I do the above pictured Markhill I am hoping it will reach at least 20 Watts. But these amplifiers sound a lot more powerful than those numbers indicate. The monoblock amps that I did went back to their owner in Canberra, The next day he calls me and asked if I had made the amplifiers more powerful? Nope! Same watts. They just sounded more powerful.
When a sound system is lacking realistic dynamics we tend to turn up the volume to compensate. But with these new amplifiers we don’t need to turn the volume up any louder than they were when they were recorded. Maybe that is why youths turn up lousy car sound systems so loud? I suspect so.
Take the time to come on Sunday 27th – hear the results for yourself. Also read the article re the equipment that will be on show.
Joe Rasmussen
At The Cross Roads
joeras@vacuumstate.com
