The Aa Line Preamplifier III – The Completion of the Amplifier Part

Hello to everybody,

today I want to show the completion of my amplifier chassis with the constant current filament regulator boards. As already stated I did use the predecessor of this preamp for several years with different battery types, since conventional dc regulation showed a lot of problems and did not sound as good as batteries. So I had to struggle instead with contact problems, broken cables and dying batteries for years, which was a real pain. I had started with lead batteries, later did I use conventional AA Nimh's, a pack of three makes 1.5 hours of listening possible. So did use several packs, the everyday use and charge cycles made life for them short, almost every year I had to exchange the most packs.
In 2010 I did learn about Rod Colemans Constant Current regulator boards, which have been designed for this particular dedication, the filament supply of DHTs and for the use in pre amplification stages as special option. These boards can be configured for almost any triode type, even the big ones like 211, GM70 or 845 with high current draw figures can be used. For each triode one board is needed separately, which means here in this case with two Aa's running in parallel mode as couple will be fed by one regulator. They don't run very hot, so they are quite easy to use in the amplifier enclosures.

The Coleman regulator boards are mounted to both chassis sides to dissipate the heat and to be very close to the filaments of both Aa couples, here shown with one board.

The regulators are dead quiet and don't show any deterioration of the sound structure, comparing them to batteries. They will be available though Rod Colemans web page and are a great help for everybody using directly heated triodes in audio amplifying devices. They need to be implanted with a raw power supply each, which should be designed to match the final preferences set by the tubes itself. Any extra power will be burnt down into heat by the regulators itself. As example for the Aa with 3.8 volts for their filament a transformer with separate windings for each raw supply with minimum 11 volts output is needed. Each Aa draws 0.5 ampere of current, together it is 1.0. The boards allow the exact setting of the current to match the variations of such old vintage tubes.

Since the Daven attenuators  are extremely difficult to find nowdays, which means it is almost impossible to find them as stacked stereo units in values like 100K or 250K, I did use one mono type for each channel. With everyday use of such a preamp, separated attenuators for each chanel would drive me crazy. With any control of volume you need to match the stereo balance again, a night mare if the amp is not positioned exactly in the middle of the speakers. In order to make the use more comfortable, I introduced a coupling system. With a o-ring both shafts are coupled together by wheels from a toy steam engine. The o-ring keeps both attenuators in balanced mode since I need to adjust one of them. If, I can hold one since I turn the other one where need. We will see if that system will do comfortable within everyday use, otherwise I have to design a spring supported clutch system…

O-ring coupled shafts for the Daven attenuators.

The back shows several sockets for different dedication. From left there is the power input socket (6 pole) where both raw voltages for the Coleman regulator boards will be supplied. The other two contacts are for the filtered anode supply. Next to this socket are two output xlr-sockets, the two main output terminals of the amplifier, these can be switches for the two corresponding rca sockets as alternative outputs at the other end of the back panel. All paired rca sockets in-between are four different input contacts, which are connected to the source switch, operated from the front panel. Four inputs should be enough nowadays, since the most sources are fed through one dac, like web radio, streaming clients, computers and similar. So there are three inputs for other sources like phono equalizer or a recording unit and here the second set of output sockets will be helpful.

After mounting the two potted chokes on top of the amplifier case and after connecting the tube sockets with four Aa tubes, the amplifier chassis is complete and is waiting for its complex dedicated power supply. When the power supply is completed and will work with the dedicated smooth anode supply of 200 volts and the two raw supplies for the regulator boards, I need to adjust these with a dummy load instead of the real tubes to prevent an overload of their filaments. But till there is quite a long way… As almost everybody knows, a good power supply is by far more complicated than the amplifier part. I will show it with an option of two different versions, a quite simple tube rectified supply with the three different voltages to keep this preamp a simple stand alone solution.

For the aficionados the series of articles will be continued showing a quite unusual pre-amplification concept, which will have the future in mind. Similar to german broadcast preamplifiers of the 1950ties I will show the build of different stages with different amplification factors. A gain amp (+50 db), a catch-up-amp (+25db) for compensation of losses and the already shown line amp for impedance matching to the power amp.
With these modules and a dedicated power supply almost any use in a modern audio set can be managed. For example a typical dac output stage will need roughly a amplification of +50 db, almost the same than a good phono stage before RIAA equalizing. When equipped with a input transformer with different windings and different ratio, like 1:10 or 1:25 (just to mention the extremes), the same amplification stage can be used for both sorts of applications. Since the former digital signal is now strong enough for the impedance amp (line), the phono signal needs further equalization with a RIAA filter with typical loss of -25db at a following stage. So another amplifier with +25db gain is here needed for compensation of the losses to level up for the line stage.

A multi voltage power supply will be needed for such a amplification chain. Such complexity needs to match highest standards and components, potted oil filled transformers and chokes would be first choice, but it is already difficult to find this stuff nowadays. The anode supply develops with tube rectification at the beginning, ideally made with a classic mesh anode rectifier tube lie RGN1054 or 1064 as perfect partner of the Aa tubes, to be continued with several chokes for HV and LV: 1. a well filtered B+ for the DHTs in the line stage; 2. two well filtered low dc-voltages for the Coleman Regulators; 3. a further filtered and tube regulated anode supply for the gain amps, and finally 4. a well filtered dc-supply for their filaments of all indirect heated triodes. I assume the weight, complexity and size of a well made power amp to get an idea about the case.


Read on soon,
Volker

The Aa Line Preamplifier II – The further Completion of the Amplifier

Hello to everybody

interested into the completion of my Aa-DHT-Preamplifier. As already stated, this line amplifier is a perfect solution for the mainly digital based systems, as it can be a exceptional supplement for any analogue chain with separate phono equalization. Implanted into a system with advanced dynamic attributes and with careful set up to harmonic structures, this preamplifier can open a window into the finesse of advanced audio performance, which I never have been able to obtain from commercial made line preamplifiers. A typical variety of incomparable attributes accompanied with directly heated triodes in small signal amplification stages are the base for such a extreme statement. DHT's for small signal output are very linear in their whole operation spectrum. They have a extraordinary refined character, sometimes show up with unique tonalities and exceptional dynamic abilities. Exactly like their more powerful sisters in power stages, but with a much deeper scaled effect because of the smaller signal ratio. The unbelievable attitudes of the Aa-Triode give a extra portion of refinement, linearity and tremendous wide open transparency, combined with harmonic detail I do not know second. Commercially made tube preamps could and can only be designed around available types of tubes, it was almost impossible to find bigger amounts of exceptional tubes since the renaissance of the tube amplifier in the 1980ties for hifi, nor it is possible today. The exceptional icons of indirect heated tubes for small signal amplification, independent of mu characteristics and impedance values, are mostly tubes made for professional dedication, like the WE 417 and 437, the red RCA's like 5691 and 5692, all the Telefunken special series like EF804s, EF801s and the red tipped types of all ECC types, the D3a's, C3-series, from other makes like Bendix, some former Soviet brands and uncountable lots of rare others tubes not named. All these exceptional quality tubes are not to be found in commercial audio amplifiers, where they could show their incredible qualities. But even these exceptional tubes cannot tame up with the same audible benefits of direct heated tubes, they are able to perform tremendous other benefits but rarely with such finesse of dht's.
The tungsten thoriated filament of directly heated tubes is acting as cathode in the signal part and creates very special peculiarities. On one hand responsible for such incredible aural benefits, but as well well for very delicate conditions in everyday use. Everybody common with DHT's in the power stage might be experienced with the typical sensitivities of such tubes for hum and vibration. In reverse to the amount of signal to ground ratio current, these sensitivities are ten times or more higher. This can make the use of such tubes very uncomfortable or even more, – it can be a real pain. I am using such tubes since almost ten years by now in such implementation for every day, within that time it has come sometimes to a point, where I just switched off my stereo, because I was unable to locate some reason for apparent hiss or hum. Completely exasperated after some time of search I decided to find reasons for the problems another day… In particular in the first years, when I did use flying batteries for the heating of the filaments, contact problems with broken cables, clips, battery containers, – flying batteries create here a permanent conflict zone. With a fixed and permanent installed power supply solution (constant current source) the most of such problems are now history. For those which are already on a shortcut, a  common or even complex regulated voltage supply is unable to support the aural qualities in the signal loop. It needs a complex constant current source with own complex (highly filtered with common mode choke) power supply per tube!It has to be said with impertinence, a direct heated triode is not a plug and go installation. If you want such a ease of use, you better should go for something else than tubes and in particular for directly heated triodes in general. Those respecting the extraordinary outlay will get gifted with the most colorful natural soundstage possible.

For those undaunted here a new life with audio might start soon. The Aa-Line Preamplifier its a fantastic solution for the digital chain. In combination with a good dac all the quite new benefits of digital stored music (i.e. better recorded standards, storage and organization improvements, a wide spread program of sources: world wide radio stations with lossless compression, cloud based streaming, etc), will find a exceptional natural sounding input stage, able to perform with a perfect holographic soundstage, if the following components (i.e. power amplifier and loud speaker) are able to support such a level of resolution. All regular readers of my blog have got already certain ideas how to get close to such a rare and uncommon performance. For me the exceptional resolution and the incomparable micro detailed dynamics are a base to operate a huge cone speaker like the Tannoy with wide medium frequency performance to a higher degree of resolution other designs can not support.

Today I will try to show some more completing of the amplifying enclosure. Now the rubber isolated perspex board has been mounted to the chassis, a distance of 20 to 30 mm between board and case is a quite good value since the Aa-tubes have a solid dimension, more known from typical power tubes of the 1930ties like coke bottled 6L6 types or similar. So the tube will be hidden with the socket inside, wile the balloon shaped glass will show out. The tube sockets are mounted at the perspex board with  some turrets to find place for the very few components and their connecting wires, which are necessary for the design of the line amp. This version of the DHT-lineamp is the "short wire version" of this topology. Since it incorporates plate chokes instead of output transformers (the long wire), the physical signal length from input to output sockets is less than 40 cm wire, with just a tube system (plate-grid-cathode/bias resistor, tubes in parallel mode), one oil-paper cap in-between. It is almost impossible to do any shorter design and this explains the tremendous influence to the final performance of the Aa-tube.

Solid silver wire with cotton hoses customized made near to the unwired socket board

Used components

To bring these 40cm to the perfect resolution I did decide to use solid 4-p silver wire (99.99%) for the signal path. I made up the wires myself from solid silver wire (0.5mm or 24 ga) and did isolate them with natural cotton hoses. The negative loop/earth path is made with thicker blank silver wires (1 mm or 18 ga) per stereo side, which get at one point together. The few resistors are carbon comp types of minimum 2 watts heat dissipation and 3 watts metal oxide types for the cathodes. The attenuators are classic potentiometers from Daven with switched 21-step carbon comp resitors for both sides. All capacitors (4 pc.) are oil-paper types, the chokes are Lundahl LL1667/10mA both coils in series (400 H  for 15 K at 30 hz), potted with wax into aluminum enclosures. The chassis for the preamp and the power supply are from the steel 14xx-series from Hammond. The used tubes are nos Valvo Aa's from war production in 1944.

Amplifying chassis from underneath: at top all input and output sockets; left and right Daven attenuators with input source switch in the middle; next level left and right coupling oil caps and b+-support capacitors; the board with input voltage devider, cathtode resistors and all silver wires; knobs with extension shafts for the front panel.

The used parts and components reflect my final selection for this amplifier type, I will definitely use the built sample for the next coming years. I did decide about the rare and expensive parts, till it might be my final line preamplifier and this is going to be a very important component in my audio chain as a result of increased digital format use. It can be made with similar parts and components (chokes, attenuators, resistors, switches, etc.), which are more easy available at a level of 95% of the same performance at the half of the costs. It might depend to the interest out there, may be I decide to make a kit with this design...
Within my next entry the amplifying part of the design will be finished with the inclusion of the regulator boards for the filaments. After this I will start to show the more complex design of the power supply for both pre amplifying stages, the line and the phono stage. Both stages need very different supplies, till the Aa tubes run in pure class a, so they need a quite simplified anode supply together with exceptional complex heater supplies per tube. A class b operated phono equalizer will benefit enormous from a complex filtering chain with chokes and a integrated tube regulation for the anode supply, combined with a simple dc heater design. If you want to follow its realization 

read on soon, Volker