If you have comments, suggestions, going to use this design please let me know!

I'll try to explain why I made all choices, if anything isn't clear or there is an error I made, please let me know.

Click here for the amplifier.

Click here for the powersupply.

• The power supply was choosen 425 volts, this is the output of the coil coupled supply with the 500 V torroidal transformer I had already build in. The bias current was choosen 100 mA, just below the maximum rating of the tube, to prolong its live. The negative bias voltage need to be -90 V with this bias point.
• The output transformer primairy impedance was choosen 4k, giving a very lineair 8 watt output.
• The 10 Ohms resistors at the kathodes will be connected to my "on panel" voltage meters (range 0-1.5 volts). This way I can easy check the current through the the tubes, and correct them if necessary. Also they protect the outputtubes, when the current gets too high, they will simply burn out, and stop all current. So use 250 mW types!
• The other complex network of resistors is the humbalance. I needed all the resistors to be able to use the very beautifull Allan-Bradley potentiometers.

• I wanted to use just one stage before the end-stage. This asks for a driver tube capable of handling high lineair voltage swing with sufficiant amplification. Also I wanted to use a directly heated triode, the final choice was the S6 made by Electro Mecano from Denmark. This is a very lineair, very well build very good looking 15 watt DHT. The sound is very great, in my eyes the ideal driver!
• The S6 gets a DC heathing voltage, solidstate rectified (yes, there is some of that new stuff used!), resistor decoupled and stabilised to get it as clean as possible. When using AC or non-stabilised DC, the hum just became too much, it keeps very hard to get a lownoise feedback free amplifier playing at 104 dB speakers....
• The S6 is choke loaded. I used a 40H signal choke (interstage without secundairies) permalloy Double-C core, with silver wire. This is to be sure the very low level signal is handled best, to get an as good as possible final result. For full range the value of the choke has to be somewhat bigger, I can use 20 Hz, because I use the amps only above 150 Hz.
• To make the amp full range, one has to chance the coupling capacitors as pointed out.

• This powersupply is build partly on a seperate enclosure, not only to save space on the amplifiers, but also because I have but one (and didn't want to buy another!) highvoltage transformer. It's a 1200 VA ILP torroidal transformer, 2*500 volts, about 1 Ampere. It is choke loaded, to have an as smooth as possible sound.
• On board of the seperate powersupply there are also the tube rectifiers (5R4WGT), two filter coils (Sowther, 10 H/250 mA/50 ohms) and capacitors (ChateauRoux, 100 uF/630 V, MKP). When coming on board of the amplifier enclosures, there's another 20 H coil and ChateauRoux 100 uF/630 V MKP capacitor.
• The power gets switched on and off by relais driven by the negative bias voltage of the monoblocks. If this voltage drops for one reason the powersupply will also be switched off, so the powertubes will not get harmed. Also it prevents the high voltage powersupply from being on without an amplifier connected, which could be a potential dangerous situation.

• My first thought was to get this powersupply from the B++ supply. Because I needed a seperate transformer for the negative bias voltage anyway, I decided to make it seperate also. It's tube rectified with a 5Y4GT, an octal double diode (5Y3GT with different pinout). With all the coils and the capacitors it the voltage is very clean and hum-less.

• To be real consequent I also made the negative biasvoltage powersupply tube rectified and stabilised. I use a 5W4 double diode as rectifier and a OA2, 150 volts stabiliser to get the voltages I want. It's decoupled with an GE 20 uF paper/oil capacitor to be sure that no garbage gets through it to the amplifier stages.

• This high current supply I rather had on a seperate chassis. This supply is quite heavy and coil decoupled. There are some big cooling blocks on them to get rid of the surplus voltage. The 4.7 mH/1 Ohm aircoil with the 30000 uF (13 Hz) will filter the garbage out of the voltage, and prevent the high capacitance from drawing to high currents out of the transformers. Big advantage is the slowness of this supply, it will only come up slowly, preventing the tubes from drawing big inrush currents.
• This supply also contains the stabilised DC supply for the driver tubes. Even before the LM317T stabilisation chip it is made very clean, just to be sure no Hf-noise (made by the switching of the solid state rectifiers) gets into the system anyway.