Note the size of this amp compared to the cigarette packet!
Hello, all you high-powered tube amp maniacs out there! Having recently been commissioned to design and build four 1kw tube amps, ........yes FOUR!!....... , I decided to use four of the classic 813 tubes rather than multiples of KT88's. In-fact twenty KT88's would have been needed to get the 1000 watts!
An 813 tube (right) compared to a KT88 (left).
A. CHASSIS No 1: THE MAMMOTH POWER SUPPLY
Check out the size of that mains transformer against the ciggie packet!
Underside of the power supply.
B. CHASSIS No 2: THE AMPLIFIER ITSELF
I have not cut any of the transformers' wires as these will of course be re-used on one of the production models in due course.
Underside of the amplifier section itself, showing the pre-amp, phase splitter, and push-pull driver stages.
C. MORE PHOTOS OF THOSE CLASSIC 813 VALVES !
The 813's are run with 2,500 volts on the plates (anodes), 750 volts on the screens (grid 2's) fixed bias of course. This humungous monster is pushing 1015 watts RMS!! You could toast your bread from the heat they are kicking out!
Hello everybody, well, I have at last got off my bum and found the time to do some technical details on this humongous monster!! Before I continue with this there are two points I would like to mention. First……..
(1). I am overwhelmed (but not surprised) at the immense amount of e-mails and phone calls that I have received about this project since it was placed on the site a few months ago. Thank you for all your interest. I realize it is hard to find something like this amplifier elsewhere!! And………
(2). If any of you out there ever tackle this enormous baby (or even a 500 watt version with 2 x 813’s and just half the HT current.....everything else stays the same apart from a plate-to-plate load of 19,000 ohms for just two bottles) please remember that you are working with extremely, potentially lethal voltages!! Apart from the obvious 2,500 volts on those top-caps, even the 750 volt screen (grid 2) supply and the lesser 500 volt rail for the pre-amp and drive can also stop the heart!! Be very careful!
OK, having gotten that off my chest (pardon the pun!) I will try here to cover all I can to help you. Over the many years that I have repaired, rebuilt, designed and built valve (tube) amplifiers, I have many-a-time stuck to a fairly rigid “tried and tested” formula learnt from these many years of experimentation and results. The CBA-500 (please see this article) follows a similar design pattern to this CBA-1000.
My own designed phase splitter circuit (ECC81) and directly coupled cathode follower/push-pull driver (ECC82/12BH7) is “world beating” and has never let me down! On the CBA-500; the twelve EL36/6CM5/EL360’s needed quite a bit of drive and this can be barely achieved straight from the phase splitter tube. We need a low impedance cathode follower supply/drive to do this, hence the ECC82. In the case of the CBA-500 the HT line is only about 435 volts. This is a trifle low for the push-pull driver so, a crafty but simple and effective trick here is to stand the cathode resistors from the said driver tube not on ground but…….on the minus, negative bias rail. This in turn now puts a greater potential across the driver tube (e.g. HT @ 435 volts but now with the -52 volts bias rail added gives us a total usable voltage of 487 volts), raising both the drive voltage and clipping points. (We always need the drive to be clean, unclipped before the output stage clips!). The other point I must make here is the fact that most bias windings on “off the shelf” transformers are usually of only a few milliamps. This being the case, we wouldn’t be able to use it for this purpose. I therefore always wind my bias windings for at least 300ma or more. This then leaves plenty of headroom for using the bias winding for just such issues as helping the driver tube, along with other items such as series connected DC pre-amp heaters and various relays etc.
On this amplifier, being the CBA-1000, the HT supply for the pre-amp/drive stages is already around 500 volts and generally speaking would be adequate but, although 813 tubes are reasonably efficient and don’t need excessive amounts of drive, (about 180 volts grid-to-grid to achieve the 1000 watts from four), I still opted for the idea of placing the push-pull driver cathode resistors onto the bias rail; this time of around -122 volts. This gives us around a total of 620 volts to play with on the driver tube!! Using a 12BH7 instead of the ECC82 (the 12BH7 being much beefier) gave me at least a further clean, unclipped drive of 20 volts above the required level and is still clean well after the output stage clips! Brilliant! The 12BH7 does run a little warm but……is fine and comfortable at this.
I have now included the SCHEMATIC DIAGRAM for this beast on the site for your help and interest. I think most of it is pretty self-explanatory; any questions please, just e-mail me.
I have always run my pre-amp/driver heaters on DC, right back as far as the sixties!! This is very beneficial all round. As the 12BH7 has a higher heater current than the ECC8? series of tubes, I have done a paralleled 6.3 volt DC chain instead on this amplifier. (Please see power supply schematic: “Y Y” heaters).
For those sufficiently knowledgeable to notice, I always tie the bottom end of my individual bias pots straight on ground. This is not normally a good practice as obviously the pots can then turn the bias off completely to the said tube(s), but….anybody who understands this would not allow this to happen anyway, and would always make sure that there was sufficient bias on the tubes’ grid 1’s before powering the amp(s) up. My reason for doing this is because it keeps the bias supply as low as possible from an impedance point of view and “rock solid”. I also use the 20 turn cermets pots as this too gives very good smooth and accurate setting up of the individual bias per tube. Lastly on this point, I also use 47uf caps on the wipers (output) of each cermets pot too. This also helps keep things “rock solid, accurate and smooth” on the individual bias adjustments.
Moving on, the 20 turn Cermets pot across the plates (anodes) of the phase splitter tube (ECC81) is such a simple, yet extremely effective idea. Usually, on virtually all amplifiers this would be two fixed value resistors, and yet, this simple addition balances the phase splitter tube, the push-pull driver tube, all associated components for both the afore said tubes, all the output tubes (no matter how many) and any deficiencies within the output transformer; all in one fail swoop! This works wonders. Try it and see!
As regards the pre-amp stages, once again I think the schematic’s are very self explanatory. If you didn’t wish to include the very front-end and tone circuits, then you just simply stop at “pre-amp schematic number two” and your input is as a slave-amp would be. If however, you did decide to go the “whole-hog” and build my own designed front-end pre-amp section too, then I can assure you that you will not be disappointed! My own-design tone circuit is incredible, even though it doesn’t use a “middle control”. I have designed this tone circuit in such a way so that when the bass control is increased, this brings in with it the low-mid. Similarly when the treble control is increased this brings in the high-mid along with the presence too. The combination of the two controls is absolutely breathtaking!! If you aren’t “gob-smacked” with my tone circuit, I’ll eat my hat! (and I do wear one!).
Two final points here, and touching back onto the fact that this amplifier uses extremely high voltages. (Well, they are high as far as most commercial amplifiers are concerned.)
(1). The two F,1amp fuses on the + and – ends of the main 2,500 volts HT rail should preferably be of the inch and a quarter type (32mm), and should have plastic covers over them, with all connections covered with shrink-sleeving too. Both fuses should also be mounted on a single (each) insulated pillar, and a minimum of one inch (25mm) from the chassis. In the event of a blown (shorting) 813 tube, once either (or both) of these fuses pop there is a chance of arcing if the fuse length is too short and/or too close to the chassis with connections exposed! And finally……….
(2). For the same reasons as above, the four 1ohm, 1%, .6 watt resistors (.6 watt is fine, the voltage drop across these is miniscule) need to be mounted at least one inch (25mm) away from each other, and at least one inch (25mm) from the chassis too. When one of the 813 tubes becomes short or flashes across inside (as happened on this prototype a couple of times with faulty tubes bought on EBay!!), this resistor, as well as being your reference point for each individual tube's bias, also acts as a fusible resistor too. However, when this resistor blew open on the prototype, the high voltage arced across to the others adjacent to it. Keeping them apart would stop this problem occurring.
Enjoy the amp, enjoy the project but...........please be careful of those high voltages!!
Cheers for all your continued support and interest.