I had been looking for a higher power amplifier design to act as a test bed for trying out EL34s, KT88s, etc. when I stumbled across this KT88/6550 push-pull amplifier in Tube Kingdom Vol. 38.
As a mono-block it’s designed to generate 40-44W/ch with manageable ~150mA @ ~415V B+. The schematic is here:
I liked this amplifier because the design uses triodes in the input stages and a long-tail-pair splitter yet it is simple enough to be implemented with point-to-point in one chassis.
Power Transformer Considerations:
The mono-block HK10 is designed to produce 40-44W/ch from B+ of 415V with total quiescent current of 150mA
The obvious candidate for the PT would be Hammond 278CX with 400-0-400 @465mA. Except that this amplifier requires filament current of 1.6A per each KT88 which exceeds this Hammond's filament winding capacity of 6.0A. Also, 278CX does not have a bias tap.
Hammond also has a Fender replacement unit, 290FX that is rated to supply 450mA at 320-0-60-320V. It appears to have the identical (well, almost..) mounting dimensions as the Dynaco PA060-S. It is relatively inexpensive at $81 at Angela. Alas, its heater winding is rated at only 5.5A at 6.5V. Not enough to power 1.6A per power tube times 4.
Another possibility would be Antek AN-4T360 with whopping 360-0-360 @1000mA capacity. With two 6.3V filament windings of 5A each, it easily accommodates a quad of KT88/6550. The problem was at 4.9” diameter and 3.1” height, I could not easily fit into a 2” high chassis that I prefer.
Edcor turned out to have a great candidate, XPWR016 that is rated to deliver 450mA at 350-0-70-350V and has two heater windings each rated at 5A @6.3V. Alas, I did not know its existence until I completed the project. Oh well.
http://www.edcorusa.com/Products/ShowPr ... spx?ID=593
In the end I settled for Triode PA060-S since it has all the needed voltages and filament windings except that its high voltage capacity
of 300mA at 360-0-360 is marginal at best for a 40W/ch stereo amplifier. The amplifier is designed to operate as a class AB1 and it is not supposed to consume significantly more current in louder passages.
I would compensate for the lack of capacity with larger PS capacitance.
Output Transformer Considerations:
The Hammond 1650N, 1650P, the Triode A431S, or the Edcor CXPP60-8-4.2 will do nicely.
I wanted to try the Edcors,
Output Tube Considerations:
I had two sets of matched Sovtek KT88 pairs that I have been saving for a yet-to-be-determined amplifier project.
I also had a matched quad of Svetlana 6550Cs.
They were fine for this project.
Driver Tube Considerations:
I had various 12AU7As and 6CG7/6FQ7 on hand. They would do nicely to start with.
Chassis Considerations:
Since this amplifier is to be a tested for trying out different configurations such as UL, triode strapping, and tube rolling, I wanted
to make it as compact as it is transportable to and from a work bench.
In the end, I selected the Hammond 1441-30BK (17"x10"x2") steel black textured chassis.
As it turned out, the chassis was not quite rigid enough to support the weight of all the iron. I borrowed Ed Brown's solution - an aluminium L-beam from Ace Hardware.
How It is Implemented:
I made a minor modification to adding a CCS to the long-tail-pair.
I also made these bias adjustment pots accessible to make tube rolling relatively painless.
"As implemented" schematic:
With 6550s:
It sounds as good as a Stereo 35 with more authority in lower frequencies.
It sounds louder, too.
/b
June 15, 2015 : Added KT88 bias circuit for the second channel