|topology: SE Class A||driver: 10Y/VT25||sound quality: TBD|
|input sensitivity: ~3V||output: 801A/VT62||nostalgia factor: 10/10|
|rated output: 1-1.5W||rectifier: RG1-250||hernia scale: 9/10|
Click here to see all project updates
The last DHT amps I was listening to in 2004-2005 was the Exodus III and had the 10Y in the output. I tore down those amps after about 2-3 years of not listening to audio. The parts have been in boxes ever since, and we need to bring some change to that.
My favorite tubes, in general, are the 26, 71A, 10 and RS241. I've always loved listening to these tubes and no doubt they will make it back into the rebuild of a full DHT system.
At the moment I am not into anal-retentive audiophile listening, and more into the emotional satisfaction of listening to music that I like. This could be Ella, this could be Daft Punk. Next to this amp I am also working on an amp system that is full-on fun factor and not so much about the ultimate sound, incorporating pretty much everything I have always disliked in audio like toroids, regulators, PCBs, pentodes... but it is all in good fun. This amp however, is a return to my last system, with some nice improvements here and there. As I never threw anything out and sold only a handful of parts I had in my stash, I basically have everything I need to build these new amps.
My love for the old globe UX210s, sight and sound, led me back to this tube (and the fact that I have all the signal iron for them in yummy Tribute amorphous variety). Right now I am still pondering if I want to waste the precious bare nickel globe UX210/CX310s, or stick with the big pile of NOS VT25 and VT62 versions I have, which will last me a lifetime.
The brainfart for now is to double up on the 10 and go Sakuma-style by using the same tube as a driver. Well, sort of.... I'd like to use the VT62/801A as the output tube to allow for more plate dissipation and voltage. Ideally I would want a bias of 500V/30mA on the output tube and load it with 16K or higher. At this time I already have 16K:16R Tribute OPT, but would like to actually use a 30K load like I mentioned in an earlier itteration of the Exodus amps, but never really tried out. So to start out I will use the 16K, but the plan will be to go to a 30K OPT down the road. I'll need to find someone that can wind a decent tranny to spec for me.
Here are the loadlines for 16K (1.5W) and 30K (1.0W) on the 801A:
The eTracer predicted that with these actual measurements and loads, I would have a THD aorund 0.9% for the 16K (1.5W out), and around 0.6% for the 30K (1.0W out). I've measure several tubes so far, including triode connected E180F, E810F, 6Z9P and 6Z11P, but this tube is still the most linear of the limited selection I tested. Where the 45 and pentodes generally show 2-3%, the 10 and 810A cut this in half. A really awesome tube to hear and on paper as well now.
Here is a plot of the 10Y in comparisson; notice how the bottom part starts to drift up, unlike the 801A.
The driver will be more or less the same tube as the output, namely a 10 variant. Seeing as the output will be using 500V
The output section will be combined with the Usagi 26 preamp, so I've aded that section in as well. The 26 will feed the driver with an interstage in either 2:1 or 1:1 plus offer an additional output for the bass amps.
I'm playing with the idea of using a 4P1L as the driver tube vs the original 10Y. The B+ supplies will allow me to use either of these, so I will definitely test and compare them.
Concept of the signal section.
The "Raw DC" supply will be in a seperate chassis from the signal section, mainly to keep the size within reason, but also to keep transformers as far away as possible from the tubes. The "raw" supply is just passive supply giving around 16V and these will used on all six tubes in the pre/power amps.
Check this project for the passive part of the filament supplies.
Inside the signal chassis there will be a few small caps, followed by a Coleman regulator for the filaments.
Check this project for the passive part of the bias supplies.
A simple CLCLC+RC filtering with two mercury vapor diodes for the rectification. As the diodes will need a few minutes to warm up and vaporize the mercury, the filament and high-voltage windings will be put on seperate transformers. I could just put a switch in the anode circuit of the diodes, but that think will spark. Also this way the high current filament taps will be kept away from the B+ windings and allow other options in the future.
The filament transformer has a a 2.5-2-0-2-2.5V/4A winding for each diode, which actually pull 3A each at 4V. I've added the 5V taps should I want to switch to an american rectifier like the 83, 5Z3 or 5U4G etc. There are a few high-voltage windings to play with different rectifiers and size of the first capacitors. This will allow me to get anything from 375V to 600V out of this supply.
There is a also a light 24V winding which is used to power a small relay. When the B+ is turned off this relay will activate the discharge resistor at the output so that there will not be 500-600V lingering on the output tubes while their filaments are turned off.
The final RC filter stage is actually in the signal section itself. The resistors will allow me to fine tune the voltage a little, within 10-20V or so if needed.
Let me know what you guys think.