• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

triode 26 as driver for triode 45

Triode 26 and 45 get lots of praise, and the cost is a bit less than tubes like 2A3 and 300B.
There are easier tubes to build with, like 6SN7 and EL84 etc. but I'd like to wind the mains x-former, filter choke and the OT x-former myself, so why not go the whole distance with the 26 and 45 while I'm at it.
And yes,-I have wound a mains x-former and OT for a Vox AC30 guitar amp with great success so I'm not completely green with such work so I know what to expect to some degree.
But the question is how does the 26 cope with driving a 45 in a SET configuration?
I like the idea of using only two stages. That's the beauty of tube amps,-they can be very simple and have a short signal path which I think is important.
Is this possible without implementing an interstage transformer or adding a whole bunch of circuits around the 26 stage?
 
Im sure others will know better than me and chime in with an answer, but I will have a go.. the 26 has a gain of about 8 along with 10y/801a/ triode 47, the 4p1l a bit more and triode 46 a bit less gain . With all of them you will need a step up transformer about 1:4 and this in all reality will have to be at the input of the amp. This will give enough gain to drive a 45. You will need to load the 26 with a choke or Ales gyrator and then you could cap couple to the 45 or you could direct couple the 26 and 45. This is what I am doing only with a triode connected 47 as input tube with a lundahl 7903 SUT the 47 loaded with a choke and DC'd to the 45. Good luck
 
EZ80 has understood this very well. With all of these DHTs - 26, 46, 47, 10Y - you will need a 1:4 step up on the input. I use the Hammond 1140-LN-C which is excellent, studio quality and the 1140-LN-D would work too. Then you can drive a 2a3, 6B4G or 45 in 2 stages, by far the best, sounds a lot cleaner than 3 stages.

You can choose to load the 26 with a choke or gyrator, or you can simply use a resistor which is what I do. I like the purity of resistor loads for classical and opera. The resistor I'm using is 27.5K and I wouldn't go lower. See my schematic on the 26 preamp thread, recent post. I cap couple with Russian teflon FT-2 caps, and wouldn't use anything else.

The 10Y is the best sounding tube here, but expensive. The 47 and 26 are both excellent but different. The 26 is very nice on voices and has the relaxed tone of a true triode. The 47 has a bit more..... shall I say focus? A bit more dynamic, and you can put more current through it. I'm happy with both - the 26 is in my system currently and the 47 and 10Y stages are on the shelf. Loads of 26 around - I have several and prefer the ST shape ones.

You should be fine. Get the V9 regs from Rod Coleman at Lyrima and use filament bias. I use filament bias for 26, 47 and 10Y. Big cathode resistor which gets hot and has to be mounted above the top plate or it's like a furnace inside. Good size heat sinks required. You'll be rewarded with a sound which is about as good as it gets.
 
A short piece of wire with gain would be nice, but since such a thing has not been invented yet a two stage setup is the next best solution.
Years ago I had a conversation with two amp designers in Norway (Elektrokompaniet and Doxa) and both swear by as short and straight a signal chain as possible and a well thought out power supply, so 2 stages it will be.
 
  • Like
Reactions: 1 users
emtor,

You are winding your own power, output, and choke.
Nice!

I would like to make a suggestion:
Use a Choke Input B+ filter
You will need more power secondary volts to get the B+ voltage you need (but you are doing the winding).
And, you will need to wind the choke to meet the Critical Inductance number.

Choke input DCV output is 0.9 x secondary Vrms (you also need to include the other voltage losses, power primary DCR x step up ratio (primary to secondary volts); secondary DCR; rectifier drop (especially if it is a tube rectifier; choke DCR.
Usually, it is a Choke, cap, resistor, cap. That resistor also has a voltage drop.

Choke Critical Inductance:
50Hz power mains; 100Hz full wave rectification; 420/Load mA
60Hz power mains; 100Hz full wave rectification; 350/Load mA

That should be a really sweet amplifier.

Happy designing, building, and listening!
 
Choke input PSU's have better regulation and makes life easier for both the power transformer and the tube rectifier,
and while winding transformers isn't hard, at least on paper, the tricky part is not falling for the temptation of overdoing parameters.
As an example, very large values of inductance in the input filter choke is tempting but that results in lots of DCR which messes up the Q unless you use very thick magnet wire which of course would result in a monster sized choke.
The PSU should be well damped, not too slow to recover and not so fast that it overshoots.
PSUD II is a good tool in that respect.

As for the 45, it needs 35 - 40 volts p-p to drive it to full output.
A 26 has a gain of 8 and as far as I understand it the 26 would then need 5 volts p - p on the input to satisfy the 45's drive requirement.
A mic transformer of 1:4 on the 26 input would lower the required input signal to 1.25 volts p - p.
Is this close to correct?
I plan on driving the amp from the line out socket on my Quad Vena ii amp. The Quad has a decent DAC,-analogue inputs, bluetooth, digital and optical inputs, wifi, ethernet and the whole works enabling me to use many different types of signal sources.
 
I don't know the details until I've done the math, but you could go here,
where you can purchase spreadsheets for power transformers, output transformers both push-pull and single-ended plus chokes.
Without such spreadsheets you'll be fighting with math for a long time since transformers require you to do many iterations of calculating before it all fits.
The spreadsheets are not very expensive either.
Also you can purchase the materials you need on the same site.

To get an idea of how to go about it after all the math is done, go here
and scroll down to find the chapters on transformer winding.

You will need a winding machine of some sort, either a cheap manual machine from China or build an automatic one yourself.
This site shows you how to make one. Scroll down and look through his older posts to find the videos.
 
  • Like
Reactions: 1 user
The Top Classical quiescent operating condition for the 45 in single ended is:
275V plate to filament
-56V bias
36mA
4600 Ohms RL
2.0 Watts output
Your driver tube for this quiescent condition should be able to swing from +56V to -56V (112V peak to peak).
True, the 45 will badly distort at this level of grid signal, but if the driver can not do that, it will distort before that signal swing.

There are 2 other classical quiescent operating conditions for the 45:
They have lower plate to filament voltage, lower bias voltage, less plate current, and lower power output.
At those quiescent conditions, the bias levels of -31V and -50V, your driver should be able to swing 62V peak to peak, or 100V peak to peak.

The 2A3 with -45V bias for its Top Classical operating conditions is -45V bias, and 3.5 Watts output.
That is an easier drive than the 45 at its maximum power output. A driver with 90V peak to peak will work.

I like both the 45 and the 2A3; have used both in single ended; and also used the 2A3 in push pull.

My favorite 45 amplifier I designed and built only put out 1.5 Watts/channel. It was a stereo amplifier, 12AU7 driver (low gain) and RC coupling.
I used it at work, and I could not turn the volume up, even on the moderate efficiency Usher S-520 speakers .
One evening after 5:00 I turned it up. A lady from several aisles away came over and told me how much she liked the sound.

Have fun designing, winding, building and listening!
 
There are 2 other classical quiescent operating conditions for the 45:
They have lower plate to filament voltage, lower bias voltage, less plate current, and lower power output.
At those quiescent conditions, the bias levels of -31V and -50V, your driver should be able to swing 62V peak to peak, or 100V peak to peak.
A lower power output is fine by me since I have a few Coral Flat 8ii full range speaker units at my disposal.
Sensitivity is 100dB and with no x-over to soak up power they don't need much to fill my 25 square meter (270 square feet) living room.
 
I use 26 to drive my EL12n outputs. Makes for a pretty simple build. EL12n is available new in Europe. Sounds very good.

EL12n + 26.png