Using the Lundahl "300B" transformer LL2792
1) Using the heater windings for AC
With tube heaters, the more precise the heater voltage, the longer tube tube life will be. So yes, most tubes have 5% tolerance, but that does not mean, 5% deviation, up or down, doesn't matter. It rather means at 6% you may have a tube defect. So the closer to 0% the better. Sometimes you need to make a choice, rounding up or down resistor values to commercial values. In that case, choose for over heating, never under heating. So when choice is 4.9 or 5.1V, choose 5.1
The heater windings of LL2792 are rated 4.7V at 3A, and 0.9 Ohms. Meaning, at 3A there will be 4.7V, and internal voltage drop is 3x 0.9 = 2.7V. So open voltage is 7.4V, and short circuit current is 7.4 / 0.9 = 8Ampere. Because of the above, there is never "one" 5V winding for all tubes. Also the tube socket has some contact loss at high current (You have even 2x that loss, because of 2 pins). The tube heater voltage, as you will understand is not specified at the socket solder lugs, but at the tube pins itself. The following table takes that into account.
It is a nice coincidence, for 5V 2.5A (like 5U4G needs) you get exactly 5V at the tube pins. For all other tubes, you need a small series resistor, basically used to increase the transformer windings impedance. For this reason, and for logic wiring afterwards, the best is, to add this resistor in the heater winding near the transformer. Heat development in that resistor is low. In this way the resistor is not seen directly as a part of the circuit any more. Which it isn't indeed.
Heater Current (A) |
Voltage at transformer |
V to tube Pins |
RS Value |
Heat in Rs (Watt) |
Power Rating |
V to tube Pins |
| 1,0 | 6,5 | 6,5 | 1,50 | 1,50 | 3,00 | 5,0 |
| 1,1 | 6,4 | 6,4 | 1,20 | 1,45 | 3,00 | 5,0 |
| 1,2 | 6,3 | 6,3 | 1,00 | 1,44 | 3,00 | 5,1 |
| 1,3 | 6,2 | 6,2 | 0,82 | 1,39 | 3,00 | 5,1 |
| 1,4 | 6,1 | 6,1 | 0,68 | 1,33 | 3,00 | 5,1 |
| 1,5 | 6,1 | 6,0 | 0,56 | 1,26 | 2,00 | 5,1 |
| 1,6 | 6,0 | 5,9 | 0,47 | 1,20 | 2,00 | 5,1 |
| 1,7 | 5,9 | 5,8 | 0,47 | 1,36 | 2,00 | 5,0 |
| 1,8 | 5,8 | 5,7 | 0,33 | 1,07 | 2,00 | 5,1 |
| 1,9 | 5,7 | 5,6 | 0,27 | 0,97 | 2,00 | 5,1 |
| 2,0 | 5,6 | 5,5 | 0,22 | 0,88 | 2,00 | 5,1 |
| 2,1 | 5,5 | 5,4 | 0,15 | 0,66 | 1,00 | 5,1 |
| 2,2 | 5,4 | 5,3 | 0,10 | 0,48 | 1,00 | 5,1 |
| 2,3 | 5,3 | 5,2 | 0,10 | 0,53 | 1,00 | 5,0 |
| 2,4 | 5,2 | 5,1 | 0,00 | 0,00 | 0,00 | 5,1 |
| 2,5 | 5,2 | 5,0 | 0,00 | 0,00 | 0,00 | 5,0 |
2) Using the heater windings for DC
This is more difficult, but basically it works with this series resistor in the heater winding as well, to adjust the DC voltage.
VERY VERY important, when building a passive rectifier circuit. This will work totally dead-quiet, as long as you set up the circuitry basically for AC heating, and then supply the passive rectified DC signal (with small hum on it) instead of the AC voltage. This is THE one and only good method, but people will not understand why it is so much better. Instead they LOVE to heat the tube unsymmetric, because they understand that circuit better, but the point is, any AC residue will not be rejected, but even amplified with the tube gain. I would say, at least try it, as advised here. You will see, you need no electronic regulation this way, and any EML tube is guarantee to have no audible heater hum like this.
3) Tube rectification
To create a 350-0-350 winding, one COULD simply tap a 700V winding in the middle, but that would result not in a very good transformer. You need to see what this is really doing. Through each of the two half windings, is flowing a pulsed, half wave rectified signal. Such a signal has a high DC component. This impairs the transformer a lot, because it results in magnetic DC bias of the core, and also the pulsing will give more audible hum as desired. Such a transformer will work, but becomes too loud, gives too much voltage drop, and becomes too warm. Believe me please, but not all transformer winders will even know this. The solution for this is, to use a double package of windings. So even when there are only two windings in the connection scheme, in reality there are four. These four, are connected in such a way, that the AC components are a parallel / series scheme, and like this, the DC component becomes eliminated. This is difficult to explain, but when done right, it works. Such a transformer will be very quiet. The user does no see this, at the connection board there is just 350-0-350.