I’m a big fan of the thoriated-tungsten filaments and I value them not just because they do look very cool, but fundamentally due to how they sound in a single-ended stage. I hope to build a nice push-pull amp with these type of valves soon.
I found a pair of HY1269 valves recently. This valve is not well-known amongst the used ones out there but they do have some interesting characteristics as a directly heated tetrode that would be interesting to consider it for a plate-to-grid (Schade) feedback configuration. With its 30W of anode dissipation capability, it’s a good candidate for an output stage. However, it’s quite likely that you will have to drive it in A2 to get the most out of this valve. Like designs using 811a and similar transmitting valves, the HY1269 can be operated in class A2 even with no signal on the grid.
As per my “Robustiano” design using the 6P21S, it would be nice to see extracting 6W or more out of this transmitting valve. I’m sure that the thoriated-tungsten touch will provide a lovely sound if properly implemented.
Tracing this DHT valve
Tracing DHT is a challenge. This valve has a handy connection to the mid-point of the filament. This will ensure the cathode and the suppressor grid to be at equidistant voltage from the filament ends.
To trace this valve using the uTracer, simply connect the filaments (pins 1 and 5) to the external DC regulated power supply, the cathode to pin 4, the grid to pin 2 and the screen grid to pin 3. Finally the top connector should be wired to the anode.
Looking at the data sheet you will find that there is very limited information on this valve for class A operation. Also given the 300V limitation of the uTracer, data available for the typical operating points is of little use. You will have to refer to the average characteristics to compare the emission of your valve:
Looking at the curves, for 300V the anode current will rise up to 170mA when Vg=0V and 125mA approximately when Vg=-10V.
The sample valve used in this test is about 90% of the expected emission levels for Vg=0V. However the variance is significant at -20V:
Again, the uTracer can reveal detailed information not available on the data sheet. In this case, the screen current at lower grid voltages is surprisingly revealing and of great value to produce the Spice model of this pentode using Derk’s “Extract Model” tool:
The Spice model
Derk used “DerkE” parameter to create the pentode model given the curves suggested a beam-like behaviour:
C:\Program Files (x86)\gnuplot\bin\wgnuplot.exe 4 !number of .utd files HY1269-pentode-150V HY1269-pentode-200V HY1269-pentode-250V HY1269-triode B DerkE 30. 0 Icmax=120
Two observations Derk noted:
- The secundary emission effects in the screen currents are not well modelled, especially for higher screen voltage (Vs)
- The anode current (Ia) characteristics show a very round ‘knee’, not representative of a beam tetrode, but more of a regular pentode. This effect can be caused by “over compensation”, i.e. the beam plates are negative rather than at zero potential.
Derk then ran a model based on ‘Derk’ rather than ‘DerkE’, to better fit the observed over-compensation. Results are below:
Which shows actually a decent fit.
As Derk points out:
“At the end this begs the question how directly heated cathodes and suppressor grids should be connected. In the end it all seems to tie down to that.”
