a mono vacuum tube audio amplifier
In my ham shack i have a low power audio amplifier, built on a cheap cake pan, that I used for experimenting and just as a general purpose audio amplifier. I thought it might be useful to have a similar device here at my second house, and also take an opportunity to experiment with some different construction techniques for vacuum tubes.
It won’t need to be much of an amplifier - it will join the kitchen table lineup at my lake house and provide some tube audio sound for the planned second generation of RF stuff, while also providing a low current, tube voltage, power supply if / when I might need one through an octal socket on the rear panel.
I have decided to try to build it on a 3D printed chassis instead of a cake pan, and already have a chassis printed and working power supply constructed.
The power transformer is a $1 hamfest scrounge, that on the line voltage here at the lake outputs 230-0-230 and 7.2 volts on the filament winding. It is mounted above the chassis by the width of a #8 nut and a second 3D printed washer so that air can flow around it for cooling and minimize heat transfer to the plastic. The chassis is beefed up a bit in this area to support the weight of the transformer.
The rectifier tube is a 6X5. This tube was selected because, AFAIK, it has the lowest voltage drop (22 volts) of any of the common 6 volt rectifier tubes; despite its somewhat poor reputation, this type has worked flawlessly in the utility amplifier in my ham shack.
The power supply is a generic CLC circuit, with a 150K bleeder, and a Hammond 154M 2H choke. The input cap is 4.7 uF and the out cap is 100 uF. The secondary of the power transformer has pretty high resistance - about 190R / leg, so I did not add any series resistance between the transformer and the tube.
Each filament line has a 0R5 resistor in series to drop the filament voltage to an acceptable level. The string is elevated to 1/6th the B+ voltage by a 150K/27K voltage divider.
All power supply voltages are available at the rear panel octal socket. The output voltage (loaded by the bleeder and the elevator divider) is 300 VDC. The filament voltage under the 6X5 load is 6.4 volts. The 6X5 data sheet indicates I should get 250 volts up to about 55 mils load on the power supply.
Most of the wiring is some stiff teflon insulated stuff that I got at a hamfest. The other parts are cheap generic AliExpress stuff. All of the metallic parts that can be touched are daisy chain grounded to the safety (AC mains) ground for some shock protection. The circuit ground is also tied to the safety ground.
Here is a temporary schematic. It is dirt simple at this point
The input is a triode - 6AB4 is 1/2 of a 12AT7, and 6C4 is one half of a 12AU7. The ideal load for a triode is a current source, and an IXYS 10M45 is used as a current source here. The trim pot sets the cathode voltage to about 1.25 volts.
The CCS IC is somewhat expensive, and a conventional resistor load can be substituted if desired. 10K - 15K is a good value for the 6AB4 at the lowish voltages here, and would probably be OK as a start point for 6C4. If more gain is desired, 6AV6 is 1/2 of a 12AX7; I would probably start around 36K as a plate load for that tube. Tie the diodes to the cathode.
The power tube is wired as a triode for simplicity. The cathode resistor needs to be set to the correct bias voltage depending on the tube and B+ voltage. The grid resistor on the power tube is low at 240K, but seems to be OK here.
At some point, I will try some different input and power tubes, and different configurations with feedback. The whole point of this besides a utility audio power amplifier / tube voltage
power supply, is just to play around and experiment with.
One objective was to determine whether or not a 3D printed chassis is suitable for vacuum tube projects. The answer so far is, yes, at least for low power stuff like this. Another objective, perhaps more unique just to me, was to determine if the Red Pitaya was as useful for audio measurement as it is for RF. Again, at least for low power stuff like this, the answer is yes.
Interestingly, the measured results of the 6X5 6AB4 6BQ5 combination is surprisingly close to the results predicted by SE Cad.
Win W5JAG