Warm Tube Multivibrator
Interest in tube technology arises either from audiophileism, or from attempts to restore an grandfather’s radio, or simply from curiosity about retro electronics. Teaching a basic electronics course at a local school, I inevitably come across questions on this topic – and at some point I decided to make such an educational craft – in order, first of all, to better understand it myself. After all, in our age of controllers and digital technology, not so many people know how to use analog technology, and even more so how to use tube technology.
So here is a multivibrator based on a 6N2P double triode and neon lights (well, it’s ugly to install LEDs in tube equipment) – with transformerless power supply. Now I’ll quickly tell you what’s what 🙂
Let’s make a reservation right away – on neon cards, due to nonlinearity with a “negative differential resistance”, you can make a winking multivibrator by adding only a capacitor and a couple of resistors – but on the one hand, such a circuit has some limitations – on the other hand, this is not our goal 🙂
Multivibrator – a type of generator – a circuit with two states that constantly switch between each other. Invented about 100 years ago, it is found in all sorts of books in the form of a recognizable symmetrical circuit on a pair of transistors. My task is generally simple – to replace the transistors with lamps, taking into account their characteristics. And somehow this matter needs to be addressed.
Its work “on the fingers” is that when one of the transistors is open (and the corresponding LED is shining), it turns off the transistor of the other half. But this is done through a capacitor on which the charge slowly changes, so that at some point the second half opens and locks the first. Here is an animated picture with transistors represented as switches:
And now – Lamps!
Or rather, one lamp is a double triode. As the name suggests, it contains two amplification elements similar to transistors. With transistors, however, there are significant differences (except for the high anode voltage):
current flows into the base of the transistor to open it – and the triode grid is controlled by voltage (we can assume that it has infinite resistance and nothing flows there)
on the base of the transistor in the operating (open) state there should be a potential slightly higher than on the emitter – and the lamp, on the contrary, has an open state by default – a small voltage below the emitter must be applied to the grid in order to stop the electron current.
With these theoretical ideas, I got a little wiser and set out to construct a model – for authenticity, by mounting it “on stiletto heels” – in the role of which were small nails. I was lucky – my ideas turned out to be correct and the scheme worked. Here's what it looks like in its entirety.
Sorry for the captions in English – initially I showed this to my friends on the bourgeois forum – but I drew the elements in our language 🙂
The multivibrator itself is on the bottom left, around a double triode that looks like an intricate “smiley face” – as you can see, the main change in comparison with the transistor circuit is that the resistors on the control electrodes are pulled down (towards minus) and not up – to lock the triodes and not open them their.
In order for the minimum voltage on the grids to be slightly lower than the cathode voltage, the cathodes need to be “supported” with resistors – then the current passing through the lamp will create a small voltage drop across them, “raising” the cathodes “above” the potential of the negative wire to which the grids are connected. As you can see here, there is one such “supporting” resistor for both cathodes – and a capacitor is paralleled with it to reduce the influence of changes in the anode current during operation of the multivibrator on the level of this support.
Above the triode is the part with the neon lights. Everything is clear here, typical limiting resistors – in neon lights they have a higher resistance than in LEDs – but the principle is the same. A neon in working condition has a stable voltage drop of about 60V – and you need to burn off all the excess on the resistor, otherwise it will “boom” as Piglet would say 🙂
Why I added two resistors in parallel with the neon wires – I’ve already forgotten – in my opinion, the circuit won’t start without them – we need the potential of the anodes to be close to that on the positive wire when both triodes are closed – and here the neon wires lose 90V in the closed state.
Nutrition
This part is no less interesting. Typically, lamp technology is powered through a hefty transformer – which has a pair of secondary windings – low-voltage and powerful for filament – and also a relatively low-current high-voltage – for the anode circuits. Such a trans weighs a kilogram or several – and in a small diagram it looks unattractive (not to mention the fact that this is a valuable thing, not a penny).
Therefore, let's ask directly. With the anode circuit, everything is not complicated – we took a diode (why not a bridge? I’ll tell you now) – and added a two-stage RC filter – and here you have something like 300 volts at the output. For the capacitors, don’t forget the “discharging” resistors – it won’t be good if you get a current from the circuit that is already turned off.
The incandescent circuit is more complicated – it has a relatively high current (hundreds of milliamps for this lamp) – but a low voltage (about 6 volts). You could install a fat resistor that would “eat up” all the extra 200-odd volts. It’s not hard to imagine that the power of such a resistor will be something between a soldering iron and a boiler 🙂
Therefore, instead of a resistor there is a capacitor. It creates “reactive” resistance to alternating current. The capacitance can be calculated using the formula for the reactance of the capacitor (it depends on the frequency) based on the desired current, according to Ohm’s law (we simply ignore the voltage drop across the filament itself). In general, it is very convenient – including for all kinds of battery charging and so on – it is only important that the capacitor must be film (not electrolyte) and therefore quite healthy (the more, of course, the greater the current is needed). And the discharging rezuk will also be useful to him.
And here lies another nuance. A filament circuit with a capacitor as a resistor – it is connected directly between the AC voltage wires. The rectifier circuit for the anode voltage is also between them. At the same time, we need the minus of the anode voltage to be on the same wire to which the filament fits!
If we had inserted a capacitor not to the right but to the left of the filament circuit (according to the diagram) – or if we had used a bridge rectifier circuit – the filament and the minus circuit (with the cathode) would not have been directly connected – and between them in each period of the mains voltage there would have been potential difference of hundreds of volts. Because of which the filament itself would become an additional anode 🙂 Of course, it will not work exactly like the main anode (due to heating and due to the lack of coating on the inside of the cathode – but it can still, depending on the circuit, be great ruining life. In a multivibrator this would perhaps be less critical – but in some amplifier it would be quite so.
Demo
That's basically all
If you decide to repeat this craft, of course, do not forget about safety measures (we all know how easy it is to miss something in a hurry) – when turning it on, work with only one hand, if possible with a glove. Use tools with insulated handles. Do not tamper with anything in the circuit while it is turned on.
