|
Clone Information
|
See http://sensi.org/~svo/zloshnik/index.en.html by svofski At first I was going to build a CRT clock using this tube. But after learning that the 3??1? has a very short lifetime (~1000 hours), I changed my mind. Zloshnik is not trying to be a clock. Instead it's a decorative piece, something that you can fire up once in a while and enjoy its looks. It has a serial port which allows its use as a vector terminal, but at the moment this function is not used. Circuit Description To power a CRT, even as small as this one, one needs pretty high voltages. To power a Nixie clock 150-180 volts are enough, and this is easy to make with a boost converter. For the 3LO1I the difference of potential between the cathode and the second anode has to be not less than 750 volts. This is why the power circuit differs significantly from that of a Nixie clock. In a lot of homemade CRT projects the high voltages are produced using a power transformer from 220V, which makes them huge and heavy. For such a small tube I wanted to use a single source of +12VDC, which can be taken from any wall adapter. The first part of voltage conversion is done with a push-pull cascade comprised by two MOSFETs IRF630N that drive a high-frequency transformer. The transformer is wound on a ferrite ring. The primary uses bifilar winding, 2×6 windings. A bifilar winding ensures high symmetry, and it is important for a push-pull transformer circuit to work efficiently (or to work at all, since asymmetry here may kill the transistors). There is only one secondary, without a central tap. The number of windings on the secondary was found experimentally. With a 12V input there is approximately 150VAC on the secondary, or ±300Vp-p. There is another secondary with just one winding that provides power for the cathode heater. The heater must be floating without reference to the ground to avoid the potential difference between itself and the cathode. To avoid short circuit while the heater is cold, it is connected via a 0.5O resistor. The voltage from the secondary is doubled with a voltage doubler made with C16, D9, D10, and C17 and smoothed with R6, R8, R9, and C7, which gives us +300V. Similarly, the tripler made with C1, D1, D2, D3, C2, and C3 provides -450V. It is convenient to have voltages like this relative to the ground because this way we can use pretty common transistors with Uce=300V in the deflection system. Between +300V and -450V there is a resistor ladder that provides all voltages needed to make a CRT work. The deflection system is made with two differential pairs of the "long-tailed" kind, formed by transistors Q1, Q3 and Q5, Q6. Tuning the amplifiers to make them really symmetric was the most fiddly part of the project. The blanking circuit is formed on transistor Q6. It allows blocking of the ray for a short period of time by changing the voltage on the grid. The deflection signals are formed by the TL5626D DAC. The blanking signal comes directly from the microcontroller. The ATmega8 microcontroller does two important things. The first is to open and close the MOSFETs of the push-pull circuits. This is done by Timer1 and output compare unit OC1A/B, and once set up this does not require any attention. The other thing that occupies the controller is the formation of coordinates X,Y for the DAC and the blanking signal Z. |