+86 400-003-5559 CN


Small, Detailed Nixie Clock Build | Hackaday

tagssmd transistor

Although digital tubes were built in the vacuum tube era in the mid-20th century, they still exist among hackers. Because of the many quirks, getting them up and running is a difficult task, so getting the entire clock to work with the Nixie tube is a badge of honor for those who try the project. For anyone who wants to try, [Tomasz] wrote

 This should help.

There are many in-depth theories behind the Nixie tube on the [Tomasz] page, which he will involve in describing the clock. As far as the actual project is concerned, this is a simplified design, the whole clock uses one board, including the lamp, driver, microcontroller, power supply and DC/DC conversion circuit. This achieved his goal of making the project as small as possible. The Nixies he chose are IN-12, which is popular in his Eastern European home, but can be purchased from eBay and shipped anywhere in the world.

There are many documents on the project site, including schematics, microcontroller code, PCB design, and even oscilloscope screenshots at various points in the circuit. Although this may not be

, It's definitely closed, easier to read, and the most detailed version we've seen in a while!

The high-end drive structure is not good. Really, really bad. The current is only determined by the 1M resistor and the h21e of the transistor. When you make a high-end driver for digital devices, it is recommended to use resistors in the emitter of the PNP and better design the resistor divider. In this way, the driver acts as a current source.

According to the schematic diagram, about 1/4 of the time is provided to each digit about 13mA. This means an average of 3.25mA. It's perfectly possible, but other readers may use a different PNP from KST92-assuming the transistor h21e is 300. This means that each digit is 10mA. Goodbye, Knicks!

If the number is too bright, you can dial back the voltage. The load current from the ground-side driver may be sensed and used in the feedback loop.

Or release the anti-blue light resistor! It does the same thing, you don't need any ground sensing.

Even a collector of several thousand ohms will make the life of a digital camera longer.

You can of course use emitter resistors and provide constant current, but not in the ratio of these voltage dividers. You only need to reduce a few volts on the base.

There are good reasons for using constant current-to balance the overall brightness of the different numbers.

This is why the transistor should be used in the specifications, not just hope that the replacement will work. People who just slap parts just because they don't know the gain or other properties are asking trouble. The design obviously works. If changes are to be made, research and necessary adjustments are required before implementing these changes.

I happen to be the author of this project, yes, you are right. I unwisely connected the anode directly to a pnp transistor without a current limiting resistor, especially for devices that exhibit similar hysteresis (here attributed to neon ignition voltage)

I even placed these resistors on the schematic (collector side), but marked them as 0 ohms, just because I don't know what value will make Nixies look best. I am sorry to mislead you in this regard. In fact, I am using a 10k resistor that provides negative feedback: a higher current will result in a greater voltage drop.

According to what happened to the static voltage divider on the pnp base: it did the job well (I multiplexed it at 400Hz) while consuming very little current from the high-voltage rail, which further reduced the overall Current consumption. no problems found.

kind regards.

I'm glad that other people don't like the inefficient and bulky power supplies on the Internet, but design them themselves. Use MOSFET drivers, design circuits and mathematical operations, and perform simulations in LTSpice for extra points.

Then use 1N4148 as a high-voltage rectifier. what

He did this for efficiency reasons, because conventional rectifier diodes are too slow for SMPS (large capacitance and long recovery time).

However, he could have used 200V switching diodes instead of using 75V diodes in series at 130V, but at the same time there is no voltage divider.

I use the MURS140 (400V 1A) ultra-fast rectifier for the digital power supply because I got it stripped. They have a recovery time of 50ns, less than 10pF, and are good enough.

Yes, these voltages cannot exceed 50V.

Of course, 1N4007 would be a better choice here, and those within the rated value are good for 1KV.

for reference only:

>The total time between the time when the diode is reverse biased and the time when the current actually stops is 3 μs.

I saw a data sheet that said 30us.

In simple English: When the input changes polarity, the diode will not turn off, and the current will leak for at least a period of time. This is a "pure type" rectifier diode used for AC rectification. It is simply not suitable for switch mode power supplies.

This is why there are different types of diodes: switching diodes, ultra-fast recovery diodes, Schottky diodes, etc., are designed for different applications.

I happen to be the author of this project.

I used two CD4148 diodes with DC components in series (

), it provides me with a total reverse voltage of 150V (I used 20V above 130V), which seems to be a very embarrassing thing, but in this case it is actually very beneficial.

Remember, I am dealing with small inductances (many designs use inductors like 330uH, I use 15uH), which means high-frequency short current pulses are used, so the reverse recovery time plays a major role here effect

(1N4148 has 4ns reverse recovery). I measured the reverse current when the output capacitor was charged to 130V, just to make sure that the clock lasted until the end of the day, and no obvious leakage (~3uA) was observed.

kind regards,


Does anyone know what kind of software he might use to draw circuit diagrams?

Schematic/PCB layout software or LTSpice for simulation?

He uses Altium Designer for schematics and PCB layout.

"It must be remembered that multiplexing comes at the expense of display intensity (or brightness), because the power-on time of a single Nixie is only 1/4 of the entire cycle. However, it can be compensated by increasing the supply voltage."

Is this pipe safe? I have used a Nixie clock before, and I often want to increase the power supply voltage to compensate for the dimmer display caused by multiplexing, but I have always worried that exceeding 10-20 volts in the specification will shorten the service life or reduce the voltage. Damaged tube (very important when running 24/7). Does anyone know this better than me?

Certain scrolling or animation processing will help prevent burning hours. You can adjust the brightness of the clock according to the time of day to extend its life. If the power supply is not software controllable, it can run 1/6 cycle or higher late at night (ie pretend to have more numbers).

Please be kind and respectful to help make the comment section great. (


The site uses Akismet to reduce spam.


By using our website and services, you expressly agree to our placement of performance, functionality and advertising cookies.