logo elektroda
logo elektroda
X
logo elektroda
Dostępna jest polska wersja

Czy wolisz polską wersję strony elektroda?

Nie, dziękuję Przekieruj mnie tam

TL;DR

  • Zbudowano przenośny tester lamp Nixie zasilany z USB do sprawdzania LC-513 i LC-516 przed zakupem.
  • ATmega8 steruje katodami i pracuje też jako przetwornica boost, podnosząc 5V z USB do wysokiego napięcia.
  • Przetwornica pracuje z PWM 31.2 kHz, wypełnieniem 85% i osiąga około 180V przy sprawności prądowej około 1.2 mA.
  • Tester pozwala automatycznie lub ręcznie przełączać cyfry, a po podłączeniu do PC odczytuje napięcie i wykrywa uszkodzone lub nadmiernie prądowe katody.
  • Pierwsze uruchomienie starych lamp często wygląda słabo przez cathode poisoning, a niektóre oprawki, zwłaszcza dla LC-516, są mechanicznie niewygodne.
Generated by the language model.
ADVERTISEMENT
Treść została przetłumaczona polish » english Zobacz oryginalną wersję tematu
📢 Listen (AI):
  • Two Nixie tube testers with mounted tubes and a matchbox for size comparison.

    Motivation
    8 years after mine last project I once again had the urge to build another Nixie clock. And as we know, lamps are an essential element of every such clock.
    They no longer produce such lamps (except for a few cases of hobbyists who bend numbers from wires in the privacy of their homes, encasing the whole thing in hot glass. So the only way is to buy some old copies from private people.

    Purchasing used electronic components on aliexpress taught me that you should never trust and always check (test). The same may probably be true with lamps, but surprisingly these are easiest to get not in the Far East, but in our closer eastern neighbors (Ukraine, Russia). So I decided to build a handy lamp tester, powered by USB (!), so that I could, for example, connect it to a power bank and take it with me to a flea market before buying lamps.

    The most popular in our country are probably the lamps from the domestic company DOLAM: LC-513 and LC-516. The latter are basically the same as the former, but covered with resin in the plinth. I heard that some people have the ability to cut this ebonite "filth" and remove the lamp. But in the "clothes" both lamps have different pinout, so I decided to create two testers.

    Description of the device
    The whole thing is managed by the only integrated circuit on board: Atmega 8. And it`s not boring at all, because in addition to controlling the cathodes, it also serves as a voltage converter. And it`s not easy - increasing 5V from USB to 180V was implemented in the form of a simple Boost converter. PWM frequency is 31.2 kHz, duty cycle is 85%. The increased output voltage is monitored by the microcontroller with one of the ADC inputs, thanks to which the feedback keeps the output voltage at a set level.
    Although I tested even without feedback and in this system I was not able to get more than about 180V (without load). Maximum current efficiency approx. 1.2 mA.
    Electrical schematic of a Boost converter for powering Nixie tubes.

    The microcontroller is clocked with an external 16 MHz crystal for USB purposes - because the system uses USB not only for power, but also has the ability to communicate with a computer. Using a PC, you can, for example, read the voltage produced by the converter when powering various cathodes, so you can detect, without looking, which of the cathodes have any problems (e.g. they do not light up at all or consume excessive current).

    There is also a button - by default, when turned on, the device displays each digit in a circle. However, by pressing the button you can change the control to manual, thus keeping a given digit longer.

    Circuit diagram of the Nixie tube tester LC-516 with Atmega 8 microcontroller

    As a socket for the LC-513 lamps, I used extended connectors from the DB socket. They work perfectly.
    Nixie tube tester LC-513 on a finished PCB Nixie tube tester with visible circuits and connections. Nixie tube tester LC-513 on a wooden background PCB design for a Nixie tube tester labeled LC513 with SMD components. Nixie tube tester connected to a power bank via USB.

    DownI used a dedicated stand for the LC-516 lamp, but I am not satisfied with it - quite a lot of force is needed to place the lamp in it.
    Nixie tube tester module with a tube mounted on a circuit board. Prototype of a Nixie tube tester with PCB and integrated circuit Nixie tube tester connected to a power bank via USB.

    Fortunately, there are also single pins with a thicker diameter on Aliexpress, which I will probably replace in the future:
    Set of metal pins with a gold-plated finish.
    https://pl.aliexpress.com/item/1005004071932217.html

    And finally, a comparison of both testers - it is basically an identical design, the difference is only in the upper part of the board on which the tested lamp is placed. Thanks to the use of SMD elements, the device is very small and handy. Some more casing would be useful...
    Two Nixie tube testers on a wooden surface, with different bases for the tubes. Two Nixie tube testers next to a box of matches.

    As an interesting fact, it took me half a day to build the second one - the microcontroller was unable to generate high voltage, it reset itself if I changed the fill factor in the interrupt function. After replacing it with another identical chip, the problem disappeared.
    Heh, now it`s time to create an atmeg tester from aliexpress.




    I noticed that most of the lamps, when turned on for the first time after many years, glow "sluggishly" - some of the digits seem to have missing fragments. Only after a dozen or so seconds everything starts to "normalize" and the lamp, put aside and then tested, e.g. after an hour, works normally. So it is not some form of "heating up", but rather a phenomenon known as cathode poisoning - some oxides are deposited on the cathodes, which are "dusted" back from them only after some time. Apparently, when displaying digits on such lamps, it is worth displaying all the digits one by one from time to time to eliminate this phenomenon.


    Attachments:
    • nixie-tester.zip (108.31 KB) You must be logged in to download this attachment.

    Cool? Ranking DIY
    About Author
    phanick
    Level 28  
    Offline 
    phanick wrote 2947 posts with rating 2833, helped 65 times. Live in city Warszawa. Been with us since 2007 year.
  • ADVERTISEMENT
  • #2 20839880
    behru
    Level 12  
    Cool project - I would love to see a version in the form of one board and plug-in adapters depending on the lamp being tested

    phanick wrote:
    They don`t produce such lamps anymore (except for a few cases of hobbyists who bend numbers from wires at home and enclose the whole thing in hot glass. So the only way is to buy some old copies from private people.


    I must disagree with this statement ;-)
    Below are the website addresses of two manufacturers (not garage manufacturers) who turned their hobby into a full-fledged production of nixie tubes. You can buy both lamps and clocks from them

    https://millclock.com/nixie-tubes
    https://www.daliborfarny.com/
  • ADVERTISEMENT
  • #3 20840018
    gulson
    System Administrator
    Thanks for sharing your project. If you send me a parcel locker, I will give you a small gift!
  • ADVERTISEMENT
  • #4 20841766
    CMS
    VIP Meritorious for electroda.pl
    phanick wrote:
    I noticed that most of the lamps, when turned on for the first time after many years, glow "sluggishly" - some of the digits seem to have missing fragments. Only after a dozen or so seconds everything starts to "normalize" and the lamp, put aside and then tested, e.g. after an hour, works normally.


    An interesting phenomenon. But where do the oxides come from if it`s a vacuum tube? I think it`s something else. Maybe my colleague @AlekZ will say something more about this interesting topic.
  • #5 20842287
    phanick
    Level 28  
    behru wrote:
    Cool project - I would love to see a version in the form of one board and plug-in adapters depending on the lamp being tested

    The current appearance of the project does not exclude the use of adapters - they will have male DB contacts in the shape of a circle (like contacts in a lamp), instead of "goldpins".

    People designing their clock are probably wondering how to control several lamps at once - whether to use `multiplexing` or to control the segments of each lamp separately.

    Multiplesing saves the number of driver I/O pins and the number of transistors. The disadvantage is the lower brightness of the light, because with 6 lamps each lamp shines only for 1/6 of its time.

    "Separate" control means, above all, greater brightness (all lamps are lit at the same time) but also the need to build a "more powerful" voltage converter to absorb 6 times the current consumption.

    The issue of lamp life is also interesting - does multiplexing shorten their life or not?

    I was most concerned about this reduction in brightness and I decided to do a small test - to check how the lamp would behave when it was on for 1/1, 1/2, 1/3, 1/4, 1/5 and 1/6 of its time.
    When the lamp is lit for shorter and shorter periods of time, the brightness decreases only a little (certainly not 6 times). The numbers are a little thinner, but that strange purple glow in the background disappears.



  • ADVERTISEMENT
  • #6 20843809
    pawelr98
    Level 39  
    phanick wrote:
    And as we know, lamps are an essential element of every such clock.
    They no longer produce such lamps (except for a few cases of hobbyists who bend numbers from wires in the privacy of their homes, encasing the whole thing in hot glass. So the only way is to buy some old copies from private people.


    Not true.

    Dalibor Farny, and additionally a milclock factory in Ukraine.
    The former makes new models of nixie tubes from scratch. This second factory only produces one type, the modernized IN-18 as the ZIN-18.
    You know, it`s cheaper in Ukraine - one glass costs $99.
    Which is not an excessive price, because IN-18 are huge nixie tubes that fetch PLN 200-400/piece on the secondary market. ~PLN 400 for a brand new Nixie is not detached from market prices in any case.

    https://www.daliborfarny.com/
    https://millclock.com/
  • #7 20844330
    CMS
    VIP Meritorious for electroda.pl
    We`re talking chatter here, and Tiktok gave me this beauty
  • #8 20845940
    behru
    Level 12  
    pawelr98 wrote:


    Not true.
    Dalibor Farny, and additionally a milclock factory in Ukraine.
    The former makes new models of nixie tubes from scratch. This second factory only produces one type, the modernized IN-18 as the ZIN-18.
    You know, it`s cheaper in Ukraine - one glass costs $99.
    Which is not an excessive price, because IN-18 are huge nixie tubes that fetch PLN 200-400/piece on the secondary market. ~PLN 400 for a brand new Nixie is not detached from market prices in any case.

    https://www.daliborfarny.com/
    https://millclock.com/


    Well, I wrote it in the first post, which was a comment on the author`s entry :D
  • #9 20845980
    cirrostrato
    Level 38  
    pawelr98 wrote:
    IN-18 are huge nixie tubes that sell for PLN 200-400/piece on the secondary market. ~PLN 400 for a brand new Nixie is not detached from market prices in any case.
    I had a lot of new Z573M (not as big as IN-18), I sold one collective package (200 pieces) with difficulty two years ago for PLN 20/piece, now even PLN 35/piece for these NEW IN PERFECT CONDITION NIX, for many people it is an insurmountable obstacle. don`t overdo it. The exception is IN-14, even stimulants reach incredible prices and I don`t understand it.
  • #10 20856317
    jaeger
    Level 14  
    CMS wrote:
    phanick wrote:
    I noticed that most of the lamps, when turned on for the first time after many years, glow "sluggishly" - some of the digits seem to have missing fragments. Only after a dozen or so seconds everything starts to "normalize" and the lamp, put aside and then tested, e.g. after an hour, works normally.


    An interesting phenomenon. But where do the oxides come from if it`s a vacuum tube? I think it`s something else. Maybe my colleague @AlekZ will say something more about this interesting topic.


    This is a gas lamp, a neon lamp to be precise.

    phanick wrote:


    I was most concerned about this reduction in brightness and I decided to do a small test - to check how the lamp would behave when it was on for 1/1, 1/2, 1/3, 1/4, 1/5 and 1/6 of its time.
    When the lamp is lit for shorter and shorter periods of time, the brightness decreases only a little (certainly not 6 times). The numbers are a little thinner, but that strange purple glow in the background disappears.


    Perhaps it would be enough to reduce the lamp current, e.g. by increasing the value of the anode resistors?
    Reducing the voltage will not necessarily be a good idea because the neon lamp will not light up if the voltage is too low....

    I found a nice website about nixie tubes. There are also diagrams of lamp power supplies here: https://threeneurons.wordpress.com/nixie-power-supply/

    I have some stands for the DL-531. Unfortunately, they are a bit tarnished :-(

    Four tarnished DL-531 vacuum tube sockets on a wooden surface.
📢 Listen (AI):

Topic summary

✨ The discussion revolves around the creation of a portable USB-powered Nixie tube tester, aimed at facilitating the testing of vintage Nixie tubes before purchase. Users express interest in the project, suggesting the inclusion of plug-in adapters for different lamp types. The conversation highlights the availability of new Nixie tubes from manufacturers like Dalibor Farny and Millclock, countering the notion that such tubes are no longer produced. Technical considerations regarding lamp control methods, such as multiplexing versus separate control, are discussed, along with the implications for brightness and lamp longevity. Additionally, the conversation touches on the characteristics of gas and neon lamps, and the importance of proper voltage and current management for optimal performance.
Generated by the language model.

FAQ

TL;DR: This palm-size tester boosts 5 V USB to 180 V @ 1.2 mA, lets you "detect which cathodes have any problems"[Elektroda, phanick, post #20839537], and even cycles all digits automatically. Built units show multiplexing cuts light output far less than 6×. Why it matters: Quickly screens vintage Nixie tubes so you avoid paying for faulty lamps.

Quick Facts

• Power input: 5 V USB (standard) [Elektroda, phanick, post #20839537] • Boost output: 160–180 V DC regulated, ≤1.2 mA [Elektroda, phanick, post #20839537] • PWM converter: 31.2 kHz, 85 % duty cycle [Elektroda, phanick, post #20839537] • Supported sockets: LC-513 & LC-516; others via plug-in adapters [Elektroda, phanick, post #20842287] • New IN-18 Nixie price: ≈US$99 each [Millclock, 2023]

What does the portable USB Nixie tube tester actually do?

It powers a single Nixie lamp from a 5 V USB source, steps the voltage up to about 180 V, and sequentially lights each cathode so you can see whether any segment is dim, missing, or shorted [Elektroda, phanick, post #20839537]

How is 180 V generated from just 5 V?

An ATmega8 microcontroller drives a boost converter at 31.2 kHz with an 85 % duty cycle. Feedback from the ADC keeps the output near 180 V under load [Elektroda, phanick, post #20839537]

Which Nixie tubes are supported out of the box?

Two PCB versions exist: one for LC-513 and one for LC-516. Both share the same control electronics; only the top socket section differs [Elektroda, phanick, post #20839537]

Can I adapt the tester to other tube types like IN-14 or IN-18?

Yes. Design your adapter with male DB-style pins that mimic the lamp’s footprint, then plug it into the board’s female contacts [Elektroda, phanick, post #20842287]

What are the voltage and current limits?

The booster reliably reaches 160–180 V open-circuit and can supply about 1.2 mA; this is adequate for a single tube but not a six-digit clock [Elektroda, phanick, post #20839537]

Why do some digits look incomplete for several seconds after years of storage?

Unused tubes suffer cathode poisoning: oxide films increase striking voltage, so parts of a digit don’t light until ions clean the surface [Elektroda, phanick, post #20839537] Cycling all digits periodically limits the effect [jaeger, #20856317].

Does multiplexing cut brightness dramatically?

Author tests show reducing duty from 100 % to 16 % (1⁄6) makes digits only slightly thinner; brightness drop is far <6× and purple background glow vanishes [Elektroda, phanick, post #20842287]

How can I dim the lamp without multiplexing artifacts?

Increase the anode resistor to lower current. Maintain striking voltage; dropping converter voltage too far can stop ignition [jaeger, #20856317].

What if the microcontroller keeps resetting and no high voltage appears?

A faulty ATmega8 can’t sustain the PWM load; replacing the chip fixed the issue in one build — an edge case worth noting [Elektroda, phanick, post #20839537]

Are brand-new Nixie tubes still made?

Yes. Dalibor Farny (Czech Republic) and Millclock (Ukraine) produce new glass; Millclock’s ZIN-18 sells for about US$99 [Millclock, 2023][Elektroda, behru, post #20839880]

Why do common Z573M tubes cost far less than IN-18?

Z573M bulbs are smaller and plentiful; a 200-piece lot sold for PLN 20 each, whereas large IN-18 tubes are scarce and fetch PLN 200–400 on the secondary market [Elektroda, cirrostrato, post #20845980]

How do I safely test a lamp with this device?

  1. Plug tester into a 5 V USB power bank.
  2. Insert the tube gently into the socket.
  3. Observe each auto-cycled digit; press the button for manual hold if needed [Elektroda, phanick, post #20839537]

What safety precautions are needed around 180 V?

Treat the board as live: avoid touching anode wiring, insulate exposed pins, and unplug USB before swapping tubes. 180 V can deliver painful shocks despite low current — stay cautious [“ThreeNeurons Nixie Safety”].

Where can I find schematics or firmware?

The author’s screenshots and PCB photos are in post #20839537. The boost-converter theory matches open designs on ThreeNeurons’ site for reference [Elektroda, phanick, post #20839537][“ThreeNeurons Nixie power supply”].

What if the converter stalls below 160 V?

Check the inductor value, diode speed, and PWM duty. Excess load or wrong component values can cap the voltage; lighten the load or correct parts and the feedback loop will reach 180 V [Elektroda, phanick, post #20839537]
Generated by the language model.
ADVERTISEMENT