Soviet lamp controllers keep all digits charged at 65% of their glow voltage. This makes transitions faster

@NotoriousPyro

4 жыл бұрын

In Soviet Russia, nixie tube charges lamp controller.

@blitzroehre1807

4 жыл бұрын

Didnt the western version IC (7441 or 74141) do this also?

@user-lf8tx7zq5i

4 жыл бұрын

To be exact at 66.6%.

@masox755

4 жыл бұрын

He is not russian, he is czech and speaks with hard czech accent

@dan8t669

4 жыл бұрын

@@masox755 such a smart comment

Answer to question in title: 22:19

@kamonic6571

4 жыл бұрын

Thank you. 26 minute video for a 15 second answer

@ningxin0702

4 жыл бұрын

thanks XD

@coachreenasharma

4 жыл бұрын

Thanks man I appreciate it

@firefly2472

4 жыл бұрын

You are the man man.

@kenan0

4 жыл бұрын

careful he is a hero

Keep alive voltages will reduce transition times between glow and not. Individual adjusted to each tube at a 65-70% transition voltage

@isettech

4 жыл бұрын

Keep alive voltages are used in Plasma TV for lighting pixels at low intensity where pixels may not strike reliably. This reduces the glow area when the pixel is dark and under a mask, which permits re-establishing glow at low currents without needing the elevated voltages to produce break over ionization. This is not used in Nixie tubes.

@isettech

4 жыл бұрын

@@aphenioxPDWtechnology Where are they still using gas lasers for communications? I thought all this went solid state

@SebastianM93

4 жыл бұрын

@@isettech i think this is because a plasma screen has isolated chambers for each pixel. In a nixi tube all cathodes share the same gas so if there was a cathode active, ions/electrons can move to the next position and this cathode can strike even faster ;)

@cambridgemart2075

3 жыл бұрын

@@isettech They don't, but diode lasers have to be kept biassed just below threshold to give fast switching.

@isettech

3 жыл бұрын

@@cambridgemart2075 This is mostly due to the capacitance of the junction. The junction already at threshold will respond to sudden changes in supply current. The capacitance if the diode is fully off, will use some current to charge the capacitance at the junction before emission.

Работа конечно проведена громадная. Заслуживает докторской диссертации . Молодцы. Сам начал увлекаться сборкой часов на неоновых индикаторах , благо запасов Советского Союза хватит на долго. Всем Мир, берегите себя и своих близких.

In case anyone is wondering (as I was at first till I paused the video and looked it up) why the gas behaviour matters for all this, it's because Nixie tubes are cold cathode tubes, i.e. the glow is from the de-excitation of the ionised gas in the tube and not incandescence of the cathode filament as is the case for vacuum tubes.

This is very cool! I hope you are plaining on making more videos like this.

@daliborfarny

4 жыл бұрын

I am, thanks for nice response! :-)

@imtrash1228

4 жыл бұрын

You can kill cancer cells with that much Hertz

@Myrtone

4 жыл бұрын

@@daliborfarny Before you do, could you please make a trial video in Czech with English subtitles to see if your English-speaking viewers would still watch it?

Awesome presentation and a great science lesson Dalibor. Very easy to understand. I really appreciated the classic chalkboard over a whiteboard, instruction. Whomever is purchasing this timer is going to have a truly great device to last for ages.

Excellent video. A perfect example of how quick simple things can get complicated when you start to close! "Oh.. we'll just hit it with a high speed camera" 6-months later... "wow... that was awsome!"

Wow. Incredible stuff. Thanks for sharing!

This is amazing! Absolutely astonishing and fascinating! The beauty if the products, the science, the exploration of a technology that is both very old and cutting-edge new! I am so happy to be just a mere follower of this channel, hope you all are having an absolute paradise time in your castle!

While watching this video, I checked the Subscribe button to confirm that I have subscribed to this amazing channel. Thank you for this detailed video.

Thank you Dalibor! I felt like I was back in the University sitting in class. I understood the need for a bias to speed up the display turn-on but did not realize how significant the difference was between cold start and regular operation. There is not much that can be done to extinguish the glow after power is removed. I suspect that the mix of gas and the concentration/pressure of the gas will play a big part in the duration of the "afterglow". Even with those limitations, the nixie performance at 100,000 Hz is quite acceptable. Excellent video explanation and confirmation of performance.

@daliborfarny

4 жыл бұрын

Hi Jeff, thank you - very precise conclusion. We can use our regular nixie tubes because they perform well enough for this application. The gas mixture and pressure is the only how to influence the afterglow and charge - this was a topic for dekatrons in the past, the old dekatron engineers played a lot with gas mixtures.

@nonix3470

2 жыл бұрын

@@daliborfarny Hi Dalibor, not that I am an expert, would it be possible for fraction of time to reverse polarity just after switch off? Not the the old way up to 100% but only about 50-60%. Would that help to limit the after glow? With today's electornics it should be achievable, IMHO. Whether it would help or not, I don't know.

@Dalbor Farný Thanks for taking the time to make and share these fantastic videos Best regards.

I love these. Amazing video, I'm so glad this channel showed up in my recommendations.

Skvělá práce. Je fajn, že to umíte i dobře vysvětlit. Moc Vám fandím. :-)

Thanks for sharing this, Dalibor! Super interesting. I love seeing the deep-dive into the small details.

@ahmedalshalchi

4 жыл бұрын

The problem is when you deep-dive to pick something and float up to surface you will find a Chinese diver is selling at one quarter of that mission cost , what to do then ??

I think the Afterglow you're looking at in the high-speed camera footage is actually an after image on the camera sensor itself.

Was waiting for this video ! Thanks a lot for sharing ! I wish all the best to this wonderful piece of engineering :)

To get the after-glow disappearing faster: Couldn't you just apply the 0V to the digit (alternatively apply VDD to the grid) for a short time before displaying the next digit? This would almost instantly de-ionize the Neon atoms. Some researchers at my university did this to use it in "benchmarking" different nixie tube manufacturers and time-efficiencies of different controllers, I believe.

This is great, thanks for sharing! And congrats to the amazing project!

Dalibor, this is some of the coolest stuff on the interwebs. I remember watching one of your earlier videos on making your fantastic tubes. Nixie tubes have always fascinated me for some reason. When I was in electronics school long ago my final project for one of my classes was a nixie tube clock. Pretty much all discrete 74xx TTL electronics, only saving grace was the availability of a TTL Nixie tube driver. Timing was off line current. It's 60 hz here in the states so it was sort of accurate for a while... I still have it laying around somewhere. All this makes me want to dig it out and play with it again. If memory serves, I used some old used IN14 tubes. This video really "went into the weeds", I loved every second of it.

Amazing work, a mixture of homemade top-notch technology crossed with state-of-the-art devices... Big thumbs up!!!

BTW: an LED can be operated extremely fast if you don't just switch it off but set ground or even a negative potential (not to close to avalanche) to its anode to empty the pn transition region from charge carriers within a several ten nano seconds. If you have opto couplers with a base connection to its photo transistor connected across a resitor to ground you can operate them at frequencies of several 10 MHz.

absolutely brilliant stuff! :) Have you thought about bipolar custom waveform driving of the tubes instead of square wave? This would surely make the electronics more complicated but then you can use negative voltage to turn digits off quickly and controlled overshoot to turn them on quickly. This is commonly used in LCD technology to reduce response time.

@cambridgemart2075

3 жыл бұрын

Yes, definitely the way to go, drive each cathode with a half-bridge (push-pull) stage and keep the cathodes just below threshold voltage. Keep inductance and capacitance as low as possible and have a high current, or more correctly, low impedance, power supply.

Great video! Continuously good work means you're really putting work into these. I love it! :D

Dalibor, thank you so much for this. I enjoyed every minutes of it. Please dont go away from making videos.

Electrically speaking, the elements in the tube are capacitors and hold a very small charge in the small pF range and the wire, socket, etc add capacitance storage. Most Nixie drivers are open collector design designed to simply interrupt the current. To remove the voltage, discharge resistors can be added to the Nixie elements to quickly remove the voltage when the transistors turn off. This may improve your extinguish times. Have you looked at the rise and fall times on your tube elements?

@SebastianM93

4 жыл бұрын

Yes, we did this, but the afterglow because oft lifetime of excited atoms was higher. At 19:21 the cathode voltage is shown. The risetime is about 4 us.

@MakisHMMY

4 жыл бұрын

Are they really Caps though? Because when no voltage is applied, they still discharge. I am now interested to look up the linear equivalent of the elements...

@isettech

4 жыл бұрын

@@MakisHMMY All metal plates, screens, and wires in a tube have capacitance. If you look up old vacuum tube info, this specification is listed as inter electrode capacitance. In amplifiers and transmitters with the high impedance devices, this had to be compensated for to prevent unwanted parasitic oscillation. Values in the Nixie tubes will be in the order of a couple Pico Farad between elements and the Cathode

These delayed ignition and extinction effects give to the display on Nixie tubes a special unique effect. The human eye still does not require such a high speed, and the effect of smooth ignition gives the glow a special softness.

Thank You for sharing all this information! I hawe followed You for some yera now, and it has been a genuinly exciting journy You hawe realy worked hard with all this and deserve the sucess. That castle idea was simply ingenious.

WONDERFUL! Thanks for the detailed analysis

So interesting! Thank you for the video! There is an additional experiment that I would like to see: During your afterglow test, you concluded that the afterglow lasted 200μs on a single digit dissipating, but in the desired application, you have a second digit (cathode) energizing right after the first one and I'm curious how this affects the afterglow--does the second digit attract the ions from the first? Keep up the beautiful work!

Man I love nixie tubes, this video was great

Thanks for such a wonderful video. Before finishing it I was about to suggest rearranging the digits so that consecutive numbers do not overlap (like in half a circle pattern).

I just had a thought about the fast digit switching. What if the "off" digits were driven positive? This would eliminate the afterglow by forcing the ions away from the digit, but also create a physically nearer sink for the electrons, possibly speeding-up the turn-on time.

Amazing video. Thank you so much, you do astonishing stuff. Great!

This is a great presentation. Very nice and informative.

Very good explained principle. Oldies but goldies.

Beautiful work. Bravo sir

Facinating. Excellent work!

4:20 I would call this "delay" a "ramp". I just learned that on custom car builds, some are programming in a ramp & fade into their LED turn signals so they mimic the old incandescent bulbs that warm up & cool off every blink.

Thank you so much for sharing the results of your study here on youtube, Mr Dalibor. Big thumbs up! Ironically though, likely none of those engineers at Burroughs et al who developed these nixie tubes and dekatrons 70 years ago are alive today to actually view the discharge develop and decay in such slow motion. Such cameras werent around back then, but I could imagine the guys sure would have given an arm and a leg to see this excellent video :-)

@RageXBlade

3 жыл бұрын

There were high speed film cameras, but they were enormous, finicky, and very very expensive

Zdravím Dalibore, ani ve snu by mne nenapadlo že to dopracujete až na závod s rychlostí světla :) Parádní video a perfektní práce. Díky a už se těším na další.

@daliborfarny

4 жыл бұрын

Děkuji, další už připravuji :-)

24:45: Because it takes a relatively long time (75us) to get the neon going from a cold start, some tubes had a keep-alive electrode (a kathode outside the visible area), e.g. the panaplex SP-351-354, which enabled leading zero suppression. The keep-alive kathode "reduces reionization time to less than 30us", according to the datasheet.

This discovery is amazing. Just shows how impressive this gas technology is. And thank you for sustaining the nixie tubes :D

"the other side of the country" "3 hours" i cant get to the other side of my relatively thin state in that timespan.

@GGigabiteM

3 жыл бұрын

If you were to be dropped in the middle of Texas, you could drive 5-7 hours in any direction and still be inside the state.

The rise time wasn't an issue, but with a cold tube, potentially the decay time was. If decay is a function of the half-life of an excited neon atom, would changing to a different gas help? Obviously the color would no longer be the classic warm Nixie amber though.

I see most of the electronic solutions to turn-on time have been mentioned already but another thing you can consider is a very weak radioactive source inside the envelope. The radioactivity can affect the gas just as the ionizing field can and this can create interesting effects like easing ignition of the digits and making ignition time more deterministic. I seem to recall it's used in krytron timing tubes for exactly this reason

Hello Dalibor. Great content and by-the-way, greetings from Shannon/Nyquist (double sampling frequency.) Yes, for precisely timed microcontroller code avoid e.g. Arduino libraries and hardcode in assembler, typically in C and/or inline assembler. This allows for faster i.e. more deterministic code/runtime control. However, in very time critical applications I verify (measure) the runtime behavior of the code like you did. Don't blindly trust any assembler/compiler. I wonder, what you can do to reduce the afterglow? It looks like you are using a square wave to "fire up" a digit. Why not using a square wave that is pulsed at a time "close to the end of its falling edge." Yes, I am aware of a dimming effect, but why not playing with the thought of modulated excitement energy? You might also want to try overshooting the voltage when you "fire up" a digit, to speed up the process (boost). In addition, at high switching frequencies capacitance and inductance of wiring, board, glowing hardware etc. can delay current rises and declines. That might be another point for optimization. Keep us posted on your journey :)

Very educational! It was very cool to see the Nixie tubes recorded with high speed camera! But if you would ever need to do another test, there is a simple way to find out how quickly the Nixie tubes switch, without using a high speed camera. Just pulse one digit on and off with some low duty cycle, sense the light output with a photodiode, and look with an oscilloscope at the diode photo-current, which would be proportional to the light signal. This would allow to see and measure the rise and fall times of the Nixie tube light output. Even many cheap (under $1) photodiodes have response time which would be essentially instantaneous for the purposes of such experiment: for example Osram SFH 206 K -- 0.020 microsecond.

@Muonium1

4 жыл бұрын

would a simple photodiode really be sensitive enough to elucidate the afterglow phenomenon, etc. I think maybe they need photomultiplier tube or avalanche pd to do it.

@cogoid

4 жыл бұрын

@@Muonium1 A photomultiplier would of course work nicely! But consider that the image sensor in the videocamera is essentially an array of tiny photodiodes, each a few microns or a few tens of microns across -- one pixel does not collect much light at all. If they can produce an image with an accceptable signal to noise ratio, as we have seen in this video, then it should be possible to do OK with the much larger discrete photodiode. (The photodiode that I have mentioned earlier has 7 square millimeters area -- it is huge comparing to the a pixels in the camera. From the datasheet, its noise floor is 40 picoWatts @ 1 MHz bandwidth.)

Wow that video was so interesting.

Could you reverse bias the previously lit digit to blast the residual atoms off? I wonder if that would speed things up even more.

Dalbor, a couple of thoughts: 1) instead of having an entire digit, how about just a grid of dots? 2) since you build your own tubes, put the driver electronics inside the glass envelope. Possibly on the "digit wire" itself. 3) instead of a 10 "digit" tube, a "20" digit tube, so that the digits inside one of two cages can be quenched while the second set runs. 4) what about arranging the digits as 7 segment wires. For a number of transitions, segments remain lit from one count to the next. Those segments don't need to be driven "off", and their small size means less capacitance/inductance.

@lordroo8484

Жыл бұрын

But it would look like an LCD display. Not having segmented didgits is one of the main selling points of Nixie tubes!

This is interesting. There are many options to improve response times. For example a sharper wave form, that spikes at the start. Thinner cathodes would provide a lower capacitance and higher field strength. (similar to a corona wire) Slightly higher gas pressure could improve the cooldown time. And taking it to the extreme, shortening the signal path to the tube would reduce inductive influence.

Ha! I can say that I was first in this case! I've documented how high speed switching can be used for cathode poisoning prevention a while ago, there's a video on my channel. I do have to say though, while nixies are very cool, they are still going to be way slower than direct emission LEDs or laser diodes. I've personally confirmed that blue LEDs and laser diodes can be operated at several MHz with very sharp slopes. I've got a couple high speed cameras that do close to and over 200kfps, as well as single point detectors that sample at over 10MHz down to nanolux, let me know if there are other neon projects you want shot fast, for free :)

@daliborfarny

4 жыл бұрын

Spirit I have to go through your videos, that poisoning prevention is interesting! Leds are way faster, nixies cant compete - but customer was used to nixies and also we are here to support nixie tubes, not LEDs :-)

@Spirit532

4 жыл бұрын

​@@daliborfarny This project definitely falls into the category of "didn't think I'd ever see _this_ used _that_ way" :P Side note though: most phosphor-based LEDs *are* slower than nixies.

@heinrichhein2605

4 жыл бұрын

you are not the first;)

@SeanBZA

4 жыл бұрын

Wonder how the turn on time of the tubes would be reduced if you had a UV LED shining into the tube base, with a red filter on the display so that it is not visible from the outside. An alternative to the old method of making the tubes fast by simply making the numerals out of something with a high concentration of Lanthium, as it is mildly radioactive and thus provides a regular source of ionised particles to start the cascade reaction. That or try making a tube with the Amerecium source out of a smoke detector placed on the base of the anode, hiving you a ready source of alpha particles to ionise the tube. just the fast nature of the tube will eecay with time, dropping to slow after about 20 years, and there might be a bit of blue helium glow in there as well by then.

@johnrehwinkel7241

4 жыл бұрын

Yes, a source of ions will speed up ionization. A light shining on the neon will help, a primer electrode will help even more. Some tubes include a radioactive source to supply the initial ionization as well, but a primer electrode is probably just as effective and easier to engineer.

Interesting stuff ! ...cheers.

One application was early electro mechanical television where a Nipkov disk at the receiving TV had an lightsource made of such a gas discharge.

You can improve cold start light-up time by having a radioactive source in the tube. That's how you increase the speed of gas discharge tubes when you want to catch fast transients. You could take the small Americium source from a smoke alarm and place inside one tube keeping it aimed towards the digits to see how big of a difference it makes.

@daliborfarny

4 жыл бұрын

Americium sources were used in some old tubes (e.g. ZM1042), the reason was to make ionization faster for some digits in the middle of the digit stack ("4") in cold temperatures. For us, there would be no benefit as we run already ionized tube.

@Richard.Andersson

4 жыл бұрын

@@daliborfarny You are right that this particular application don't need good cold-start performance. However other applications might be different, and having radioactive sources would be difficult from a legal point of view. Perhaps in such a case it could be smart to have an extra cathode hidden at the bottom of the tube, always glowing and keeping some gas ionized, so that when a digit is turned on it would be almost immediate.

Haha, if you're looking for a programmer, I'm from the US and would love to escape. XD This is really interesting stuff guys, happy to see your channel!

I like how you explain plasma. also you may have some fun with different kind of neon tubes to experiment with (some make their own)... it's seem that gas pressure may change the result. good continuation.

Cheers Dalibor! 👍😎

Thanks for sharing so much of your process in this video and others. My question isn't entirely related to this video, but I was curious if you have tried other gas species mixes to vary colors?

Incredible info .. thank you

Really nice video!

I think Techmoan would just love this.

@voxelfusion9894

4 жыл бұрын

He should do a collab with 8bit guy, maybe to demonstrate old telephones?

I congratulate you friend for your initiative to emerge and grow as a person! The Nixie tubes you make are incredible !!! I was speechless. I live in Venezuela, I have few resources! and how I would like to get those NIXIE Tubes! I would love to make a clock as a project! because my daughter got a lot of attention !!! Thanks in advance...

Dalibor, thx for sharing. It was interesting. Your customer has an interesting application for your Nixies indeed. You could have used a mechanical shutter on a wheel that should be fast enough. However, hat is not the reason for my post. A lot of people use open collector drives to switch the cathode voltage. Keep in mind that the on time for the voltage or in this case since it is being grounded is driven by the electronics used for a switch. Now for the off time or the time it takes for the cathode voltage to relax to a level that turns off the current, is left to chance. There may be enough capacitance in the system to hold the voltage to a level that causes the emissions to persist for a long time electronically speaking. Whew, a lot of words. Consider a high drive and a low drive for the switching. That would force the shortest voltage switching time for each cathode. That would reveal the actual recombination time of the ions around each cathode. Hope this helps.

If you flicker the power to each cathode (numbers) at 240hz with a sharp rise is current, separated by long pause, you can greatly enhance the life of the tube, since it will run at much lower energy levels overall. Also, dimming the tubes or turning down the drive voltage just enough to lower the lumen output some, for dark room or night time use, winter etc, it will also make the tubes last longer & can also reduce power consumption, though will increase complexity of drive logic board etc // just some ideas. Also keeping the tubes lightly ionized, pre threshold voltage will make triggering the light output of each cathode easier with a smaller rise in voltage (switching)

Love your videos. However it was quite hard to understand you when you were at the chalkboard. Maybe a lapel mic next time? P.S. Been following for a while and love your work!

@daliborfarny

4 жыл бұрын

Hi! Thanks for the feedback, I want to buy a shotgun mic next month - this would be perfect for this particular purpose.

@analogaudiorules1724

4 жыл бұрын

It's actually pretty easy to understand him.

Congrats!

I remember reading somewhere (might have been Wikipedia) that neon can glow blue/cyan under high frequency currents. Am I remembering something wrong, or is the frequency just not enough here to manifest that effect? (iirc it was in megahertz range)

@Dalibor: It's not so easy to answer your question about the afterglow. First I'm not sure what's the major mechanism leading to the glow. Your youtube explanation is a little bit oversimplified en.wikipedia.org/wiki/Glow_discharge#Mechanism .... First you should check if your electronics can change the potential on the electrodes fast enough - there is probably quite some capacity to be discharged. However, fast electronics should be able to do that in ~ 50 ns ... personally I use pulsers on beam blanking plates which can drive 60V in 10 ns. After switching off the potentials, there is no electrical field remaining for the electrons to get accelerated. Only space charge from the plasma itself and surfaces charging on insulating parts will still lead to a gradient in the tube. Therefor there should not be a lot of ionization and excitation by electrons or the ions. The ions still have kinetic potential. Depending on your pressure you can try to calculate the mean free path. Do you know the software SRIM? With 100 eV in total, neon on neon scattering will probably not lead to light emission. Way too slow. When the ions hit the cathode or anode, they will cause sputtering and secondary electron emission. Since they will not be accelerated, there should not be a lot of light either. Still excited Neon (atoms or ions) will decay. However, how fast depends on the configuration and can vary a lot. For forbidden transitions it can take minutes. For Neon, everything should be very fast ... I would assume fs to tens of ps - but I'm not an expert on that field. The major contribution for the afterglow is propably Ne+ that captures an electron somewhere. Here the time depends strongly on the pressure in the tube and if the Ne+ hit's a wall. There are still electrical fields in the tube (space charge from the plasma or surface charges)... so they will fly in some direction. Since the glow is very limited to the cathode region, the majority never reaches a wall and they recombine prior to that. Even the first ionization energy is enough 21 eV (UV light), to trigger more secondary emission ... The mean free path in vacuum is for 1 mbar in the order of 100µm. en.wikipedia.org/wiki/Mean_free_path The question is, when will they find a free electron - or more probable, a Ne- since all electrons will be more or less captured (semi-stable) by Neon atoms (not stable, but still). Now, do you know the ionization level in your plasma? (can vary between 1% and 100%)... and now let's say 90% of them must hit the other polarity. ... That get's complicated. I'm at the end of my expertise, but I'm also not a plasma guy :( ... I think your experimental way was a good alternative. However, what you could try still: Not switch off the cathode voltage completely, just lower it below ~ -20V (or lower) instead of turning it off. The voltage should be too low for the electrons to ionize the neon, but could be enough to accelerate the remaining Ne+ onto the cathode. Unfortunately you (edit) not will visit the high speed camera company again, right? Good work, keep going! Best "Ion beam guy from Dresden" Side questions: How does sputtering of Neon ions on the Cathode influence the glow effect and the lifte time of the tube? Whats the cathode material?

@daliborfarny

4 жыл бұрын

Wonderful, thank you very much for your inputs! We checked the electric potential on the cathode with scope and it is discharged very fast - I dont remember exact values, but it was definitely already discharged when the digits were still (after)glowing. We wanted to do visual test just in case the electric signal on cathode differs from the light output. There are lot of interesting information in your comment - how did you calculate the fs/ps recombination/decay time for Ne? I couldnt find any good info on this.. As for the sputtering effect on the glow, the sputtering is at very low rate, so I dont think it makes visible difference. Material - stainless steel.

@demolitionb

4 жыл бұрын

@@daliborfarny The fs/ps decay time for electrons in excited states of Ne is not calculated, it's more an educated guess from my side... Auger cascades are usually fast: iopscience.iop.org/article/10.1088/1367-2630/13/11/113003 ... Here is a master thesis: www.hi-jena.de/hij-publication-download-17241 that I found in the internet... from heisenbergs uncertainty princinple, the liftetime of a decay can be calculated from it's width, here the 20meV would be something like 16 fs. .... the recombination of Ne+ with Ne- or electrons could be much loooooonger since they have to meet and greet in real space. Here I could not come to a conclusive result. Ok, let me continue... Ne at Room temperature (probably this is not the case in the nixie tube plasma), neon should have a speed of www.pfeiffer-vacuum.com/de/know-how/einfuehrung-in-die-vakuumtechnik/grundlagen/thermische-teilchengeschwindigkeit/ v=500 m/s ... and the mean free path of ~100µm, the time between collisions would be in the order of 200 ns. If you would have a 100% plasma and it's enough the 95% of them neutralize, and we would assume 50%Ne+, 50%Ne-.... after one collision, you would have 25% Ne+, 25%Ne-, 50% Ne.... One collision in average more, 12.5%, ... 6% and so on. So 4 collisions in average, ... ~ 800 ns. ... at least only 1.5 orders of magnitude away from reality / your experiment. Maybe the neutralization during the collision of Ne+ and Ne- is not 100%... you have a different pressure, velocity, mean free path in the plasma region, everything is propably not in thermal equilibirum (no boltzmann distribution) or I missed something completely. Here I a quote from lectures come to my mind, that my profs was telling us that nobody really knows what happens inside a plasma and we should not even think that we know everything because it's nearly impossible to probe a plasma without disturbing it.

"What is the half-life of excited neon?" That seems to require "Fermi's Golden Rule" which sadly I don't really understand. *Question:* what would happen if each element was able to independently switch from a cathode to an anode? The idea is similar to powering those old neon glow lamps with an AC source. Would switching the polarity of an active element disperse the afterglow around it? If this even worked, would there be any issues with blowing out the driver circuit or increased wear on the number elements?

@SebastianM93

4 жыл бұрын

I think it would increase the complexity of the driver and to blow the particles away, the cathode needs to be clamped to a higher voltage than the anode. This will cause additional problems.

@paragonseeker2589

4 жыл бұрын

There should be minimal wear in a Vaccum tube as there is little to no gas to react with. Since the anode is Tungsten Steel and it runs the high voltage to be split across the cathodes the reaction is actually ionized plasma, AFAIK as long as there isn't a heavy gas present the metal should not corrode easily, on top of that it looks like he is using stainless/steel which should be more than sufficient for long lifespan. As for neon tubes they do not function the same, the gas is excited and the coating on the tubes gives it color, heat, and spikes of voltages actually reduce the lifespan of the lights. if you want your neon tubes to last, you need to get a proper transformer and keep the lights cool. Since these nixie tubes don't run off the same High voltage line, I would assume they don't run into the same problems.

@spambot7110

4 жыл бұрын

@@paragonseeker2589 nixie tubes aren't vacuum tubes, they have mostly neon at a low pressure

@paragonseeker2589

4 жыл бұрын

@@spambot7110 yes! they're at a low pressure and the gas is not dense.

Great video

Is there any kind of cathode discharge circuit? Cathode capacitance may be the reason for continued afterglow.

Awesome project Dalibor! Greetings from Germany 🇩🇪

I'd be interested to see the light degradation/dissipation time from the cold tube and how it compared to the standard 200ms dim time - it would be even better to include the time to dim from the full speed/"hot" tube so a comparison could be made between all 3. After all, the dim time from the full speed tube is most relevant to the project, and it appears as if the 200ms measurement they showed was from a more average environment as opposed to the high-speed "hot" setting.

This is so cool. Woooow. Moving particles in slomo is fucking awesome 😍

Nice video! Did you try to add some pullup resistors on the cathodes to pull them up to anode voltage? Maybe that would make the digits extinguish faster. I would also try a push-pull driver that actively pulls the cathodes up to a voltage lower than the anode voltage to avoid discharge between digits.

I know it's not what you're aiming for but I quite like the demo at 4:52 of delayed turn on and slow turn off.... it might look nice for seconds on a "normal" Nixie clock. 18:56 AVR Assembly

CHEERS! THANKS FOR THE WORK!

Awesome video! Thanks. What kind of driver are you using? I assume push-pull between GND and +HV (anode voltage) to get the faster fall/rise time? Might it be possible to speed up the fall time and the cut down on the after glow by instead of switching back to +HV, to switch to a higher voltage than +HV? Say, +HV + 50V? This might 'suck up' the extra charge. edit: Looking at the scope readout, you are probably not using a push pull driver, but rather the classic transistor to the ground.

@Konecny_M

4 жыл бұрын

The use of prebias voltage is nothing unusual, same applies for very high repetetion rate LED stroboscopes - the diode is held just at below the VA curve knee. Defacto this comes almost for free with current mode driver.

@jackychan6190

2 жыл бұрын

Great reserach on the Nixie Tube (still have nixie clock in my home and my son's home) characteristic and possible make the nixie tube reborn again, warm and comfortable glowing at night, great invention and worth to continuous.

Very interesting ! And beautiful ...

My understanding is that the after-glow is in part due to metastable electron states that have some characteristic time before they decay and you can't do anything about it. However some part of the afterglow might be due to slow dissapation of electrons, or capacitive effects etc. that might be possible to speed up by reversing polarity for a short time directly when turning off the digit (perhaps not full reversal, but some bias in the wrong way). What do you think?

@SebastianM93

4 жыл бұрын

If I remember well the measuring, the rise and fall times were within 300 to 500ns. So, they are faster than the physical afterglow and there was no need to make the electronic more complex with push-pull drivers or something else. But thanks for the hint 😉

@daliborfarny

4 жыл бұрын

You might be right, as Sebastian mentioned, it would make the electronics very complex and it was not necessary. But very good point, thank you!

You kan make a NOT-gate ether on software (program) or aphysical switch, or a transistor (that can turn ON and OFF in incredable high speed) on everyone of the digit so, if one is ON, no other digit, can go ON at the same time.

Very cool vidéo thanks for sharing experience

I would say this test would be much better looking in film because the low light of the after glow would not appear as bright thus a better result. This would correlate to the type of camera that would be used in the sale era of the Nixie tube.

Only a clock maker would apologize for the audio quality of a KZread video. ;) Had you thought about using AC signaling to cut down those trailing times? For instance, rather than simply turning the tube on and off, could you flip the anode and cathode briefly to help dissipate the energy faster?

Voce também constrói analizadores de espectro de áudio em tubo nixie?

Very informative

Super interesting! 👍

fantastic! i love this video, so much interesting information!

That's amazing it can run at that frequency.

10 μs per digit does not translate to 100 kHz. Assuming all 10 digits are cycled equally, each digit gets 10 μs of light and a 90 μs pause while the other digits turn on and off. So the entire period is 100 μs, i.e. 10 kHz, with a 1:10 duty cycle.

@thebarryfish

2 жыл бұрын

And he has digits that display 1uS resolution??

Could you light up the nixie tube faster by giving it a slightly larger voltage surge? How long is the lifespan of them and what usually fails when the tube dies?

@OtherDalfite

4 жыл бұрын

What were seeing is the Nixie tubes failure to cool down and stop producing light. It's a limitation of the element radiating it's heat into vacuum. Unlikely much else could help this problem.

@vaclavtrpisovsky

4 жыл бұрын

@@OtherDalfite No! Nixies (unlike numitrons) are *_cold_* _cathode_ tubes, meaning they do not produce light like incandescent lamps through a heating filament, but by ionizing neon & argon gas around them when a negative voltage is applied to cathodes. Google _neon lamp._

@vaclavtrpisovsky

4 жыл бұрын

The lifespan is in the order of decades, Dalibor Farný gives a 10-year warranty. What can happen is cathode poisoning, which is the deposition of small oxide particles on digit cathodes, forming a non-conductive layer. Alternatively, like in indicator neon lamps, the same layer may build up on the glass and obstruct visibility, buth with an anode in between this is rarely the case. By far the most common failure, though, is leakage of air into the tube due to glass breaking, mechanical shock or a _slightly_ not airtight stem.

@OtherDalfite

4 жыл бұрын

@@vaclavtrpisovsky you are right! My apologies for spreading misinformation. I just assumed they were similar to low watt lightbulbs and simply were a filament. Thank you for clearing it up for me

How does capacitance of the plates effect the speed? On cathode ray types, we must keep the grid capacitance very low, pico-farad range.

I would think another main limitation on the switching speed would be due to the capacitance of the tube, and that would vary from digit to digit. What if you tried to lower the duty cycle for an on digit to 50% or less? maybe that would greatly reduce the after glow.

Absolutely magnificently beautiful !

Interesting insights.

Have you thought about adding a small amount of quenching gas in your tubes, like hydrogen (as in surge arrester gdt) or bromine (as in GM tube)?

Very interesting! Having access to the high speed camera (free?) was ideal! Here in the USA I doubt a company would have bothered! At least not without a big cost. The reason for using nixies instead of solid state displays is typical of that customer! They do lots of stupid things that could be done much more effectively!

@daliborfarny

4 жыл бұрын

Man, you spoiled my childish joy from working for that customer. But I still feel honoured to see you here, Ron! :-)