You don't need latched outputs - you just shift quickly, and use the output enable to blank while shifting. You also don't actually need drivers rated to the full anode voltage due to the forward voltage drop of the nixies.
Maybe, "open pins" voltage can be clamped, say, by ~10V zener. You have measured 120 V to 10 MOhm resistance, so the current drawed was only ~12 uA. Incresing this value even 10 times should not cause any problem.
Dave, I'm a mechanical engineer and I always learn something new when I watch your videos. You make these videos so incredibly interesting and relevant. Sure wish I had someone like you around to show me the way when it was time to decide what kind of engineering I should study 35 yrs ago. Thanks for taking the time to record, edit and share with us!
Dave, considering datasheet. Voltage at which discharge happens: not more than 170V Current for digits: not more than 2.5mA Current for dot: not more than: 0.3mA Cathode brightness: not less than 100 [candels/(square meter)] Power supply voltage (DC or AC RMS with half bridge): not less than 200V Maintaining discharge voltage: 120 - 170V Operating curent at DC voltage: for digits: 2 - 3.5mA for dot: 0.7mA Operating current at 50Hz AC with half bridge: for digits: not more than 2mA for dot: not more than 0.2mA No noble metals contained.
I really panicked when you got the translation wrong. Good thing you double checked, mark of a good engineer, good on you mate. I was scared you'd burn the thing because of the decimal needing a smaller current.
Important! 0,3 mA is indication current of decimal point. 2.5 mA for numbers. Maximum for DC: .7 mA for decimal point and 2-3.5 mA for numbers. For AC (50 Hz half-way rectifier) max current: 2 mA for numbers and .2 mA for decimal point. UPD Looked further and you used google translate. Few)
Hey Dave, have you considered a fundamentals friday episode on different output types for IC's? Totem pole, open collector, etc. Its one of those things that I've picked up bits of over the years, but i'm sure newbies, and people like myself could do with a refresher, or learn some extra tricks around different types!
simontay1984 it's two transistors in series, one sources to ground and the other drains to voltage, they are driven by the same input on the base or gate, so they have to be complementary, meaning one is npn and the other is pnp, it is called totem pole because one is on top of the other, like a totem lol
simontay1984 Totem pole is a push-pull type of output, it is able to actively drive the output high or low, compared to an open collector that can only actively pull low.
Dave, suggestion from a professional solder; if you cut the legs on the components after soldering, you should resolder. The kick from cutting with a plier can cause micro cracks in the solder joint causing the board to fail after some time. Maybe not super important but nice for viewers to know.
As I know that was adding Kr85 which with half-life of 11 years so till now all should be safe. On other hand uranium (uranium oxide) pottery glaze was used for centuries. Basically all houses build before 1940 used bathroom or kitchen tiles that had been glazed with varying amounts of uranium and most ceramic glazed dinnerware used it. Even now some companies use depleted uranium (rather than the original natural uranium) for production of glaze (for example Homer Laughlin China Company of Newell, West Virginia).
Just because something contains radioactive isotopes doesn't make it dangerous..... A banana is relatively radioactive due to the naturally occurring potassium, and probably orders of magnitude more radioactive than any uranium glass, or isotope added to a Nixie tube. The Uranium glass is perfectly safe to use, even for food or what-have you. You ingest Uranium daily anyway, simply by eating natural foods, and a much larger amount than any Uranium glass would possibly add (which is nearly 0, because it's bound in the glass, it's not going anywhere).
Correct ~ The Kr85 releases a beta particle which was used to create reliable initiation of ionization in the tubes, but a very short half life means that the isotopes are by now mostly inactive.
The channel Mr. Carlson's Lab did a nice project with Nixie tubes. He designed it to be a frequency counter / digital display for radios. It will be interesting to see how it's done here.
Just ordered a set of nixies, yes I'm making a clock, and was happy to come across this. Has been too long since I watched anything on this channel. May be time to binge some mailbag.
Cracking stuff, thank you! Fran got me started on these beauties and I've been learning by getting them working and counting just using decade counters and 10 hv transistors and resistors per tube. I think this should take my meagre understanding to a new level. Looking forward to the next one!
He mentioned fitting 'some kind of display' in his Apple Pi cluster computer project. He said he wanted to remove the DWD RW drive and fit the display in the vacant bay, this looks like it might fit in there ;)
This was an absolutely wonderful episode, Dave! I really enjoyed seeing and hearing your thought process during the design phase of a project. I can't wait to see more videos on this project and to see the end result!
I really love nixie tubes! As a hobbyist, I wanted to make one of these for my first homemade clock, main difference being that I want to use IN-18 tubes! Another factor is that I need money for that...oh well, I'll get my enjoyment from these videos.
Doubt it! They will without a doubt go up in price with time as less and less are around. Unless they find some batch of a few thousand of them hidden in some warehouse, they will probably keep getting more expensive. Guess we've got to settle with IN-14's or IN-16's :(
Yo, Russian translator here. ;) 0.3mA - is the nominal current for the decimal point indication. 0.7 is max. * surprisingly google translate got most of the things correctly.
Fallout 4 is a post-apocalyptic role-playing game developed by Bethesda Game Studios[2] and published by Bethesda Softworks. It is the fifth major installment in the Fallout series (7th overall), and was released on Microsoft Windows, PlayStation 4, and Xbox One on November 10th, 2015. The player is the Sole Survivor of Vault 111, who emerges 210 years to the day and time after the Great War. Prior to this, there is a brief period of gameplay during the pre-War era showing the player living with their spouse and child. As revealed in the trailer on June 3, 2015, the setting takes place in Boston, Massachusetts. Famous local landmarks like the Paul Revere Monument, the USS Constitution, as well as the Massachusetts State House with its unique Golden Dome are included in the game world. Other notable present-day locations that make an appearance in the game are Scollay Square, Bunker Hill memorial, and Fenway Park, which is renamed Diamond City.
This was a blast. I love my nixie collection. These 12's are tough as nails! I do have PCBs for Steins:Gate Divergence meters using a different nixie. The Supertex Serial to Parallel Converter 220V 32Ch and Open Drain Outputs. PQFP-44 packaging. These are really handy for nixie driver projects.
I recently made a Nixie clock/calendar with no less than 14 nixies. I used four of the SN65518FN from TI for driving the tubes. They are a bit hard to find but turned out to be a great solution. You don't need to have a switching solution that can work to the full 170V. The SN65518FN works up to 60V and this is fine. The trick is to switch the low side of the tube to a 60V bias when you want the digit to be OFF and to 0V when you want the digit on. When a digit is off it is naturally open circuit. One resistor can be shared between the tube and an additional one in series for the decimal point. One thing to watch out for is that the 60V 'bias' somehow requires a few mA of current, which doesn't seem like much but does mean that you need a proper 60V source in the circuit -- a simple shunt regulator isn't suitable.
You don't need a 200V rated transistor for each cathode as long as each tube always has a digit lit. You only need something rated for 40-50V. A ULN2803 or ULN2003 transistor array works great and you can protect all of them against overvoltage with a single zener diode. The nixie driver ICs have a similar voltage rating and require one digit to always be on. I used ULN2003's in my nixie clock and it has been running on my desk for years just fine. It runs with all 6 digits multiplexed. They are certainly bright enough to read in normal lighting, but not in direct sunlight.
Neon bulbs strike better when the gas is pre-ionised. This also means that the nixies are light sensitive. At threshold voltages they will likely only start to glow when there's enough light shining on them. I never tried it, but it's an interesting topic for one of your experiments: xenon-flash-start an undervolted nixie tube.
The display driver is the thing....what you´re displaying is second order, as it will pe probably within a µC anyway. If it would be discrete chips without any µC i won´t go that way either; for a clock i would stay in the tube field an go for counter tubes and wire them up directly to the nixies, that would be a nice clock or a counter. You also could use these counter tubes to generate the 1hz clock input siganl direct from the mains AC. It´s a rather simple 8 counter 6 nixie component count and you´re pretty much set, some resistors and blocking caps for the 1Hz impulse generator and that´s about it.
You´ll need something like a nice reference and some HiRes ADCs along with some storage memory and readout conversions. Have you ever opened a nixie based voltmeter...? My Grundig Nixie 5 digit meter has some three big (20 by 40 centimeters) double sided PCBs inside, a lot of nuvistors (little tubes) and some 12 large double width DIP-Chips on them which do all the measurement, conversation and display jobs.
I was thinking about it possibly just being an external display for an existing multimeter (ie. tying into some automation/programming connectors), but it would be quite interesting to see what would go into designing a high precision nixie-based multimeter!
Really awesome video Dave. Been a while since I last was this excited about your upcoming videos. Really nice seeing you design something, even if we don't know yet what it is ;-)
Mode for voltage is 120-170 V, indication current for "numbers" - 2-3.5mA, for "comma" - 0.3-0.7mA. 30 degrees - viewing angle, brightness - not less than 100 cd/m^2.
I helped a friend build a 6 digit (hh-mm-ss) nixie clock back in the early 1970's. It was a kit from Henry's Radio who were very big at the time. It used the typical 7490/74141 combination. Cost was was about £80 for the kit which was a lot of money then. There are still plenty of 74141 on E-bay. Bob
170 V is the maximum voltage when they start glowing, 2.5 mA is the maximum current for digits, 0.3 mA is the maximal current for decimal point, 200 V is the minimum voltage of power supply, 120-170 V - voltage to keep glowing, and also it says that there is no precious metals
Main electrical and lighting parameters. Start of glow voltage, no more 170 V Current of indication for digits, no more 2,5 mA Current of indication for comma, no more 0,3 mA Luminous intensity for cathodes, no less 100 cd/m^2 Angle of view, no less 30 degree Acceptable modes of operation Power supply voltage (DC or AC RMS when used with half-wave rectification), no less 200V Discharge hold voltage 120-170 V Working current when voltage is constant, mA for digits 2-3,5 for comma, no more 0,7 Average working current (50 Hz Power supply with half-wave rectification), mA for digits, no more 2 for comma, no more 0,2 Doesn't contain precious metals.
As far as I know a push pull driver would not work with the Nixie tube. Since the segments are all in a row the current won't be flowing from the grid to the segment but from a segment that is pulled high to the segment that is pulled low. I have a feeling that this is not right.
I was wondering about that too. With those totem pole outputs you'd basically get 9 additional anodes and one cathode. Maybe Dave is planning on leaving the high-voltage supply input of the drivers disconnected? That might do the trick although I have a feeling that's not really how the drivers are intended to be used.
Dave, here values from datasheet: Indication current: 2.5 mA for digits, allowed maximum us 2-3.5mA (for DC) 0.3 mA for dot, allowed maximum is 0.7 mA (for DC)
not allowed maximum, but nominal, i guess. Also, it says that there are no precious metals :D Recommended to solder pins not closer then 4mmfrom glass.
So, the instruction states the following: 12 wires, 1 for anode, 10 for digits, 1 for comma (digit separator). You own PN-12B model. Ignition voltage, max: 170 V. Discharge voltage range: 120-170 V. Discharge current: 2-3,5 mA (I assume it depends on the surface, you can adjust the brightness by controlling the current). Discharge current for comma: 0,7 mA. Which means that you would need only a f***ing high-voltage mosfet package, which is small, one or more microcontrollers to drive the thing, you can have them interconnected and that's all, mostly. In fact, you could make a universal solution with a boost converter on the very same mosfets in the package - one simple solution for many situation. The hardest thing is to write the program for the microcontroller or microcontrollers. With mosfet packages you can even program the brightness, with extra mosfets for each digit of high-voltage DAC conversion. How come it takes 3 videos for some simple circuit designs ?
Can confirm this to be true. I have a beautiful Nixie clock in the living room, and so far precisely 0 people who have visited me _hasn't_ asked questions about it :)
Nice projects you're getting going lately Dave, Looking forward to more parts for this and the Raspberry Pi too. :) If I remember rightly the Suzuki GT750 nicknamed 'The Kettle' because it was a water cooled 2 stroke Motorcycle had a Nixie tube in the clock cluster to display what gear you were in. I had a Suzuki GT250 again 2 stroke and it had the gear shift contacts in the gearbox to install something along the lines of the GT750. I wanted to wire a 7 segment display with decoder to respond to which line was grounded depending what gear you was in but never got around to it before the bike died on me. :(
Can you please share all of your educational videos onto a torrent network? This would be really great. I cannot stress how invaluable your videos are. Please don't keep them locked away on RUclips.
Yes I understand the immensity of transcoding everything and seeding it on torrents or hosting it at Archive.org. And I do know you can download videos on RUclips with your browser, believe me. But your entire collection really is deserving to be hosted on its own. It shouldn't be locked away here always requiring an internet connection. Please consider releasing your work in a more proper manner, it would be a great service for everyone.
Perhaps one of Dave's viewers might be able to download all the YT videos in HD and create a collection, with Dave's blessing of course. Volunteers please?
Also, you can drive the thing by supporting a constant high-frequency corona discharge inside it. It would allow you to drive it without heater at much lower voltage, as you won't need initial ionization. The correct solution would be using no registers or whatever and use a microcontroller with high voltage DAC circuit for each of the tubes, as it would give a universal solution with less wires and many protocols supported.
My design of nixie power supply used a XL6008 boost converter in TO-252 package and a TTRN-060S transformer, configured in flyback topology. The total component count is less than 10, cost less than $2 and fits in a 16x20mm PCB. Input can go from 5V to 24V, with 170V/20mA output.
if you stay in the tube field, it´s so easy to set up a clock, go for counter tubes and wire them directly to the nixies and count them via some more counter tubes...you just need 8 counter tubes and 6 nixies for a full clock a bunch of sockets, resistors and some blocking caps...that´s it...
In the final version of my clock I used a Atmega8, some 74HC595 and russian 74141 plus a MC34063A stepup. I would say that is much easier but how you build your clock in the end is personal preference.
These tubes were still being used by Bell Canada in 1980 in SxS - ANI (Step-by-Step - Automatic Number Identification) manufactured by Northern Electric
I tried this out by looking into a little box here at a reasonable range, it looks like around 2mm difference at that depth. You do see the 1 point closure of the edges. You would have to be looking more or less straight on for it to correct, so maybe they wouldn't care.
FIY the datasheet translates to: """ Main electrical and lighting parameters: Discharge voltage, no more than 170V Digit indication current, no more than 2.5mA Comma indication current, no more than 0.3mA Cathode luminescence, no less than 100 cd/sq.m. #candella per square meter Viewing angle, no less than 30 degrees Allowable operating modes: supply voltage (direct current, or effective pulse/peak voltage of half-wave rectified alternating current) of not less than 200V Voltage required to sustain the discharge 120-170V Operating current with applied constant voltage: Digit 2-3.5mA Comma 0.7mA Average operating current (with half-wave rectified 50Hz A/C): Digit 2mA Comma 0.2ma Does not contain precious metals """
Hey Dave, just wanted you to know that the mighty engineers of my great country created a special IC to power these things - К155ИД1. You just hook each cathode of the tube to one pin of the IC, and based on one of the 2^6=16 possible high/low inputs on other four pins, it allows current to flow to the ground of your circuit through the tube. Please let me know if you need some of them - can send it to you :)
For me individual drive solution is much better than multiplexing, flickering nixies. When I am making my devices, I use flicker flickering/multiplexing only if it is absolutely necessary (huge matrices such as LCD screens), and even then I try to implement a filter (e.g. LC).
Could you have looked at getting a binary decoder? something with 4 digital inputs and 16 output pins? It would have a couple redundant pins, but hey, whatevs!
9:28 - On the HV supply, L1 looks like it is marked on the board as a 'laid down' component (i.e. mount it horizontal). I guess that this is why the cap and inductor were squabbling over that small piece of PCB? Strangely interesting vid though looking for chips on DigiKey, although not sure which one he ended up getting... the 16 bit 0.5mm pitch PITA, or the 8 bit jobbie.
They are not new production, but "new-old-stock".(NOS). They were made decades ago, by the Soviet central planners. Still a lot of them in the warehoused in the former Soviet Union (Russia, et al). Same goes with many tubes. Yes, a few tubes, are still being made, but most "types", are just old inventory, that's been sitting on the shelves, for decades. I use to buy the Russian 74141s in lots of 100, years ago, for a kit, I sold. Use to pay ~70 cents for them. The inventory, though readily available, is thinning out, and the price has been going up. Use to have a lot more of them out there.
belive me, pc817 + resistor works a treat as a anode driver for nixie tubes. still working after 4+ years. a simple 555 boost with no feedback is good enough for the power supply.
Just curious as to why you chose 22k for the dropper resistor value, based on the datasheet wouldn't somewhere around 48k be better to achieve the 3.5mA? Or is this due to the resistance of the nixie itself? Thanks!
The 170V on the anode only serves to kick-start the discharge. Once the process is started, the nixie only needs about 120V to sustain the glow. If you apply 170V directly without drop resistor, you'll end-up burning your nixie. The resistor must be calculated for the ~50V drop across it, so that's why you end-up with a 22k resistor. See threeneurons.wordpress.com/nixie-power-supply/ for more details on how to power a nixie.
If you are going to go to the trouble of laying out a board then why don't you just jump to the 0.5mm pitch QFN package? With some careful layout you might be able to squeeze the driver inside the pinout for the nixie tube. It would make a really clean design with just a bus for clk, data, latch and power running between the digits. Great video.
You don't need latched outputs - you just shift quickly, and use the output enable to blank while shifting.
You also don't actually need drivers rated to the full anode voltage due to the forward voltage drop of the nixies.
Blanking is morally offensive :-> But yeah.
As for the voltage requirement, I did measure the open voltage on one unit and the highest voltage was 122V.
Mike, you are wise.
See my 2nd channel video for a demo of this.
Maybe, "open pins" voltage can be clamped, say, by ~10V zener. You have measured 120 V to 10 MOhm resistance, so the current drawed was only ~12 uA. Incresing this value even 10 times should not cause any problem.
Dave, I'm a mechanical engineer and I always learn something new when I watch your videos. You make these videos so incredibly interesting and relevant. Sure wish I had someone like you around to show me the way when it was time to decide what kind of engineering I should study 35 yrs ago. Thanks for taking the time to record, edit and share with us!
Dave, considering datasheet.
Voltage at which discharge happens: not more than 170V
Current for digits: not more than 2.5mA
Current for dot: not more than: 0.3mA
Cathode brightness: not less than 100 [candels/(square meter)]
Power supply voltage (DC or AC RMS with half bridge): not less than 200V
Maintaining discharge voltage: 120 - 170V
Operating curent at DC voltage:
for digits: 2 - 3.5mA
for dot: 0.7mA
Operating current at 50Hz AC with half bridge:
for digits: not more than 2mA
for dot: not more than 0.2mA
No noble metals contained.
Thanks for translating Mikhail!
I really panicked when you got the translation wrong.
Good thing you double checked, mark of a good engineer, good on you mate.
I was scared you'd burn the thing because of the decimal needing a smaller current.
Important! 0,3 mA is indication current of decimal point. 2.5 mA for numbers. Maximum for DC: .7 mA for decimal point and 2-3.5 mA for numbers. For AC (50 Hz half-way rectifier) max current: 2 mA for numbers and .2 mA for decimal point.
UPD Looked further and you used google translate. Few)
indeed!
Hey Dave, have you considered a fundamentals friday episode on different output types for IC's? Totem pole, open collector, etc. Its one of those things that I've picked up bits of over the years, but i'm sure newbies, and people like myself could do with a refresher, or learn some extra tricks around different types!
simontay1984 it's two transistors in series, one sources to ground and the other drains to voltage, they are driven by the same input on the base or gate, so they have to be complementary, meaning one is npn and the other is pnp, it is called totem pole because one is on top of the other, like a totem lol
simontay1984
Totem pole is a push-pull type of output, it is able to actively drive the output high or low, compared to an open collector that can only actively pull low.
Dave, suggestion from a professional solder; if you cut the legs on the components after soldering, you should resolder. The kick from cutting with a plier can cause micro cracks in the solder joint causing the board to fail after some time. Maybe not super important but nice for viewers to know.
Awesome video Dave! I liked how you had all the different chips you looked at in there, instead of just showing us the one you ultimately went with.
Some of the Russian nixie tubes have a radioactive isotope added to decrease the low-light turn on time.
As I know that was adding Kr85 which with half-life of 11 years so till now all should be safe.
On other hand uranium (uranium oxide) pottery glaze was used for centuries. Basically all houses build before 1940 used bathroom or kitchen tiles that had been glazed with varying amounts of uranium and most ceramic glazed dinnerware used it.
Even now some companies use depleted uranium (rather than the original natural uranium) for production of glaze (for example Homer Laughlin China Company of Newell, West Virginia).
Just because something contains radioactive isotopes doesn't make it dangerous..... A banana is relatively radioactive due to the naturally occurring potassium, and probably orders of magnitude more radioactive than any uranium glass, or isotope added to a Nixie tube. The Uranium glass is perfectly safe to use, even for food or what-have you. You ingest Uranium daily anyway, simply by eating natural foods, and a much larger amount than any Uranium glass would possibly add (which is nearly 0, because it's bound in the glass, it's not going anywhere).
+Mythricia radioactive banana mmmmm
Correct ~ The Kr85 releases a beta particle which was used to create reliable initiation of ionization in the tubes, but a very short half life means that the isotopes are by now mostly inactive.
Fun fact all bananas you can buy in the normal stores are sterile mutants.
Ah, EEVBlog. You always put a smile on my face.
The channel Mr. Carlson's Lab did a nice project with Nixie tubes. He designed it to be a frequency counter / digital display for radios. It will be interesting to see how it's done here.
The beginning of some classic EEVBlog.
Looking forward to the rest of these vids.
Just ordered a set of nixies, yes I'm making a clock, and was happy to come across this. Has been too long since I watched anything on this channel. May be time to binge some mailbag.
Dave, I'm loving these project videos. I know they're a lot of work, but they're fun to follow along with!
Cracking stuff, thank you! Fran got me started on these beauties and I've been learning by getting them working and counting just using decade counters and 10 hv transistors and resistors per tube. I think this should take my meagre understanding to a new level. Looking forward to the next one!
I Quite like the touch screen support for DaveCad! It looks so bold and responsive!
8 Nixie tubes with decimal points and saying that it's going to be a project that we'll "find out later"?
Calculator.
frequency meter perhaps or some type of lab equipment
Spoder Man PLACE BET NOW! :P
I'm placing a bet on a Dave bullshit meter. Don't let it near solar roadways or it will explode
No, it's to display the count of his unopened Mailbag items. He needs all the digits.
He mentioned fitting 'some kind of display' in his Apple Pi cluster computer project. He said he wanted to remove the DWD RW drive and fit the display in the vacant bay, this looks like it might fit in there ;)
This was an absolutely wonderful episode, Dave! I really enjoyed seeing and hearing your thought process during the design phase of a project. I can't wait to see more videos on this project and to see the end result!
So that's how you looking for specific chips. Nice work. Will wait for all progress you do.
Steins;Gate anyone?, I can't be the only one thinking that lol.
Gotta love Nixie tubes.
wrong kind of tube for the divergence meter though, you'd want the 90 degrees/freestanding ones.
Loved the divergence meter, it looked so much like a real electronics project :)
What world line are we in?
LOL i was thinking the same thing :D
TUTURUUUU
I really love nixie tubes! As a hobbyist, I wanted to make one of these for my first homemade clock, main difference being that I want to use IN-18 tubes! Another factor is that I need money for that...oh well, I'll get my enjoyment from these videos.
Doubt it! They will without a doubt go up in price with time as less and less are around. Unless they find some batch of a few thousand of them hidden in some warehouse, they will probably keep getting more expensive. Guess we've got to settle with IN-14's or IN-16's :(
Yo, Russian translator here. ;)
0.3mA - is the nominal current for the decimal point indication. 0.7 is max.
* surprisingly google translate got most of the things correctly.
Who else immediately thought of Fallout 4 when they saw this?
What's Fallout 4?
LOL
Fallout 4 is a post-apocalyptic role-playing game developed by Bethesda Game Studios[2] and published by Bethesda Softworks. It is the fifth major installment in the Fallout series (7th overall), and was released on Microsoft Windows, PlayStation 4, and Xbox One on November 10th, 2015.
The player is the Sole Survivor of Vault 111, who emerges 210 years to the day and time after the Great War. Prior to this, there is a brief period of gameplay during the pre-War era showing the player living with their spouse and child.
As revealed in the trailer on June 3, 2015, the setting takes place in Boston, Massachusetts. Famous local landmarks like the Paul Revere Monument, the USS Constitution, as well as the Massachusetts State House with its unique Golden Dome are included in the game world. Other notable present-day locations that make an appearance in the game are Scollay Square, Bunker Hill memorial, and Fenway Park, which is renamed Diamond City.
it's a world without the transistor... :(
What's a joke?
This was a blast. I love my nixie collection. These 12's are tough as nails! I do have PCBs for Steins:Gate Divergence meters using a different nixie. The Supertex Serial to Parallel Converter 220V 32Ch and Open Drain Outputs. PQFP-44 packaging. These are really handy for nixie driver projects.
I recently made a Nixie clock/calendar with no less than 14 nixies. I used four of the SN65518FN from TI for driving the tubes. They are a bit hard to find but turned out to be a great solution.
You don't need to have a switching solution that can work to the full 170V. The SN65518FN works up to 60V and this is fine. The trick is to switch the low side of the tube to a 60V bias when you want the digit to be OFF and to 0V when you want the digit on. When a digit is off it is naturally open circuit. One resistor can be shared between the tube and an additional one in series for the decimal point. One thing to watch out for is that the 60V 'bias' somehow requires a few mA of current, which doesn't seem like much but does mean that you need a proper 60V source in the circuit -- a simple shunt regulator isn't suitable.
Nixie tubes are cool. I need to get my hands on a few.
Vv yo gji fat t ifYu guy yo
Offfddddddddxtgg CD xz a try g G u s
mate look at the reference picture for that high voltage power supply, there's a marked area where you're supposed to bend over that inductor
Really liking this (sort of) project! As a Ben Heck fan, i like seeing you do a more hands-on project soldering and such
Really fall in love with these tubes, very nice! Never seen that small.
Fan of Mr Carlson's Lab?
He's got a bunch of great vids playing with Nixies. Fun stuff.
YES!! When I saw the title I instantly watched the video, gotta love nixie!
excellent dave. subscriber counter?? you know that i love leds but this is so cool
this was awesome, would love to see more programs like this, very informative and useful! thank you so much!
The silkscreen shows one of the fouling parts (101) on it's side.
You don't need a 200V rated transistor for each cathode as long as each tube always has a digit lit.
You only need something rated for 40-50V. A ULN2803 or ULN2003 transistor array works great and you can protect all of them against overvoltage with a single zener diode.
The nixie driver ICs have a similar voltage rating and require one digit to always be on.
I used ULN2003's in my nixie clock and it has been running on my desk for years just fine. It runs with all 6 digits multiplexed. They are certainly bright enough to read in normal lighting, but not in direct sunlight.
It's not that easy. See my 2nd channel video for a demo of up to 122V open circuit voltage on the off cathodes.
But yes, the 74141's look to have zener clamps on each pin.
Putting a zener diode between 0V and the com pin on the ULN2803 provides the overvoltage protection like the 74141.
Ah, yes, I forgot the common pin, yes, that will certainly work with a single zener. 13 chips for 88 lines.
these look SO COOL
Neon bulbs strike better when the gas is pre-ionised. This also means that the nixies are light sensitive. At threshold voltages they will likely only start to glow when there's enough light shining on them. I never tried it, but it's an interesting topic for one of your experiments: xenon-flash-start an undervolted nixie tube.
Is it going to be a clock? Please don't build a clock. Everybody builds a clock.
A multimeter would be nice.
I second that. Maybe an external display for a 7 1/2 digit multimeter?
A youtube subscribers counter ?
The display driver is the thing....what you´re displaying is second order, as it will pe probably within a µC anyway. If it would be discrete chips without any µC i won´t go that way either; for a clock i would stay in the tube field an go for counter tubes and wire them up directly to the nixies, that would be a nice clock or a counter.
You also could use these counter tubes to generate the 1hz clock input siganl direct from the mains AC.
It´s a rather simple 8 counter 6 nixie component count and you´re pretty much set, some resistors and blocking caps for the 1Hz impulse generator and that´s about it.
You´ll need something like a nice reference and some HiRes ADCs along with some storage memory and readout conversions. Have you ever opened a nixie based voltmeter...?
My Grundig Nixie 5 digit meter has some three big (20 by 40 centimeters) double sided PCBs inside, a lot of nuvistors (little tubes) and some 12 large double width DIP-Chips on them which do all the measurement, conversation and display jobs.
I was thinking about it possibly just being an external display for an existing multimeter (ie. tying into some automation/programming connectors), but it would be quite interesting to see what would go into designing a high precision nixie-based multimeter!
So glad you didn't just do yet another basic Nixie clock!
Really awesome video Dave. Been a while since I last was this excited about your upcoming videos.
Really nice seeing you design something, even if we don't know yet what it is ;-)
Yes, finally! I was waiting this.
Nixie are so beautiful.
Mode for voltage is 120-170 V, indication current for "numbers" - 2-3.5mA, for "comma" - 0.3-0.7mA. 30 degrees - viewing angle, brightness - not less than 100 cd/m^2.
Dave, V_bias might be a really useful feature as it allows easy brightness regulation for your display
that´s why they put it into....if you ever saw a nixie clock in a dark room you know why...it´s just too bright....
30 minutes of raw engineering awesomeness!!!!
I find it amusing how you call 7-segment displays traditional compared to these.
Gotta love a Nixie tube. They're just works of art, really.
I helped a friend build a 6 digit (hh-mm-ss) nixie clock back in the early 1970's. It was a kit from Henry's Radio who were very big at the time. It used the typical 7490/74141 combination. Cost was was about £80 for the kit which was a lot of money then. There are still plenty of 74141 on E-bay. Bob
awe, this gave me flashbacks of playing minecraft and building decoders, alu´s and displays with redstone. good times.
thy for the nostalgia.
170 V is the maximum voltage when they start glowing, 2.5 mA is the maximum current for digits, 0.3 mA is the maximal current for decimal point, 200 V is the minimum voltage of power supply, 120-170 V - voltage to keep glowing, and also it says that there is no precious metals
Oh, i didn't see that you used a google translator later
yep!
One the things I now pick up on is how people seem to think that low pressure is not a vacuum. Sure it isn't a "true vacuum" but it is still a vacuum.
Main electrical and lighting parameters.
Start of glow voltage, no more 170 V
Current of indication for digits, no more 2,5 mA
Current of indication for comma, no more 0,3 mA
Luminous intensity for cathodes, no less 100 cd/m^2
Angle of view, no less 30 degree
Acceptable modes of operation
Power supply voltage (DC or AC RMS when used with half-wave rectification), no less 200V
Discharge hold voltage 120-170 V
Working current when voltage is constant, mA
for digits 2-3,5
for comma, no more 0,7
Average working current (50 Hz Power supply with half-wave rectification), mA
for digits, no more 2
for comma, no more 0,2
Doesn't contain precious metals.
As far as I know a push pull driver would not work with the Nixie tube. Since the segments are all in a row the current won't be flowing from the grid to the segment but from a segment that is pulled high to the segment that is pulled low. I have a feeling that this is not right.
I was wondering about that too. With those totem pole outputs you'd basically get 9 additional anodes and one cathode. Maybe Dave is planning on leaving the high-voltage supply input of the drivers disconnected? That might do the trick although I have a feeling that's not really how the drivers are intended to be used.
Exactly and he already pointed out in the forum that it does not work that great. He is going to show in the next part.
"Excuse me. Flight attendant? The man next to me would appear to be detonating a bomb."
Dave, here values from datasheet:
Indication current:
2.5 mA for digits, allowed maximum us 2-3.5mA (for DC)
0.3 mA for dot, allowed maximum is 0.7 mA (for DC)
not allowed maximum, but nominal, i guess.
Also, it says that there are no precious metals :D
Recommended to solder pins not closer then 4mmfrom glass.
100 Cd/m2 is light intensity
30 degree view angle
Also, dot could be light up with digit, when it's should be lower current, maybe you should keep that in mind
So, the instruction states the following:
12 wires, 1 for anode, 10 for digits, 1 for comma (digit separator). You own PN-12B model.
Ignition voltage, max: 170 V.
Discharge voltage range: 120-170 V.
Discharge current: 2-3,5 mA (I assume it depends on the surface, you can adjust the brightness by controlling the current).
Discharge current for comma: 0,7 mA.
Which means that you would need only a f***ing high-voltage mosfet package, which is small, one or more microcontrollers to drive the thing, you can have them interconnected and that's all, mostly. In fact, you could make a universal solution with a boost converter on the very same mosfets in the package - one simple solution for many situation. The hardest thing is to write the program for the microcontroller or microcontrollers. With mosfet packages you can even program the brightness, with extra mosfets for each digit of high-voltage DAC conversion. How come it takes 3 videos for some simple circuit designs ?
I've got some great IN-4 tubes around, prefer them to even these type of tubes. Very pretty :)
Dave is just a genius! GREAT!
very nice to see a project again! :)
Watching some of this makes me realize how much I really don't know coming fresh out of getting EE degree.
The painful truth right there.
I wonder if i'd have been better off in the interview with 'look at all this cool shit i've made in the last 4 years'
A clock made of these would make a great conversation piece
Can confirm this to be true. I have a beautiful Nixie clock in the living room, and so far precisely 0 people who have visited me _hasn't_ asked questions about it :)
Mythricia I'm planning on making a clock with nixie tubes. What are some cheeper ones I could get?
Also, any tips?
Nice projects you're getting going lately Dave, Looking forward to more parts for this and the Raspberry Pi too. :)
If I remember rightly the Suzuki GT750 nicknamed 'The Kettle' because it was a water cooled 2 stroke Motorcycle had a Nixie tube in the clock cluster to display what gear you were in.
I had a Suzuki GT250 again 2 stroke and it had the gear shift contacts in the gearbox to install something along the lines of the GT750. I wanted to wire a 7 segment display with decoder to respond to which line was grounded depending what gear you was in but never got around to it before the bike died on me. :(
Great video Dave, one of your best IMHO.
Can you please share all of your educational videos onto a torrent network? This would be really great. I cannot stress how invaluable your videos are. Please don't keep them locked away on RUclips.
Why don't you download from RUclips? Each videos is several GB long, I have almost 1000 videos.
Yes I understand the immensity of transcoding everything and seeding it on torrents or hosting it at Archive.org. And I do know you can download videos on RUclips with your browser, believe me. But your entire collection really is deserving to be hosted on its own. It shouldn't be locked away here always requiring an internet connection. Please consider releasing your work in a more proper manner, it would be a great service for everyone.
Perhaps one of Dave's viewers might be able to download all the YT videos in HD and create a collection, with Dave's blessing of course. Volunteers please?
Also, you can drive the thing by supporting a constant high-frequency corona discharge inside it. It would allow you to drive it without heater at much lower voltage, as you won't need initial ionization.
The correct solution would be using no registers or whatever and use a microcontroller with high voltage DAC circuit for each of the tubes, as it would give a universal solution with less wires and many protocols supported.
My design of nixie power supply used a XL6008 boost converter in TO-252 package and a TTRN-060S transformer, configured in flyback topology. The total component count is less than 10, cost less than $2 and fits in a 16x20mm PCB. Input can go from 5V to 24V, with 170V/20mA output.
Dave, you can find 74141s on eBay if you want to re-consider that route.
No need for a hv transistor. A BC546 is just fine. The 74141 outputstages only withstand around 60V too, like the BC546.
Without zener diodes. I had that at first for my nixie clock. But all that consumed too much space, so I used russian 74141 insted.
if you stay in the tube field, it´s so easy to set up a clock, go for counter tubes and wire them directly to the nixies and count them via some more counter tubes...you just need 8 counter tubes and 6 nixies for a full clock a bunch of sockets, resistors and some blocking caps...that´s it...
In the final version of my clock I used a Atmega8, some 74HC595 and russian 74141 plus a MC34063A stepup. I would say that is much easier but how you build your clock in the end is personal preference.
of course ;)
These tubes were still being used by Bell Canada in 1980 in SxS - ANI (Step-by-Step - Automatic Number Identification) manufactured by Northern Electric
I wonder if the digits towards the pin side are bigger to cancel perspectival reduction.
Over less than an inch?
I tried this out by looking into a little box here at a reasonable range, it looks like around 2mm difference at that depth. You do see the 1 point closure of the edges. You would have to be looking more or less straight on for it to correct, so maybe they wouldn't care.
I'm about to build a Tsyklon labs chaos divider which uses the same Nixie tube hence I found this video very interesting.
FIY the datasheet translates to:
"""
Main electrical and lighting parameters:
Discharge voltage, no more than 170V
Digit indication current, no more than 2.5mA
Comma indication current, no more than 0.3mA
Cathode luminescence, no less than 100 cd/sq.m. #candella per square meter
Viewing angle, no less than 30 degrees
Allowable operating modes:
supply voltage (direct current, or effective pulse/peak voltage of half-wave rectified alternating current) of not less than 200V
Voltage required to sustain the discharge 120-170V
Operating current with applied constant voltage:
Digit 2-3.5mA
Comma 0.7mA
Average operating current (with half-wave rectified 50Hz A/C):
Digit 2mA
Comma 0.2ma
Does not contain precious metals
"""
these design process videos are great! learnt a lot
Looks like a fun project. Looking forward to the future vids.
YAY project time! :D Thanks Dave.,
You should be able to cycle the digits much faster than visible with a plain shift register, but nice, I never knew the Microchip device existed.
Great video! Loved the parametric search bit
04:13 - No! It is not minimum sustaining current, it is maximum allowed current for "decimal point" electrode.
9:29 The inductor should have been bent down on the board and soldered like this. :D
Great video series!
I have used the HV5530 in designs. 32-bit shift register with high voltage open drain outputs. Can get 10 in a sample order from microchip too.
spoke too soon, you found it lol.
The inductor on the high-voltage supply board is supposed to be mounted on its side as you can see in all of the photos.
This is great! considering making a clock too! thanks for sharing! :)
Hey Dave, just wanted you to know that the mighty engineers of my great country created a special IC to power these things - К155ИД1. You just hook each cathode of the tube to one pin of the IC, and based on one of the 2^6=16 possible high/low inputs on other four pins, it allows current to flow to the ground of your circuit through the tube. Please let me know if you need some of them - can send it to you :)
is that DaveCAD 2.1? liking the grid and fullscreen function
These are awesome! i have been wanting to make a Nixie tube clock
"I love Nixie tubes, they're fantastic".
Sounded like Trump for a second. ;)
Make Nixie Tubes Great Again!
Sounds like Louis Rossman on his Apple product repair channel. Lol
For me individual drive solution is much better than multiplexing, flickering nixies.
When I am making my devices, I use flicker flickering/multiplexing only if it is absolutely necessary (huge matrices such as LCD screens), and even then I try to implement a filter (e.g. LC).
DUDE ! SOOO AAAWSUUUNMMM ! THANK YOU FOR THE TOTALLY KEWWLLL BUILD ! I LOVE YOU LIKE A BROTHER I NEVER HAD ! 💝💝💝💝💝💝💝💝💝💝💝💝💝💝💝💝💝
Really interesting to see how you search for parts :)
8:21 you can see the inductor laying on its side 9:28 oh Dave i see what you did there XD
Could you have looked at getting a binary decoder? something with 4 digital inputs and 16 output pins? It would have a couple redundant pins, but hey, whatevs!
You didn't watch the whole video
EEVblog I must have missed it. ah well.
Yay projects! I love project videos :)
I love this. Thanks for sharing!
9:28 - On the HV supply, L1 looks like it is marked on the board as a 'laid down' component (i.e. mount it horizontal). I guess that this is why the cap and inductor were squabbling over that small piece of PCB? Strangely interesting vid though looking for chips on DigiKey, although not sure which one he ended up getting... the 16 bit 0.5mm pitch PITA, or the 8 bit jobbie.
74141's are plentiful on eBay. 95 results for "74141 nixie".
Good to see some modern parts and what to look for if you needed to build larger setups though.
They are not new production, but "new-old-stock".(NOS). They were made decades ago, by the Soviet central planners. Still a lot of them in the warehoused in the former Soviet Union (Russia, et al). Same goes with many tubes. Yes, a few tubes, are still being made, but most "types", are just old inventory, that's been sitting on the shelves, for decades. I use to buy the Russian 74141s in lots of 100, years ago, for a kit, I sold. Use to pay ~70 cents for them. The inventory, though readily available, is thinning out, and the price has been going up. Use to have a lot more of them out there.
Hey Dave, if you have a look at the silkscreen it looks as though the inductor is supposed to be soldered horizontally (hence the spacing).
Love this project!
belive me, pc817 + resistor works a treat as a anode driver for nixie tubes. still working after 4+ years. a simple 555 boost with no feedback is good enough for the power supply.
"Love the nipple." - Dave
Anyway I look forward to the rest of this. I love Nixie tubes.
It's really a shame that these really aren't made anymore I wish I could get a hold of some they are such cool technologies
Just curious as to why you chose 22k for the dropper resistor value, based on the datasheet wouldn't somewhere around 48k be better to achieve the 3.5mA? Or is this due to the resistance of the nixie itself? Thanks!
I wanted about 2mA
wouldn't that require an 85k resistor? Sorry if im overlooking something.
The 170V on the anode only serves to kick-start the discharge. Once the process is started, the nixie only needs about 120V to sustain the glow. If you apply 170V directly without drop resistor, you'll end-up burning your nixie.
The resistor must be calculated for the ~50V drop across it, so that's why you end-up with a 22k resistor.
See threeneurons.wordpress.com/nixie-power-supply/ for more details on how to power a nixie.
***** That clears things up, thanks!
You know, anyone from ex-USSR can send you any amount of cheap К155ИД1 chips, specialy designed for driving a nixie tubes.
If you are going to go to the trouble of laying out a board then why don't you just jump to the 0.5mm pitch QFN package? With some careful layout you might be able to squeeze the driver inside the pinout for the nixie tube. It would make a really clean design with just a bus for clk, data, latch and power running between the digits. Great video.