You need TWO current limiters based on the B7971 datasheet information. As you mentioned, each cathode needs the current limited; I use an NPN transistor with a resistor at the emitter, to limit the current. The base is driven from a known voltage, say 3.3V, and the base-emitter voltage is 0.7V. So, to control the current, select a resistor value RE=(3.3-0.7)/current. The smallest segments require 4mA, so their resistor will be ~650 ohms. The remaining segments need 5.0,5.5, or 6.0mA so their resistors will be 520, 470, or 430 ohms. Now, the second current limiter is required because the max current of ALL cathodes combined must be 21mA or less. So, at the anode side, you need a similar current limiter. I use a PNP transistor. It sounds like a lot of extra cost and hassle, but electronic components are essentially free and in unlimited supply, whereas B7971 tubes are getting very expensive (typically 200USD each), and increasingly harder to find. It will take time to find a set of 6. Glad I got mine (and a few others) years ago.
All true. I did assess all the options, particularly with the various differences in optimal segment current, as the slightly shorter cathodes need more limiting, etc... I will get around to rebuilding it - and I have many other tubes, so I wasn't worried about putting wear on this one, but my bigger project requires an array of proper driver boards.
Fran....If these tubes are driven correctly and to spec they will last forever. I have (5) 6 digit and (2) 4 digit B-7971 clocks run that have been running essentially 24 x 7 for well over 12 years with no issues, degredation or loss of brightness. Considering these tubes are pulls from the stock brokerage displays and who knows how many hours they had on them before those displays were de-commissioned, I consider them some of the most robust Nixies out there. I have no issues with the tubes getting hot or even warm for that matter. I think once you have your design with individual cathode current regulation you'll see a big difference. Finally, I know of others that have been running these tubes 24 x 7 since the mid 70's without issue. Good luck and can't wait to see what you come up with.
You are right! My Datasheet says max. 21mA current if all segments are lighted. This is in theory a dissipation of 3.57W at 170V for the whole tube. In real world it will be much less because the voltage drops if the segment has ignited. There are different currents for the individual segments. For a long live it will be good to run not more than 4 mA through each Segment.
Fran I guess you got carried away a little with the helium. Never heard any story that vacuum tubes get contaminated with helium. It's usually the cathode losing emission over time. Going on air is most likely due to failing seals I would think. The getter will try to counter this but there is only so much available....
About 50 years ago, the local Burstein-Applebee store carried a bit of surplus and I bought a bunch of those boards for $5 each! I built a nixie clock out of 6 tubes and it ran for over 20 years before I replaced it with a smaller nixie clock using the Russian tubes. Some day I may have to plug the big clock back in to see if it still works.
The data sheet gives you the current limit per segment and you really should follow that. One resistor per segment is more components on the PCB but the tube will "last forever" with a better even brightness. If the white area is growing significantly you are likely over driving in general. I have a clock running for years and seen no change in the tubes and they are never hot or even warm to touch. Standard numeric nixie tubes will also specify current drive per numeral as the different sizes have different demands, most designs also ignore that requirement.
It reminds me of my first LED 7 segment display where I only had one current limit resistor, and the display dimmed more when more segments lit. After I redid the circuit to have each segment limited, the display remained bright no matter how many segments lit. The best bright display is when you can multiplex the display, and run more current for a short time, and your vision persistence makes up for the off time. Do you have the background grounded or floating? If it is grounded, perhaps it attracts the electrons, which causes the bleaching? As far as helium goes, it is able to leak past seals and threaded joints. which means it doesn't have to travel through material except under the high pressure conditions.
Or a TPIC6B595 shift register. Considerably simplifies the wiring back to the microcontroller as you can drive a virtually infinite number of daisy-chained displays without the need for any extra microcontroller outputs, or any multiplexing of the tubes. Then just a single currently limiting resistor on each segment, rated according to the segment size and recommended maximum current as given in the B7971 datasheet. This is my much preferred way of driving these tubes. Plus add a boost converter to generate the 170v DC, and then you can run everything from a single 12v DC supply.
@@stuartmcconnachie cool, I'm only worried about the max Vds of this chip being 50V. Seems a bit low for me. That's why I'd combine a 74HC595 with SN75468.
@@KeritechElectronics Works just fine. The outputs of the 6B595 are diode clamped at the 50v to provide transient protection for driving relays, motors, etc. So the voltage drop across the segment when off is your anode voltage less 50v, or 170 - 50 = 120v. Not enough to reach the strike voltage, so the segment doesn’t illuminate. And unless the segment is actually illuminated, zero (or virtually zero) current will actually flow. I think it was Dave Jones who used these in his Nixie project, although granted they weren’t B7971 tubes.
@ No, but principle is the same regardless of tube type. I have used 6B595 to drive my IN18 clock and B7971 FLW projects here for several years. About the only downside is they use 5v logic, and so can be driven directly from lower voltage microcontrollers (I use 3.3v ESP here). But if Fran is using a 5v Arduino that issue would go away.
2:00 “Complicated driver”? It’s literally just a single current limiting resistor on each segment, as indeed you later say! Using a suitable open collector shift register as the driver and you can get rid of those transistors. My B7971 never get even vaguely warm, let alone hot. As others have said these tubes will last for decades or more if correctly driven. They certainly don’t need to be driven hard for domestic use, as was probably the case in their original installations.
You need TWO current limiters based on the B7971 datasheet information. As you mentioned, each cathode needs the current limited; I use an NPN transistor with a resistor at the emitter, to limit the current. The base is driven from a known voltage, say 3.3V, and the base-emitter voltage is 0.7V. So, to control the current, select a resistor value RE=(3.3-0.7)/current. The smallest segments require 4mA, so their resistor will be ~650 ohms. The remaining segments need 5.0,5.5, or 6.0mA so their resistors will be 520, 470, or 430 ohms.
Now, the second current limiter is required because the max current of ALL cathodes combined must be 21mA or less. So, at the anode side, you need a similar current limiter. I use a PNP transistor.
It sounds like a lot of extra cost and hassle, but electronic components are essentially free and in unlimited supply, whereas B7971 tubes are getting very expensive (typically 200USD each), and increasingly harder to find. It will take time to find a set of 6. Glad I got mine (and a few others) years ago.
All true. I did assess all the options, particularly with the various differences in optimal segment current, as the slightly shorter cathodes need more limiting, etc... I will get around to rebuilding it - and I have many other tubes, so I wasn't worried about putting wear on this one, but my bigger project requires an array of proper driver boards.
Fran....If these tubes are driven correctly and to spec they will last forever. I have (5) 6 digit and (2) 4 digit B-7971 clocks run that have been running essentially 24 x 7 for well over 12 years with no issues, degredation or loss of brightness. Considering these tubes are pulls from the stock brokerage displays and who knows how many hours they had on them before those displays were de-commissioned, I consider them some of the most robust Nixies out there. I have no issues with the tubes getting hot or even warm for that matter. I think once you have your design with individual cathode current regulation you'll see a big difference. Finally, I know of others that have been running these tubes 24 x 7 since the mid 70's without issue. Good luck and can't wait to see what you come up with.
You are right!
My Datasheet says max. 21mA current if all segments are lighted.
This is in theory a dissipation of 3.57W at 170V for the whole tube.
In real world it will be much less because the voltage drops if the segment has ignited.
There are different currents for the individual segments.
For a long live it will be good to run not more than 4 mA through each Segment.
I'm going to have to look up a picture of that stock market sign. I never even knew tubes like this existed.
Fran I guess you got carried away a little with the helium. Never heard any story that vacuum tubes get contaminated with helium. It's usually the cathode losing emission over time. Going on air is most likely due to failing seals I would think. The getter will try to counter this but there is only so much available....
About 50 years ago, the local Burstein-Applebee store carried a bit of surplus and I bought a bunch of those boards for $5 each! I built a nixie clock out of 6 tubes and it ran for over 20 years before I replaced it with a smaller nixie clock using the Russian tubes. Some day I may have to plug the big clock back in to see if it still works.
The data sheet gives you the current limit per segment and you really should follow that. One resistor per segment is more components on the PCB but the tube will "last forever" with a better even brightness. If the white area is growing significantly you are likely over driving in general. I have a clock running for years and seen no change in the tubes and they are never hot or even warm to touch.
Standard numeric nixie tubes will also specify current drive per numeral as the different sizes have different demands, most designs also ignore that requirement.
Helium has left the room
♪ ♫ ♬dun dundun ♪ ♫ ♬ ........it's fran... ♪ ♫ ♬......in the lab.
franlab.
It reminds me of my first LED 7 segment display where I only had one current limit resistor, and the display dimmed more when more segments lit. After I redid the circuit to have each segment limited, the display remained bright no matter how many segments lit. The best bright display is when you can multiplex the display, and run more current for a short time, and your vision persistence makes up for the off time.
Do you have the background grounded or floating? If it is grounded, perhaps it attracts the electrons, which causes the bleaching?
As far as helium goes, it is able to leak past seals and threaded joints. which means it doesn't have to travel through material except under the high pressure conditions.
I'd look into controlling them with two SN75468 drivers for the cathode side.
Or a TPIC6B595 shift register. Considerably simplifies the wiring back to the microcontroller as you can drive a virtually infinite number of daisy-chained displays without the need for any extra microcontroller outputs, or any multiplexing of the tubes. Then just a single currently limiting resistor on each segment, rated according to the segment size and recommended maximum current as given in the B7971 datasheet. This is my much preferred way of driving these tubes. Plus add a boost converter to generate the 170v DC, and then you can run everything from a single 12v DC supply.
@@stuartmcconnachie cool, I'm only worried about the max Vds of this chip being 50V. Seems a bit low for me. That's why I'd combine a 74HC595 with SN75468.
@@KeritechElectronics Works just fine. The outputs of the 6B595 are diode clamped at the 50v to provide transient protection for driving relays, motors, etc. So the voltage drop across the segment when off is your anode voltage less 50v, or 170 - 50 = 120v. Not enough to reach the strike voltage, so the segment doesn’t illuminate. And unless the segment is actually illuminated, zero (or virtually zero) current will actually flow. I think it was Dave Jones who used these in his Nixie project, although granted they weren’t B7971 tubes.
@@stuartmcconnachie he definitely used IN-12 tubes for his subscriber counter silver button project, but I'm not sure if he did other nixie projects.
@ No, but principle is the same regardless of tube type. I have used 6B595 to drive my IN18 clock and B7971 FLW projects here for several years. About the only downside is they use 5v logic, and so can be driven directly from lower voltage microcontrollers (I use 3.3v ESP here). But if Fran is using a 5v Arduino that issue would go away.
2:00 “Complicated driver”? It’s literally just a single current limiting resistor on each segment, as indeed you later say! Using a suitable open collector shift register as the driver and you can get rid of those transistors. My B7971 never get even vaguely warm, let alone hot. As others have said these tubes will last for decades or more if correctly driven. They certainly don’t need to be driven hard for domestic use, as was probably the case in their original installations.
Hi Fran, with all the numbering malarky, how did you stop your brain turning to porridge?
Based on some comments it seems that perhaps I did not!
Thanks Fran..
IT IS called defusion, what the Helium does.
Fran, you playing with that noise machine machine in your recent videos is VERY irritating; I hope you discontinue that trend.....
You could probably alter the base resistors and limit the current using amplification of the transistors.
❤️🔥FRAN❤️🔥
interesting 👍