An odd, but interesting observation of a weak B7971 tube

186 views
Skip to first unread message

Jim KO5V

unread,
Oct 7, 2023, 1:26:27 PM10/7/23
to neonixie-l
I have a B7971 that doesn't want to light up - the ends of some segments will glow,  but none will illuminate along their entire length. I have had tubes in the past that were probably a bit gassy, and they cleaned up after running for a few minutes to a few hours.

So, I put it into the "10 second" position of my Mod 6 clock, and ran it over night with no improvement. However after I had been up an hour or so, I noticed that the tube's segments were lighting up along about 3/4 of their lengths, and also that the low morning sun was shining on the clock. This state lasted for as long as the sunlight was on the tube - sometime after light moved off of the clock, the tube went back to it's wounded state. 

This morning I watched as the sunlight moved toward the clock, and as soon as the light hit the clock, the tube immediately recovered, and performed like it did yesterday. However, that state lasted for about 3-1/2 hours after the sunlight moved away. I then shined a flashlight on the tube, and it recovered as long as the light was there. 

I guess this is some kind of photo-voltaic effect. I have a bit of education in basic physics (for engineering), and this baffles me - but it's cool!

I have a reading lamp that puts out a full spectrum, so I will put that on the clock today and see what happens. I may also set up a test to run the tube at a bit higher voltage (180-ish V). It may never recover completely, but I think this is a fun exercise.

Anyway, and ideas are welcome. I may not be able to save the tube, but I might actually learn something.  Thanks.  Jim

guus.a...@wolmail.nl

unread,
Oct 8, 2023, 9:01:55 AM10/8/23
to neoni...@googlegroups.com, Jim KO5V
Hello Jim,
 
If you have a source of UV-light, that will do a good job.
Maybe an UV-Led beneath the tube will do the trick....
 
BR/
Guus
Op 07-10-2023 19:26 CEST schreef Jim KO5V <jr...@earthlink.net>:
--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/d2784c4a-c76f-4ae6-b0b1-86f14185bb22n%40googlegroups.com.

gregebert

unread,
Oct 8, 2023, 12:13:47 PM10/8/23
to neonixie-l
Can you do some bench-testing to see if the segments fully light with a bit more voltage ?
I collected current-voltage (I-V) data on all my 7971's so I can check them for aging effects.

Paul Andrews

unread,
Oct 8, 2023, 1:42:40 PM10/8/23
to neonixie-l

I’ve had some of these tubes that require a higher than normal voltage to even trigger. Most people drive nixies at 170V, but this can be a marginal trigger voltage for some tubes and completely insufficient for others such as the GR10G https://www.nixies.us/bwg_gallery/gr10g/

Shining light on a tube ionizes the gas inside, the brighter the light, the more ions. Tubes used in dark environments can take longer to light up. Once they are lit, the number of free ions in the gas is usually enough to keep it going. Several approaches were used to circumvent this. Some tubes have an extra cathode driven with a very low current that is always on. Some tubes were doped with Krypton 85, a radioactive gas, e.g. https://www.nixies.us/bwg_gallery/122p224/

Nicholas Stock

unread,
Oct 8, 2023, 2:13:23 PM10/8/23
to neoni...@googlegroups.com
I've noticed a few 7971's needing a couple of hours to fully illuminate after which they're apparently very happy in the long term if in operation. Is this sign of a very small leak or something else? I'm not sure what to make of the 'gassy' term to be honest....were 7971's doped with Hg?

Nick

Sent from my iPhone

On Oct 8, 2023, at 10:42, Paul Andrews <pa...@nixies.us> wrote:



Jim Faulkner

unread,
Oct 8, 2023, 3:44:01 PM10/8/23
to neoni...@googlegroups.com

Hi Nick and Ira, and guus, and gregebert and Paul and anyone else I have missed,

 

Thanks for the replies. The light affecting the ionization was what puzzled me, but a couple of you explained that, and it makes sense. I have a way to test the tube on the bench, so maybe I can bring it back that way.

 

I used the word "gassy" because that was the term we used working with big ion lasers (40+ years ago). I don't mean a leak, because any leak should be fatal.

 

The laser tubes used getters that we heated with RF induction heaters to evaporate barium. The barium would trap any leftover impurities, and this was the last step in tube manufacturing. As the tube ran, material would sputter off of the cathode, and maybe other parts (there was ionized Ar and/or Kr inside of the tube, it was hot!). This would trap small amounts of gas. The getter material also continued to find impurities to remove. All of this would slightly lower the tube's pressure. 

 

The laser had an automatic gas filling system to keep the tube's pressure within a very narrow range, because a small deviation in the tube's pressure would greatly affect the laser's power output.

 

If a tube that had been running long enough to have had some gas added was then left unused for a long time, some of the trapped gas could escape as it sat. This increased the tube's pressure very slightly - getting "gassy" - and that would affect the laser's output power. Running it for a while could cause it to again trap some gas, usually lowering the pressure enough to recover its original power output.

 

That was how it was explained to me buy much smarter people, and it made sense.

 

Over the years, some NIXIE tubes that I've had in storage have shown weak ionization. Running them for a while has cleaned some of them up, so when I have had a weak, but working NIXIE, I assumed that it is the same issue that I saw with the lasers many years ago. Maybe I'm wrong.

 

Thanks again for the help.  Jim

 

liam bartosiewicz

unread,
Oct 8, 2023, 3:52:40 PM10/8/23
to neoni...@googlegroups.com
Another note about firing voltage is that in general, using higher voltages for nixies, ie. >200V increases lifespan, assuming the proper anode resistor is used.

On Oct 8, 2023, at 11:13 AM, Nicholas Stock <nick...@gmail.com> wrote:

I've noticed a few 7971's needing a couple of hours to fully illuminate after which they're apparently very happy in the long term if in operation. Is this sign of a very small leak or something else? I'm not sure what to make of the 'gassy' term to be honest....were 7971's doped with Hg?

Audrey

unread,
Oct 8, 2023, 4:02:27 PM10/8/23
to neoni...@googlegroups.com
I thought I had heard/read that lower voltage was better, though a cursory search suggests that the rate of sputtering is nearly directly proportional to current, so that makes sense, the only reason aside from arcing that I could think of would be making the internal wires glow

gregebert

unread,
Oct 8, 2023, 5:51:02 PM10/8/23
to neonixie-l
My first nixie clock with 5092 tubes has been running since 2011. I use +340VDC for the anodes, and of course the anode resistors are appropriately increased to give the optimum 2.2mA of current.

So, why did I choose 340VDC when the mains here in the US are 120V RMS ? This clock has no transformer, so either I would use approx +170VDC from rectifying the line directly and risk being too close to the minimum voltage, or just use a voltage-doubler and deal with the slightly higher wasted energy.

Rob B

unread,
Oct 9, 2023, 11:22:05 AM10/9/23
to neonixie-l
When I was a kid [~50 yrs ago], I had several power strips with NE-2 indicators, where the indicator starting flicker and then eventually failed.  I noticed that when sunlight hit the NE-2s, they lit more solidly: flickering converted to solid, and off converted to flickering.  I always assumed it was the energy from the sun's photons that was adding just enough joules to the equation to push failing bulbs back over their ionization threshold.  Same effect, probably?

...R

gregebert

unread,
Oct 9, 2023, 2:35:41 PM10/9/23
to neonixie-l
None of my 7971's has aged enough for the segments to get flaky; my clock went into operation in 2017. Knocking on wood.....

I've seen 2 types of incomplete segment/numeral illumination. On my IN-18's, it's cathode poisoning for the months-digit and that clears-up within a few days of the new month. Obviously my depoisoning algorithm isn't perfect, yet oddly enough the years digits dont show any poisoning so there is some factor here with the tube. BTW, it's a 14-tube clock using MM-DD-YYYY HH:MM:SS format.

The other incomplete illumination appears to be end-of-life for the tube, because no amount of voltage or current has been able to fix the tube. You get what you pay for when you buy used nixies online; many are OK, some are not.

Jim KO5V

unread,
Oct 9, 2023, 3:23:25 PM10/9/23
to neonixie-l
OK, light therapy may work! (or it's a coincidence...).

History recap: The tube would glow in some areas, but not light up any segments. I ran it for a total of 48 hours in two different clocks with no improvement. Then I noticed that sunlight shining on the tube caused about 60% of the segment lengths to illuminate. When the light moved, the improvement went away. The next day after the sun moved away, the tube ran in its 60% recovered condition for about 3 hours before crapping out.

Yesterday, I illuminated it for about 12 hours with a reading light with a fluorescent bulb that allegedly produces a spectrum similar to that of the sun (just FYI, on the spectrum - it may not matter). The segments immediately recovered to about 60%. However, with time, the tube began to slightly improve. I briefly turned the light off several times during the day, and watched the segments degrade, but there was some improvement each time. At midnight, I turned the light off for the night, and the tube was probably at 80-90%. This morning, the tube is working 100%. I moved it to the other clock, and it is also working there.

As suggested, I think the photons from the lamp helped with the ionization just enough to allow the tube's segments to eventually light up enough to heal themselves. Maybe it would have eventually recovered by constantly running it without the lamp, but I saw no improvement in 48 hours, so it might take a very long time - if ever. There might be something real here (or not). Just FYI. This is not very scientific, but it's still interesting!

Jim

Mac Doktor

unread,
Oct 9, 2023, 4:09:04 PM10/9/23
to neonixie-l
On Oct 9, 2023, at 3:23 PM, Jim KO5V <jr...@earthlink.net> wrote:

As suggested, I think the photons from the lamp helped with the ionization just enough to allow the tube's segments to eventually light up enough to heal themselves.

I don't know about healing but light helps the gas to ionize. I would assume that the shorter the wavelength the better as the energy increases. Try using a blue or near-ultraviolet source. Ultraviolet would also work but may not penetrate the glass.

Ionizing radiation will work even better because it...you know. Try using Thorium as a beta source—that should penetrate the glass. Gamma sources are even better. I just happen to have some Radium (clock dials, TR cell) and Cesium-137 around the house. Sadly, the alpha particles from

While you're experimenting, try firing up a Tesla coil that's too far away to fully ionize the gas by itself. Or hold it near a plasma sphere.


Terry Bowman, KA4HJH
"The Mac Doctor"

https://www.astarcloseup.com

"Every kid starts out as a natural-born scientist, and then we beat it out of them. A few trickle through the system with their wonder and enthusiasm for science intact."—Carl Sagan, Psychology Today, 1996

jf03...@gmail.com

unread,
Oct 9, 2023, 5:13:46 PM10/9/23
to neonixie-l
> On Monday, October 9, 2023 at 11:35:41 AM UTC-7 gregebert wrote:
>  None of my 7971's has aged enough for the segments to get flaky; my clock went into operation in 2017. Knocking on wood....

On my first clock, my original version  did have partially illuminated segments when I used +170vdc.  There were no partial segments after I increased it to about +190vdc.  The next two clocks started with +190vdc and never showed partial segments.

The first clock was built in 1975, and the next two were built in 1977, all using dollar specials from Meshna, presumably NYSE pulls with thousands of hours of previous service. None of the tubes have failed or shown significant degradation.  Other components have failed (clock IC, HV transistor, power transformer), but there have been no tube failures after over 300k hours.  There has been some darkening on the glass envelope (sputtered cathode?) and a slight color shift to a more pinkish glow.

Mac Doktor

unread,
Oct 9, 2023, 5:44:56 PM10/9/23
to neonixie-l
Incomplete sentence...


On Oct 9, 2023, at 4:09 PM, Mac Doktor <themac...@gmail.com> wrote:


Sadly, the alpha particles from

...an alpha source such as a household smoke detector won't make it though the glass.


Terry Bowman, KA4HJH
"The Mac Doctor"


“...the book said something astonishing, a very big thought. The stars, it said, were suns but very far away. The Sun was a star but close up.”—Carl Sagan, "The Backbone Of Night", Cosmos, 1980


Reply all
Reply to author
Forward
0 new messages