Nixie tube failure - not outgassed, but does not maintain ignition?

[Kevin A.]

Hi all,

A friend of mine, also purveyor of cold cathode and other unique displays, has collected two 8971 nixies that seem to be having an odd failure. 

When powering the tube on, it ignites normally and begins to operate. However, after a period of 30 seconds or so, none of the cathodes can stay ignited and they fizzle down into a state of very faint partial illumination.

Has anyone experienced this before? The most common failure mode has to be outgassing, but the tube is still full of neon as far as I can tell. It just does not stay ignited for more than a period of seconds. Can this be remedied by applying some kind of high current or high(er) voltage pulse the same way cathode poisoning sometimes can?   

[gregebert]

Can you monitor the voltage and current of the tube while it's running ? I cant help wondering if higher voltage will keep it on.

I made my own nixie bench supply which has variable voltage and current, so I usually set the voltage above 200V, and limit the current to the datasheet value, which is 4.5 to 6mA per segment for the 8971.

Do all segments show the same behavior, or just some of them ?

[Kevin A.]

I will definitely make some measurements once he has a chance to hand them over. From what he told me, the whole tube just self extinguishes during this phenomenon. Very odd. 

--
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/9ca0e0f5-53b7-49c7-a654-a817ca45bfd9n%40googlegroups.com.

[Nicholas Stock]

I know this is a very obvious question, but the power supply your friend is using isn't doing something strange instead? That's certainly odd behaviour for a nixie...

Nick

Sent from my iPhone

On Nov 24, 2020, at 20:14, Kevin A. <ka...@scarletmail.rutgers.edu> wrote:



To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/CALcVLK%2Biu8W%2BbKELTYenrkvgMHeGp%3Dxk4_PbrgK3VnQUmKP_jQ%40mail.gmail.com.

[Kevin A.]

He picked up the two problem tubes as spares for his black emerald 8971 clock. The original 6 8971 tubes work just fine in the clock, no matter what socket they are in. Only when one of these two problem 8971s are popped in place, does this issue occur. This leads me to believe it is a problem with the tube rather than the psu or driving circuit. 

I'll ask for a video and will definitely be making measurements once I get a hold of them for experimenting. 

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/63B7ED85-D391-4AC2-9170-274EE3B10DE5%40gmail.com.

[Nicholas Stock]

That would appear to be a logical conclusion....I wonder if there's a poor anode connection that separates after heating? Odd for sure....looking forward to your results! They're rare tubes....

Sent from my iPhone

On Nov 24, 2020, at 20:25, Kevin A. <ka...@scarletmail.rutgers.edu> wrote:



To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/CALcVLKLDxVKzDu7bAEgLLhBNWEeFbdjKi0NnBwc%2BqeAsWQcs1A%40mail.gmail.com.

[Paul Andrews]

It can be a result of out gassing. I had a ZM1082 that got damaged and slowly  outgassed. Towards the end, this is exactly what it did.

A neon test using some high voltage source doesn’t tell you what the pressure is in there, or how much other gas there is. I have another tube that passes this test, but will nit light up at all no matter what voltage and current I use. Conversely I have some planar displays that do not pass that test, but light up just fine.

[Nick Andrews]

Has anyone tried to re-gas these tubes?  Risky for sure but maybe doable with the right equipment.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/aa52987e-8333-4920-b3e3-1ab33d2a2c01n%40googlegroups.com.

[Christian Riise Wagner]

I suppose it could be regassed, though it probably wouldn't be easy as the evacuation port is typically close to the rest of the glass envelope. The problem is that if the tube is leaky, it's just a matter of time before it will fail again.

[Nick Andrews]

It would certainly be an experiment.   You could slather a little high vacuum epoxy around the pins to help seal the metal to glass joints if that was suspected. 

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/db9d1a29-4ab4-4ee8-be9e-641f431335cdn%40googlegroups.com.

[Kevin A.]

Performed some experiments tonight with unexpected results.

My friend sent me the two 8971s in question. One is a fine grid, one is a coarse grid tube. Both reportedly exhibit the same "partial illumination" problem when running in the black emerald clock. So I began by cleaning and labeling the two tubes. I then took my black emerald, removed one of my good 8971s and installed a faulty tube in the rightmost "seconds" position. Sure enough, partial glow as expected. No matter how long the tube ran, it was always partially illuminated. 

My first plan of attack was to use the source measure unit (SMU) to apply repetitive high current pulses to all the cathodes. The 8971 datasheet describes the maximum current of all cathodes simultaneously to be limited to 14ma. I set the SMU to operate in pulsed mode and zapped the cathodes in pairs of two (anode connected to positive, two cathodes connected to negative). The pulse per set of two cathodes was delivered as follows:

75 milliseconds period (between pulses)

50 millisecond duration (per pulse) 

200 pulses total

0-200 volts peak to peak

current compliance: 14ma

After making sure each cathode pair (randomly chosen) got a cycle of 200 pulses, I reinstalled the tube into the black emerald. During the first 30 seconds, the segments began to come back to life. After a minute, the questionable tube was illuminating normally. I could not believe it. Did the cathodes need to be "revived" somehow after a prolonged (30+ years) period of inactivity?

I set the experimental tubes aside again for an hour, to let them "cool" off, and make sure the effect was not temporary after having warmed them up with the high current pulsing. Sure enough, I inserted them into the black emerald and they resumed functioning as normal. No more fizzling out. 

As exciting as this is, the results are still very preliminary. I will provide further updates once they have been mailed back to the original owner and he has a change to try them out. Then time will tell if they are truly resuscitated!

 20201214_234701.mp4

 20201215_000459.mp4

 20201215_002725.mp4

 

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/CAJD4P-hKL%3D59wocKwazh5vWO8kXa0CjnCRB8Tj3OpWRBd5jQsw%40mail.gmail.com.

[Kevin A.]

Another problem I was able to fix: two of my own 8971s had partially illuminated segments. The same shock therapy cleared them right up! Not sure if cathode poisoning was the original culprit in all cases, or something else. 

[Christian Riise Wagner]

Kevin, I had a similar experience with an IN-8-2, though I don't have an SMU, so it was much less controlled. The tube was pretty dead, with only the upper half of the cathodes lighting at nominal current. The consensus seems to be that this indicates a leak.

Anyway, since it was pretty much dead, I thought I might as well have some fun abusing it and see when it gives up. I connected it to an NCH6100HV supply without a resistor and slowly turned up the voltage. The tube got all the way to 40mA, at which point it was a bright pink colour and flashes (arcing?) started to show somewhere inside the tube. I turned it off, then turned it on briefly a couple of times at this current. Afterwards I tried running it at normal current, and that cathode was fully and evenly lit! I don't remember the state of the other cathodes but I did keep the tube with a label, so I might  be able to find it again.

[Kevin A.]

It certainly would be interesting to see if the adjacent cathodes were poisoned or if the tube functions as expected. I know cathode sputtering is a part of nixie tube operation, but can a cathode really "evaporate" in a decent mercury doped tube if the anode current is kept in check?

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/35f357a3-5bf0-40e7-9b3c-1fe0c6b686een%40googlegroups.com.

[Dekatron42]

Some old datasheets mention that you need to exercise each digit for some time before they will illuminate fully after long time in storage, I've seen this time being mentioned as 24 hours when the Nixies have been stored for more than a year. The same goes for the glow that sometimes appear on connecting wires which they say will disappear when the glow moves onto the digit when the Nixie has been exercised. This can usually be seen easiest in Russian datasheets for Nixies but I have seen it in other datasheets too.

/Martin

[H. Carl Ott]

No idea if it is related to cathode "poisoning". But yeah. Seen this effect on a lot of older russion tubes that have been sitting for decades. 

 Especially seen it on IN-9s and less so on IN-13s. Used to have to "burn" then in at elevated currents for a while before they'd behave.  

Note that it did not always work. Some tubes never came back.  I wonder what causes it on tubes that have basically never been run. Contamination?   

 

carl
--------------------------------------------------------
Henry Carl Ott   N2RVQ    hcar...@gmail.com

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/73b08d5f-9f04-45d1-a983-058e194296a4n%40googlegroups.com.

[Kevin A.]

One theory in my head is that the cathodes could potentially act like getters if the tube doesn't have one? I don't think these 8971s have a getter. If contaminants outgas over a long time (even if the tube is not in service), maybe they are attracted to the cathode metal and form deposits that can prevent glowing?

Maybe Dalibor has some insight into this, being that he has forgotten more about nixies than I will ever know. 

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/CAP_hP_TuY3orVsnTK2P6xSkHC9w6aWJCXOE-f595XoWwFEbtJQ%40mail.gmail.com.

[MichaelB]

Yep, they are a finicky tube. I just rejuvenated one of my 8971's last night that had a partially lit carat. About an hour of constant over current fixed it right up and the segment now works perfectly. But then again, I have tried in vain with a couple other 8971's and no matter how long I exposed them to high current, the bad segments did not come back. It sounds like you're on to something here though Kevin, applying pulse current blasts vs. a constant bombardment.

[Kevin A.]

Glad that you've had some success with reviving dead segments. I noticed during my experiments that some segments were more difficult to revive than others. I had to do a few rounds of pulsing, and between each I would check the tube in the black emerald to see how it looked during "normal" operation. Fortunately I didn't strike out on the 4 tubes I worked on, so maybe we've found a way to bring them back. If you've got any problem tubes laying around, feel free to send them over. As difficult at the 8971 is to find, every revived tube put back in service is worth the effort! 

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/65bb9d52-8496-49b1-b5df-cc0a853d1affn%40googlegroups.com.