Is it possible to identify a single nixie?

[newxito]

Is it possible to identify a single nixie tube by some unique characteristics?

Is it possible to detect a nixie tube change in the clock?

I ask because I would like to add some statistic functions to my nixie clocks. The first thing I would like to add is the total “up-time” of the clock and the total up-time of the nixies (different because of PIR and software defined down times). It would be nice to do that for every single nixie, but I think there is no easy way to do that.

I also want to add an up-time counter to my ZM1350 und B-7971 sockets. 

Maybe my favorite nixie manufacturer could add a 30-cent microcontroller to the base of his great tubes with the serial number, production date and up-time counter :-)

[newxito]

After thinking a bit more about it, I came to the conclusion that it is not worth the effort trying to detect a tube change with the microcontroller. If I have to replace a broken nixie with a NOS one, I can just manually reset the up-time counter with the android app.

[gregebert]

For each of my expensive tubes (b7971, R|z568m) I measure and plot the I-V characteristics. There aren't any serial numbers on b7971's so I just write one on the backside.

If the tube fails later on, I check the I-V characteristics to see if anything changes; so far only 1 b7971 has failed when segments shorted together. None of the unaffected segments showed any measurable changes.

I definitely need to add the runtime datalogging.

[newxito]

That’s very interesting.  Is there a measurable difference between the I-V characteristics of two working B7971?

[gregebert]

Yes; for a given segment, there are tube-to-tube variations in segment current for a given voltage.

The data below is for 8 tubes, for segment #1.

At 200uA, the voltage varies from 127 to 138V. At 6mA, the voltage varies from 169 to 181V

Using just a resistor for setting the current will result in large variations across tubes and supply-voltage. That's why I use current-regulators on each cathode. It's probably overkill, but given the ridiculous pricing for b7971's, I'm not taking any chances. I have no spares for my 8-tube clock; just 2 tubes that have a shorted or dead segment.

I have no idea what happens over time, but I suspect that as tubes age the current, hence the brightness, decreases for a given voltage. In a few years, I will retake the data and see what has happened. This clock went into service in 2017.

0.2 0.5 1 2 3 4 5 5.5 6          <-- Current

130 133 138 145 152 159 167 170 172      <--- Tube 1

135 138 143 151 158 166 174 178 181

134 137 142 150 158 166 174 178 180

138 138 140 145 151 158 164 169 172

137 138 142 148 156 163 172 176 180

135 136 140 147 154 162 171 174 177

131 134 138 144 151 158 166 170 173

127 130 134 141 147 155 162 166 169     <--- Tube 8

[Richard Scales]

For what it's worth I successfully 'repaired' a B-7971 that had two segments shorted together following helpful advice on this forum. It process seems dramatic but if it means a valuable tube gets back in to service then you're up on the deal!

Search on this forum for details but essentially it involved exerting forces on a tube that you would perhaps otherwise view as crazy - lots of banging it it about - though fortunately, in my situation, a couple of short taps on to a moderately soft surface did the trick and I was very pleased. 

YMMV

Richard

[Richard Scales]

Another thing I do is have the clocks on only when motion is detected. I have clocks featuring all the tube types that you have mentioned and their actual 'on' time is probably much less than 1hr a day depending on activity around the house. 

Given that segments are switching on and off all of the time anyway - is leaving them off for most of the time a bad thing?

... and if no-one is there to see the clocks - what is the point of having them on?

... or is it something like - if a tree falls over in a forest and no-one is there to hear it - does it make any sound? (My firm believe is that it does!)

I know that many will have their B-7971's and ZM1350's running either 24/7 or perhaps off during the small hours only so, in my opinion, my tubes are seeing considerably less action than many which makes me feel that I am making some effort to preserve them as long as possible.

Richard

On Friday, 15 May 2020 11:08:16 UTC+1, newxito wrote:

[Richard Scales]

It worked well for me.

Thank you.

Sent from my Huawei phone

-------- Original message --------
From: Michail Wilson <M...@Michail.com>
Date: Sat, 16 May 2020, 05:27
To: neoni...@googlegroups.com
Subject: RE: [neonixie-l] Re: Is it possible to identify a single nixie?

Heh,

 

I think most of the posts on how to do it were written by me.

 

I’ve repaired at least 20 tubes.   I’ve only busted one.

 

Michail Wilson

206-920-6312

--
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/59ef7012-8cdf-42a2-bb05-982405df3ed0%40googlegroups.com.

--
You received this message because you are subscribed to a topic in the Google Groups "neonixie-l" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/neonixie-l/CfyklEiKZKM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/MW2PR0102MB3435FAED15689A35984A9CF082BA0%40MW2PR0102MB3435.prod.exchangelabs.com.

[Robert G. Schaffrath]

I had only one tube that had shorted segments: 4 & 5. I was able to fix them last fall with the banging method but due to having played with the tube over the last 40 years, there were other issues with it. Plus, after banging it to remove the short, I noted that segment 2 is now flaky and not always lighting up. I have debated putting it up on eBay as it still glows, I do not use it and it would be a less expensive way for someone to have a B-7971 to play with that is not fully functional with segments 2, 4 and 5 essentially being non-functional. I would then buy myself an IN-18 to play with.

Robert

[Terry Kennedy]

On Friday, May 15, 2020 at 9:51:45 AM UTC-4, gregebert wrote:

For each of my expensive tubes (b7971, R|z568m) I measure and plot the I-V characteristics. There aren't any serial numbers on b7971's so I just write one on the backside.

If the tube fails later on, I check the I-V characteristics to see if anything changes; so far only 1 b7971 has failed when segments shorted together. None of the unaffected segments showed any measurable changes.

B7971s are amazingly robust. The MOD-SIX in the room where I'm typing this reports 76,571 tube-on hours (nearly 9 years) and is on its original tubes (the CPU, PSU and case have been replaced over the years, with the tube-on hours copied to the newer CPUs). I've never had one that worked when I got it fail on me. In around 100 I've purchased over the years I've had 4 or so DOAs - 2 cracked, one with an unfixable short and one with a poisoned cathode I couldn't clean up.

[jf...@my-deja.com]

Based on my experience, you should expect at least another 300k hours even if you "overdrive" them.   That said, I don't remember anyone reporting a natural death from old age and excessive usage.

[jf...@my-deja.com]

Over that voltage range, I would expect three branches on the IV.  Did you also track the firing voltages and extinction currents?  Dark rooms vs well-lit?  Ambient temperature?

[gregebert]

I had a very difficult time trying to measure the firing voltage, probably due to photons and other particles whizzing around the workbench. Typically, it was about 10-15 volts higher than the smallest measurable current where I could see any hint of glowing. The value varied from reading-to-reading.

I recall that I could not repeatedly/predictably measure currents below 100uA, because the segment would just extinguish and the current went to zero.

Whenever I get my last 2 R|Z568m tubes, I'll spend more time on ionization measurements. Right now, those tubes are still waiting on a dock or tarmac to get shipped to the US.

[jf...@my-deja.com]

In addition to photons and cosmic rays, illuminating other elements can also promote the firing (isn't that the reason for the "keep alive" on the Panaplex?). 

It has been a long time since I did it, but ISTR that you could stay on the negative resistance branch by aggressively limiting the current.  In the olden days, that was using bigger resistors, but nowadays I would build a transistorized variable current regulator.