Author Topic: Vintage calculator project: how variable are neon indicator lamps?  (Read 760 times)

0 Members and 1 Guest are viewing this topic.

Offline rea5245Topic starter

  • Frequent Contributor
  • **
  • Posts: 585
  • Country: us
How much do small neon indicator lamps vary in their striking and extinguishing voltages, and how much does they vary over time? Are modern bulbs more consistent than ones from the '60s?

I ask because I'm intrigued by a Hackaday article about a calculator project from the 1960s (https://hackaday.com/2024/06/16/early-computer-kit-really-just-a-fancy-calculator/). I'm toying with the idea of making it.

The magazine article with the instructions (link can be found in the Hackaday article) talks about buying a bunch of neon bulbs, running them for 48 hours, then characterizing their striking and extinguishing voltages. Apparently, there's variation even among bulbs with the same part number, and there's variation over time (at least over an initial burn-in time).

It's critical to the working of the calculator because as near as I can tell, the circuit is using the state of the neon bulb as memory for the current value in the calculator.

My concern is that I'd go through all the effort to make this, and then the bulb threshold voltages will change over time anyway, and the calculator will stop working.

Is that a valid concern? If I buy modern bulbs, will they be more uniform?

Thank you.
 

Offline xvr

  • Frequent Contributor
  • **
  • Posts: 371
  • Country: ie
    • LinkedIn
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #1 on: June 29, 2024, 10:52:48 am »
Counting on neon lamps is not a best idea. Decatron will be much more suitable.
https://en.wikipedia.org/wiki/Dekatron
 

Offline rea5245Topic starter

  • Frequent Contributor
  • **
  • Posts: 585
  • Country: us
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #2 on: June 29, 2024, 12:08:37 pm »
Decatron will be much more suitable.

I'm not sure "suitability" is the goal here. ;-)
 

Offline CaptDon

  • Super Contributor
  • ***
  • Posts: 1886
  • Country: is
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #3 on: June 29, 2024, 02:34:43 pm »
Neon bulbs and devices vary all over the place. Some of the lowest strike voltage were the old NE-51 bayonette bulbs that looked similar to a #47 bulb in size and shape. Ambient light will change things by a large amount as will local sources of radiation. In the early days of Nixie style voltmeters and counters it was discovered the nixies wouldn't ionize in underground caves. Some nixies had a small dot of radioactive paint applied externally centered on the base and others had an internal isotope. Some neon devices have a 'keep alive' cathode that maintains a very small ionized glow so that the rest of the device will work correctly. Old neon devices get lazy and require a higher strike voltage or tend to flicker around the cathode. Also there is the transfer of metal due to ion bombardment. You can see this on old neon lamps and on a nixie where the deposits reflect like a mirror and look like a getter. The transfer of metal on nixies is so bad with high current operation that numerals will short to each other or the tube will have a bridging resistance between the cathodes and the anode and the tube won't light at all. Your calculator will be more stable if it operates near a steady source of light and all of the lamps receive the same amount of light. If the neon bulbs haven't been used for a few days or even weeks the striking voltage will start to increase again. It is better to have an available strike voltage source that is 1.5X the needed strike voltage and then use a higher value of current limiting resistor. If the voltage source is to low some bulbs may not ionize.
Collector and repairer of vintage and not so vintage electronic gadgets and test equipment. What's the difference between a pizza and a musician? A pizza can feed a family of four!! Classically trained guitarist. Sound engineer.
 

Offline rea5245Topic starter

  • Frequent Contributor
  • **
  • Posts: 585
  • Country: us
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #4 on: June 29, 2024, 03:16:51 pm »
CaptDon, that sounds bad.

Would it be fair to say that if anyone actually made these calculators, they probably ended up disappointed by the reliability of the finished project?
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14472
  • Country: de
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #5 on: June 29, 2024, 05:15:03 pm »
One aged a little the neon tubes should not change that much over time. Especially if one has a stable supply voltage, it should work reasonable OK with selected parts. So it will be some work to measure and sort the parts. To some degree one may be able to trim externally if really needed - this may be easier than unsoldering the neons.

The first thing would likely be to test a small coutner, like 1 decade to get some experiance. For just a demo project 2 or 3 digits may be good enough.
 

Online tggzzz

  • Super Contributor
  • ***
  • Posts: 20002
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #6 on: June 29, 2024, 05:52:52 pm »
Full of useful circuits and tables about "Using and Understanding Miniature Neon Lamps"
https://www.effectrode.com/wp-content/uploads/2020/05/using_and_understanding_miniature_neon_lamps_w_a_miller.pdf
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Offline PCB.Wiz

  • Super Contributor
  • ***
  • Posts: 1670
  • Country: au
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #7 on: June 30, 2024, 02:02:13 am »
CaptDon, that sounds bad.

Would it be fair to say that if anyone actually made these calculators, they probably ended up disappointed by the reliability of the finished project?

There is already a massive hint to the poor stability, in the burn in and matching steps .....
The magazine article with the instructions (link can be found in the Hackaday article) talks about buying a bunch of neon bulbs, running them for 48 hours, then characterizing their striking and extinguishing voltages. Apparently, there's variation even among bulbs with the same part number, and there's variation over time (at least over an initial burn-in time).



Addit : this page has various tests of 'assisted' NEON designs, that sneak in 1 or 2 transistors, in breakover mode, to remove the neon variations.
https://www.dos4ever.com/ring/ring.html

Or you could use similar TRC logic with LEDs ?
« Last Edit: June 30, 2024, 07:48:39 am by PCB.Wiz »
 

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 3492
  • Country: fr
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #8 on: June 30, 2024, 05:03:43 am »
see extrensive neon lamp applications and papers from 1950s, and 1960s.

Neon Lamps: Signalite , NJ (made special lamps for electronics use) book by Ed BAUMAN

https://www.one-electron.com/Archives/Tube/Signalite/

Nixie numeric displays: Burroughs (the original inventor)

https://en.wikipedia.org/wiki/Nixie_tube

Jon
Jean-Paul  the Internet Dinosaur
 

Online Haenk

  • Super Contributor
  • ***
  • Posts: 1162
  • Country: de
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #9 on: June 30, 2024, 08:17:06 am »
Would it be fair to say that if anyone actually made these calculators, they probably ended up disappointed by the reliability of the finished project?

The only calculator that I know of using tubes (miniature tubes and a dekatron tube, plus selenium rectifiers) were Sumlock Anitas. The other manufacturers of that time went fully semiconductor. The Anitas were known for their extremely high failure rate, still they sold rather well, for a short time.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21972
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #10 on: June 30, 2024, 12:51:05 pm »
Generic neons are... well for one thing nearly extinct, not to say unavailable, but production is very down AFAIK.  But availability aside, what is available (new or old) tend not to be the cleanest of vacuum tubes.  That is, due to impurities in the gas fill, electrodes, surface oxides and other contaminants, exact metal composition, maybe even some things about the glass -- the glow pattern, voltages, lifetime, and how everything varies during that lifetime, all all pretty dubious for circuit use.

Your starting point would be -- first, I mean you can buy regular tubes and test them, see if it works, at all, or for how long, maybe measure, characterize and sort them, separate out the most wobbly ones.  You'll find audio noise (flicker, sizzling, popping, etc.), unstable light output or pattern (glow coverage of the electrode(s)), variable strike and hold voltage, etc. vary from part to part, with operating conditions (Ia), and sometimes you get lucky and one is good as ever, and many are just kind of crap, when you need any kind of precision anyways.

The next best thing then, would be premium quality ones -- I doubt anyone is producing these anymore (but I guess a lot of searching around might be worthwhile?), but other than that, a look through an old e.g. GE glow lamp manual ought to produce ideas (many of the above posted links/references are excellent as well), and you can look up datasheets for voltage regulator, oscillator/switching/computing types, etc., and see if you can find like a carton or whatever of them floating around somewhere for cheap.  AFAIK, these are largely your bog standard NE-2's or whatever, just made with better materials and process control, so that the properties are stable, noise is low, and lifetime is acceptable.  There may also be types with getter included -- though they might not be leaded bulb types, but more proper pin-base tubes as such.  (I *think* you don't want voltage regulator tubes, 0G3/85A2/etc., as they're specified for fairly narrow current levels, for which the regulation voltage is nominal, and they may drift at currents outside that range -- may still be usable but be careful with pulse discharge sorts of circuits.)

Note, my assumption: a change in operating condition, may exacerbate chemical changes within the tube, that weren't controlled for in design and manufacture.  NE-2's are basically made to glow and that's it; and even that, they weren't exactly great at(!).  So, using them for VR, switching, etc. applications is very pot-luck, or dubious, or finicky and short-lived, and that's why special types exist (and even those weren't very good at it).

Tim
« Last Edit: June 30, 2024, 12:53:10 pm by T3sl4co1l »
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline LaserSteve

  • Super Contributor
  • ***
  • Posts: 1311
  • Country: us
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #11 on: June 30, 2024, 10:19:34 pm »
Many of the Russian bulbs I was buying are pre aged, they have a brown stain inside the envelope from massive  overdrive before and after processing. 

Aging, including a short, very aggressive aging, cures many problems.  I have a very large bag of same date code US made neons on
 my desk, how many do you think you'd need for an aging test?


Assuming you're some place I can easily ship to.

Steve

"What the devil kind of Engineer are thou, that canst not slay a hedgehog with your naked arse?"
 

Offline rea5245Topic starter

  • Frequent Contributor
  • **
  • Posts: 585
  • Country: us
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #12 on: July 01, 2024, 12:18:12 am »
I have a very large bag of same date code US made neons on my desk, how many do you think you'd need for an aging test?
Steve

Thank you for the offer, Steve. But at this point I'm skeptical that this calculator can work reliably, so I'm probably not going to make it.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21972
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Vintage calculator project: how variable are neon indicator lamps?
« Reply #13 on: July 01, 2024, 01:07:51 am »
You might still consider a hybrid circuit.  Here's a 7-tube LFSR, for example:



Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf