No hard figures - it's dependent on a bunch of factors; principally manufacturing quality, how you use the tube, operating temperature and gas fill. If you keep the operating current within spec and preferably at the lower end, avoid the tube getting heated from its surroundings and above all else keep the glow moving around all of the electrodes, they will last a seriously long time. The electrical properties will change gradually - the maintaining voltage rises and the latitude around the voltages required for reliable stepping decreases. So you were absolutely right to point out the importance of taking note of these elements of the spec when designing - you can get away with a lot on a NIB tube, but progressively less with age. Of course if the application requires clear glass to see the lovely glow, then the tube may reach end of its useful life in that application as sputtering gradually obscures the view, and that may happen well before it ceases to step reliably.
I'm not surprised by gregebert's description of his A101 - his application sounds like a great recipe for a long dekatron life. The slow speed ones are tough as anything and they love to work! Just for context, I believe that the large majority of dekatrons working in the stores of the WITCH today are from the original complement of tubes the machine was built with at the start of the 1950s. Can't formally prove it of course, but the date codes are consistent with that. Also, as we've noted before on other Russian glow tubes, the guaranteed life spans on the datasheets are ridiculously conservative if the tubes are treated well. The A101 datasheet gives a 1000 hour life - his A101 has done 70x that! I have a similar experience with my IN9 clock - the prototype unit is still going strong on its original tubes after 120K hours - datasheet life is 1K hours. Note that we're talking here about use-related life limitation - the death in storage of the high-speed dekatrons is a whole different topic. Jon. On Tuesday, February 16, 2021 at 8:49:07 PM UTC Dekatron42 wrote: > No, unfortunately not, no hard figures - maybe Jon has. I've only read in > some books about dekatron construction (neon tube construction in general > where dekatrons are shown as special variations) that they have similar > lifespans compared to small neon lamps but that the complex design > complicates the failure modes/rate and also that keeping currents/voltages > within the values specified in the datasheet will guarantee that they work > as long as possible. > > It depends on when you say a dekatron fails, it can get a silvery finish > on the glass so you almost can't see the glow but it still works and it can > fail a lot earlier than that due to internal flash-overs due to sputtering > onto the ceramic material which means a current can flow in this sputtered > material as the distance is shorter there compared to other paths (I have a > few dekatrons with that failure and of course a few with a silvery look). > > I have to count my dekatrons one day to see if I have enough to build a > second Harwell WITCH..... ;) :) > > /Martin > > On Tuesday, 16 February 2021 at 17:16:07 UTC+1 gregebert wrote: > >> *Martin* - Do you know how long dekatrons last ? I have an A-101 running >> as a spinner in one of my clocks, and it's been going 60RPM 24/7 for >> almost 8 years with no signs of degradation. I use 30k dropping resistors >> at the cathodes, and the current is right at the spec value of 450uA, so >> that gives about 15V for "steering", and the driving waveform is 3-phase >> with overlap. I >> >> On Tuesday, February 16, 2021 at 4:15:05 AM UTC-8 Dekatron42 wrote: >> >>> There is one more dimension to think about when running a dekatron and >>> that is that over time it will degrade due to sputtering affecting the >>> electrodes and to maintain a correct counting when it ages you should keep >>> the voltages as described in the datasheet, especially the guide and bias >>> voltages as those are needed when the dekatron nears its end of life - it >>> is the same effect that is in play if you want to raise a dekatron from the >>> dead if it has been heavily used or just stored for a very long time >>> without use, raise the bias/pulse voltages and sometimes the anode voltage >>> to surpass the effects of the sputtering - effectively increasing the >>> current flow between the anode and the guides/cathodes. >>> >>> So even if dekatrons work at other voltages you will see the effects of >>> lower/incorrect voltages when they near their end of life, this has been >>> described in older litterature where the internals of dekatrons has been >>> disected in detail - you might just need some 10V below the glowing cathode >>> to move the glow on a new dekatron but as it ages it will stop working at >>> that voltage. >>> >>> /Martin >>> >>> On Tuesday, 16 February 2021 at 00:53:36 UTC+1 bung...@gmail.com wrote: >>> >>>> Those values were left over from the 5v circuit. I was more concerned >>>> with the level shifting from the PIC at 0 to +5v. >>>> Peter >>>> >>>> On Mon, Feb 15, 2021 at 6:48 PM Jon <deka...@nomotron.com> wrote: >>>> >>>>> I'd had a similar thought about increasing the bias resistors - it's >>>>> not necessary to run so much current (5mA) through that part of the >>>>> circuit. >>>>> >>>>> OK, so your proposed conditions are that the dekatron will see are >>>>> 400V anode to main cathodes; guides swinging from +25V to -25V with >>>>> respect >>>>> to main cathodes. Tube current will be about 345uA. >>>>> I've replicated these on my dekatron tester using a nice NIB GS10C/S >>>>> as the test subject and can confirm that they do work - the tube stepped >>>>> fine up at speeds up to over 4kpps once I'd woken it up a bit. >>>>> >>>>> Caveats: >>>>> 1) I only tried one tube. >>>>> 2) My circuit is rather different to yours so although the static >>>>> voltages are the same, the pulse shapes are almost certainly different. >>>>> And >>>>> we didn't even talk yet about pulse durations, so I've no idea what your >>>>> PIC is spitting out. But as long as you're not trying to cut things too >>>>> fine, there's lots of latitude to find patterns that work. Keep things >>>>> north of 100us per phase and you'll be fine unless you've got a really >>>>> reluctant tube. >>>>> 3) I was running at slightly lower current (300uA) so you've probably >>>>> got a bit more margin for speed than I had. Recommended operating >>>>> conditions are 325uA +/- 20%, so we're both inside that range. But a >>>>> little >>>>> more rather than less current is useful when pushing higher speeds. I'm >>>>> guessing though that you're looking for a much slower stepping speed for >>>>> this application though. >>>>> >>>>> If you've not already done so, I'd definitely second Martin's >>>>> recommendation to take a look at Michael Moorrees' dekatron work (he >>>>> looks >>>>> in here from time to time too). He used a couple of elegant design tricks >>>>> which simplify the interfacing of dekatrons to modern electronics / >>>>> microcontrollers - I've followed his approach in pretty much all the >>>>> stuff >>>>> I've built and it works a treat. Not to say that other approaches aren't >>>>> equally useful too of course. >>>>> >>>>> Jon. >>>>> >>>>> On Monday, February 15, 2021 at 9:39:13 AM UTC Dekatron42 wrote: >>>>> >>>>>> I'd raise the resistance to at least some 100k for the two bias >>>>>> resistors R5 & R6 in your diagram above. >>>>>> >>>>>> I'd also use the correct bias voltage and just use an MPSA42 with its >>>>>> emitter to common ground for the driver to simplify the circuit as >>>>>> Ronald >>>>>> Dekker and Michael Moorrees with their dekatron circuits. >>>>>> >>>>>> Different dekatrons need different bias and pulse voltages on the >>>>>> guide electrodes to count properly so accomodating for those >>>>>> requiremenst >>>>>> will remove a lot of problems and keep down the fault finding time. >>>>>> >>>>>> /Martin >>>>>> On Monday, 15 February 2021 at 03:05:12 UTC+1 bung...@gmail.com >>>>>> wrote: >>>>>> >>>>>>> You are absolutely right. 5v was not high enough despite Ronald's >>>>>>> and my confirmation that it would work. >>>>>>> I am changing to +25v and -25v and will have an optocoupler with the >>>>>>> diode driven from the 5v PIC and the transistor at the bottom of the >>>>>>> resistor between the two power supplies. It keeps it simple. I have to >>>>>>> order the optocoupler because none of my old ones have higher than 30 v >>>>>>> rating. >>>>>>> I will have to think about the extra two power supplies. Maybe I >>>>>>> can't avoid them. It will be a few days until I get the parts. >>>>>>> Peter >>>>>>> [image: Dekatron Circuit.jpg] >>>>>>> >>>>>>> On Sun, Feb 14, 2021 at 3:18 PM Jon <deka...@nomotron.com> wrote: >>>>>>> >>>>>>>> Interesting approach - not seen it rigged up quite like that. Let >>>>>>>> us know how it goes! >>>>>>>> >>>>>>>> My immediate question is whether there's a big enough potential >>>>>>>> difference between an inactive guide and a main cathode to get a >>>>>>>> reliable >>>>>>>> transfer forward from a deactivating G2 to the 'next' main cathode >>>>>>>> rather >>>>>>>> than back to the adjacent recently used G1 - 5V is much lower than the >>>>>>>> datasheet guide bias. Might be OK at slow stepping speeds with long >>>>>>>> guide >>>>>>>> pulses. Also the leading edge of your guide pulses is going to be >>>>>>>> fairly >>>>>>>> slow as Q1/2 come out of saturation and the guides are passively >>>>>>>> pulled >>>>>>>> down to the 'active' voltage. Most guide drive circuits use a NPN >>>>>>>> pull-down >>>>>>>> to the active state which creates a sharp leading edge and then a >>>>>>>> slower >>>>>>>> return to the inactive state. >>>>>>>> >>>>>>>> Jon. >>>>>>>> >>>>>>>> On Sunday, February 14, 2021 at 7:06:28 PM UTC bung...@gmail.com >>>>>>>> wrote: >>>>>>>> >>>>>>>>> Having finished the Amperex 8453 display I started on my Dekatron. >>>>>>>>> Thanks to all for the suggestions. >>>>>>>>> I read up on Ronald Dekker's clock project >>>>>>>>> https://www.dos4ever.com/decatron/decatronweb.html >>>>>>>>> and decided against a direct drive from a 74141 because, even if >>>>>>>>> it was practical, it would not look as good as using all the pins. It >>>>>>>>> would >>>>>>>>> look like the 8453 I just finished except without the number mask. >>>>>>>>> A few quick experiments showed that a -24v power supply was needed >>>>>>>>> for the easiest implementation.. This is my design. I will let you >>>>>>>>> know if >>>>>>>>> it works. >>>>>>>>> >>>>>>>>> A PIC drives the circuit: it starts with Q3 off to force a start >>>>>>>>> at 1. A high on R3 or R4 is the same as the switches in Ron's test >>>>>>>>> circuit >>>>>>>>> placing -24v on the guides. As my PIC sends the BCD for the other >>>>>>>>> displays >>>>>>>>> and clocks the E1T it will generate the sequence to advance or retard >>>>>>>>> this >>>>>>>>> Dekatron.. >>>>>>>>> [image: Dekatron Circuit.jpg] >>>>>>>>> >>>>>>>> -- >>>>>>>> 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/62979e37-ef21-46e1-9b7c-45a4c4080238n%40googlegroups.com >>>>>>>> >>>>>>>> <https://groups.google.com/d/msgid/neonixie-l/62979e37-ef21-46e1-9b7c-45a4c4080238n%40googlegroups.com?utm_medium=email&utm_source=footer> >>>>>>>> . >>>>>>>> >>>>>>> -- >>>>> >>>> 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/dQn3tFBYfoc/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/75f258db-83ea-4df3-9d50-52dcaed1461en%40googlegroups.com >>>>> >>>>> <https://groups.google.com/d/msgid/neonixie-l/75f258db-83ea-4df3-9d50-52dcaed1461en%40googlegroups.com?utm_medium=email&utm_source=footer> >>>>> . >>>>> >>>> -- 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+unsubscr...@googlegroups.com. To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/bcaa0282-1c44-4b37-9f6f-98845208efecn%40googlegroups.com.
