Wow, thank you, for that analysis ! I guess the short answer, is change the current slowly, in particular, from the OFF to ON. Slow being a few milliseconds, since these thing tend to strike in the few 10s of microseconds. Maybe, a bit longer for those bargraphs.
On Monday, February 11, 2013 6:40:22 AM UTC-8, AlexTsekenis wrote: > > Hi guys, > > I would like to contribute a few points to the discussion. > > In contrast to the IN-13, the IN-9 does not have an auxiliary cathode to > anker the glow discharge. This we all know. > However it does have a zirconium bump/spike at the bottom of the main > cathode which is otherwise made out of molybdenum. The combination of > Neon/Zirconium gives a lower breakdown voltage than Neon/Molybdenum, > everything else being the same. The purpose remains the same: the glow to > start at the bottom of the tube. > > For the auxiliary cathode or zirconium spike methods to be effective, the > voltage must be applied with a limited rate of increase. In other words, > turn up the voltage slowly. This gives the physics in the IN-9 tube (have a > read here http://www.saltechips.com/products/thermneon/theory.html) > enough time to strike at the bottom of the tube before the applied voltage > reaches the threshold where the rest of the cathode (molybdenum) strikes. > This could be why Tim is finding the tubes to strike more reliably with > half-wave rectified DC - the voltage rises slowly. I certainly don't think > the tubes have developed the sort of nostalgia we have for neon, but for > their original power supplies! :-) > > Assuming there is a glow going at the bottom of the tube, that still does > not guarantee that the bargraph will not 'break' if a fast rise in current > (i.e bar length) is allowed. By increasing the current the glow will try to > cover more of the cathode surface. The glow at the bottom of the tube is > providing ions locally, thus giving the glow a *preference *to continue > growing from there if enough time is given (back to formative lag). If the > current rises very quickly, distant positions on the cathode become > 'disconnected' and will strike independently. > > At this point I must mention that argon-filled tubes have a shorter > formative lag than purely neon tubes. This could be why Jon has had better > luck with neon IN-9s. > > I experienced this first hand when working with the IN-13. Even if the aux > cathode was active, the bargraph column would break when I wanted to > 'instantly' move from the bottom to the top of the tube. The first measure > taken was in hardware, by implementing an RC lowpass filter on the signal > that controls the bargraph length. This slows down the signal to the tube's > driver, and also forms part of a DAC. Have a look at the implementation on > page 4 here > http://www.saltechips.com/docs/thermneon-assembly-manual-v1.1-web.pdf > The second measure was in software, the bargraph was incremented in > thousands of small, fast steps, rather than in one big step. This > eliminated the column breaking up issue. > > Regarding cleaning by cathodic sputtering, I found there was significant > variation in a batch of 20 IN-13s. At the rated 5-ish mA, a few would light > up up to the middle, a few all the way to the top and variations thereof. I > was fortunate enough to come across and salvage a PS325 programmable HV > power supply, so I used that to pulse the tubes at 100% overvoltage and > overcurrent (280 V, 10mA) for 10ms every 1s. The current was limited to > 10mA. I used a higher voltage to minimise the formative lag and a pulsing > scheme to prevent overheating of the cathode. The current was chosen on a > guestimate basis, the physics of cathodic sputtering are above my head. It > took a minute or less for each tube. I am aware these tubes were heavily > sputtered at the factory to give a high purity cathode surface - crucial > for linearity. Briefly connecting two pieces of bare wire will work just as > fine, but either put a current limiting resistor in series or wear goggles. > > Hope you find this useful. > > Alex. > > > > > On Saturday, January 5, 2013 9:54:55 AM UTC, Alex wrote: >> >> Hi guys, >> >> I have been stock piling bar-graph tubes for a while now due to >> their relatively cheap price and potential for interesting looking clocks >> or other animated displays. >> Due to a few being smashed in a recent shipment I decided to sit down and >> work through testing all 450 or so IN-9's that I have to see what they are >> like. >> >> Now, the first interesting point is that it seems that the white topped >> IN-9 can be either Neon or Argon, it seems most of mine are Argon which is >> quite annoying as the neon is a much richer red! I actually paid a bit more >> for a couple of purple topped argon tubes assuming them to be a lovely >> purple colour hence I was disappointed to find they are identical to about >> 70% of my conventional IN-9's (and badly cathode poisoned)! >> >> This brings me to the second point and a quick question, has anyone got >> any advice on getting these things to behave a bit better, some have fairly >> epic cathode poisoning with the glow starting in the middle or snapping to >> / hugging the other end or making large jumps up the tube. Are these things >> usually quite awful or have I just got bad batches? I have tried burning >> some in at upto about 40mA which has resolved some of the minor issues but >> on some seems to make them even more keen to glow from the other end... >> >> I have a few hundred IN-13's to test as well soon and am just hoping they >> are not as bad as these have turned out to be... >> >> Any advice on these tubes would be appreciated, hopefully a 64 channel VU >> meter will follow before long ;-) >> >> - Alex >> > -- 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 [email protected]. To post to this group, send an email to [email protected]. To view this discussion on the web, visit https://groups.google.com/d/msg/neonixie-l/-/mAmqU-LP5_MJ. For more options, visit https://groups.google.com/groups/opt_out.
