Great, thanks a lot Mike !!!
Le dimanche 20 octobre 2024 à 14:00:47 UTC+2, Mike Mitchell a écrit : > I originally used QS30 Nixies, but soon replaced them with Z5600M. I've > been running them continuously for several years without a problem. > I used 1/4 watt through hole resistors with a 300v rating. I used 1/2 > watt for the dekatron anode resistor and the 4.7M power supply discharge > resistors because I could get them with a higher voltage rating. > I used 1N4006 diodes, 1N4007 will work too. The 47uf capacitors in the > the high voltage supply are rated at 250v, the 33uf capacitors in the high > voltage supply are rated at 450v. The 47uf capacitor in the 40v supply > (actually 39 volts) is rated at 63v. The other capacitors are ceramic, my > notes say 470nf @ 100v; 47nf @ 250v; 22nf @ 630v; 1.5nf @ 630v. > > Mike Mitchell > > On Friday, October 18, 2024 at 5:13:29 AM UTC-4 Benoit Tourret wrote: > >> Thanks a lot Mike. >> >> sorry for my English, it's not my native language. I mixed Which and >> Witch... maybe a Halloween effect??? ;-) >> >> I'm trying to imagine my card. I think I'll use 4 identical cards. >> >> each board should have two dekatrons and two nixies, wired basically as >> if it were a divide by 100 board. >> by changing the cathode wiring with jumper instead of resistors, I could >> use it for either hours, minutes, seconds, or even 50/60 Hz division. >> since I'm on 50 Hz, it might be fun to double the 50 HZ to 100 HZ with a >> diode bridge and divide by 100, I might have something prettier to >> display... >> >> all the time reset action, resetting the counter to 0 to 60 or 24 will be >> done on a smaller board plugged into the 4 main boards >> it will be easier to set the time with an ntp clock, in the near future. >> 50 Hz is pretty accurate here in France, but I want the clock to at least >> self-adjust to winter/summer time. >> an esp32 will do just fine. but I want the clock to calculate the time >> itself. >> >> I have a handful of IN-4s, but I wonder if they will last... the QS30 is >> similar quality. the IN-12 is good, but won't match the dekatrons. >> are the z560m mercury doped? >> what are the top view nixie tubes available that last a long time... ??? >> >> Excluding the power supply, >> for the resistors, will the classic 1/4w carbon ones do the trick? >> and ceramic for the capacitors? >> >> >> [image: 1.png] >> [image: 2.png] >> I should use smaller resistors and thru the board capacitors... it is >> just to get the idea. >> >> but.... but ... the A-101 and the nixie tube cannot have the same >> position for the connectors, they don't have the same height. !!!!! >> faire et défaire c'est toujours travailler !!! >> >> Regards, >> Benoit. >> Le lundi 14 octobre 2024 à 15:12:47 UTC+2, Mike Mitchell a écrit : >> >>> There are eight dekatrons in the circuit. The first (G1) divides the >>> 60Hz signal by 6, giving a 10Hz signal. To change it to support 50 Hz, >>> move the emitter of the PNP transistor of the reset circuitry to the "5" >>> cathode (K5) of the dekatron. The furthest to the right "Witch Tube" (sic) >>> note in the diagram refers to G1, the first divide-by-six dekatron. >>> The second "Witch Tube" note referencing K8, K9, and K5 are for manually >>> resetting the seconds counter to 59.8 seconds. K8 references G2, K9 >>> references G3, and K5 references G4. Pressing the seconds reset button >>> sends a 10Hz pulse to the indicated cathodes, setting them to 59.8. They'll >>> then roll over to 00.0 after the button is released. >>> >>> The high voltage target was at least 475v. I'm using a voltage tripler >>> with a nominal 120v RMS at 60 Hz supply, giving 120v * 1.414 * 3 = 509v >>> assuming no diode losses. I measure very close to 500v, due to the diode >>> losses and the 22-ohm current limiting resistor. It's a 5 watt fusible >>> link resistor I had laying around. The only polarized capacitors are in >>> the power supply section, supplying +500v, +170v, -170v, and +40v. The >>> zener diode for the +40v supply is actually a 39v zener, as 40v ones are >>> very hard to find. A 39v supply is close enough to 40v for the dekatron. >>> >>> The "reset counter 60Hz? or Witch Tube" (sic) note references G4, the >>> tens-of-seconds counter. When that counter reaches "6" it is reset back to >>> zero. The complex reset circuitry furthest to the left detects when G7 >>> reads "3" and G8 reads "1" (hour 13), sending a reset pulse to K1 of G7 and >>> K0 of G8, reseting them to hour 01. The tens-of-hours dekatron spends a >>> lot of time displaying "0", and I found that the K0 cathode would wear >>> down. To extend the life of the Dekatron I connected the odd cathodes to >>> the NPN driver for the zero cathode of the Nixie tube, and all the even >>> cathodes to the NPN driver for the one cathode of the Nixie tube. Since >>> all the even cathodes are connected together, sending a reset pulse moves >>> the glow to a random even cathode. Advancing the tens-of-hours from there >>> goes to a random odd cathode. Making the clock handle 24-hours instead of >>> 12 hours is more difficult, as the simple even/odd grouping has to instead >>> have three groups. In that case I'd leave the K9 cathode open and make >>> groupings of K0, K3, K6 for "zero", K1, K4, K7 for "one", and K2, K5, K8 >>> for "two". The reset connection to K3 on G7 would be moved to K4 on G7, >>> the connection to K1 on G8 would move to K2 on G8, and the connection to K1 >>> on G7 would move to K0 on G7. That would detect "24" on the hours and >>> reset to "00". You'll also need another NPN driver for the "2" on the >>> Nixie. >>> >>> Converting from 120v to 240v isn't as straight forward. You still need >>> at least 475 volts for the Dekatrons, at least 170v for the Nixies, and a >>> large negative voltage for the reset circuitry. The 39 volts for the >>> Dekatron gates can be derived with a zener from the Nixie supply. First >>> the voltage tripler has to be replaced with a doubler instead, giving 240v >>> * 1.414 * 2 = 679v. The Nixie supply become 240v * 1.414 = 339v. You'll >>> have to find higher-voltage capacitors for the supply filters and increase >>> the anode resistors for the Dekatron and Nixies. The transistors used are >>> rated for 300v (MPSA42 and MPSA92) and 450v (STQ3N45K3). Note that the >>> MOSFET part number is incorrect in the diagram, it's STQ3N45K3 and not >>> STQ3N4K3. >>> All-in-all I think the easiest thing to do is use a travel transformer >>> to change your supply from 240v to 120v, something like >>> https://www.amazon.com/Watt-Step-Down-Converter-International/dp/B077CFPD1T/ref=sr_1_42 >>> >>> I did the same thing for a thyratron ring counter clock I built, the >>> schematics were for a 240v 50Hz supply and I used a travel transformer to >>> step up the US 120v 60Hz supply to 240v 60Hz. I had to change a >>> divide-by-five counter to a divide-by-six, but that was easy in comparison >>> to modifying the circuit to use 120v instead of 240v. >>> >>> Mike Mitchell >>> >>> On Friday, October 11, 2024 at 1:58:11 PM UTC-4 Benoit Tourret wrote: >>> >>>> I tried to better understand mike's drawing. >>>> >>>> what is exactly the purpose of the Dekatron 1 (in red) >>>> I see G1 and G2 receiving the AC voltage. I guess it is 60 Hz? >>>> I just have to use the pin K5 for 50 Hz ? >>>> K1 is to reset the clock >>>> K0 should activate the mosfet driving the second tube at 10 Hz >>>> the second tube divide by 10 to send seconds. and so on until the end >>>> of the day. do not pull the clock in water after midnight !!! Gremlins >>>> inside !!! >>>> >>>> the columns are driven by tube 2 ? >>>> >>>> I am using 240 V, how should I adapt the power supply part ? >>>> I see 500V on the power part and 475V on the Dekatron anode, this the >>>> same ? >>>> >>>> [image: DSCN0009 copie.jpg] >>>> >>>> Well, I am far to be a Jedi... >>>> >>>> Regards, >>>> Benoit. >>>> Le dimanche 24 septembre 2023 à 19:29:33 UTC+2, Benoit Tourret a écrit : >>>> >>>>> [image: essai.jpg] >>>>> >>>>> >>>>> Trying to make something more ... modern ??? >>>>> >>>>> it is the left part of the schematics. without the resistors on the >>>>> A-101 ... I need to work on it. >>>>> I am not sure that this project is the best to begin with kicad... >>>>> Le samedi 23 septembre 2023 à 22:24:14 UTC+2, Benoit Tourret a écrit : >>>>> >>>>>> Thanks Mike and Terry. >>>>>> >>>>>> one of my friends cleaned up the plan a little, it's already easier >>>>>> to read. >>>>>> >>>>>> [image: DSCN0009 copie (002).jpg] >>>>>> >>>>>> I will see witch method I will use for connect the A-101. >>>>>> it will mainly depend on witch tube / socket I will use for the >>>>>> nixies... >>>>>> >>>>>> Regards, >>>>>> Benoit. >>>>>> >>>>>> Le samedi 23 septembre 2023 à 19:17:41 UTC+2, Mac Doktor a écrit : >>>>>> >>>>>>> >>>>>>> On Sep 22, 2023, at 2:44 PM, Benoit Tourret <[email protected]> >>>>>>> wrote: >>>>>>> >>>>>>> How did you made the A-101 sockets ? >>>>>>> >>>>>>> >>>>>>> Here's a complete guide to the method. It includes photos (scroll >>>>>>> down a couple of pages): >>>>>>> >>>>>>> https://threeneurons.wordpress.com/arduino/dekatron-shield-kit/ >>>>>>> >>>>>>> >>>>>>> There's a diagram that you can use to get the pattern for the holes: >>>>>>> >>>>>>> >>>>>>> I built this kit. Don't make the mistake of sliding the pins all the >>>>>>> way in the "tubes". It was a nightmare getting the Dekatron back out >>>>>>> after >>>>>>> soldering the female Molex connectors in place. I had to get a friend >>>>>>> to do >>>>>>> it for me. I would have destroyed it. >>>>>>> >>>>>>> >>>>>>> Terry Bowman, KA4HJH >>>>>>> "The Mac Doctor" >>>>>>> >>>>>>> https://www.astarcloseup.com >>>>>>> >>>>>>> Edward R. Murrow: “Who owns the patent on this vaccine?” >>>>>>> Dr. Jonas Salk: “Well, the people, I would say. There is no patent. >>>>>>> Could you patent the sun?”—*See It Now*, 12 April 1955 >>>>>>> >>>>>>> -- 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 view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/bf60fb68-60bc-41be-a988-8ff6df08c4dan%40googlegroups.com.
